The Metallurgy of Gold (IA Metallurgygold00roserich)

THE LIBRARY
OF
THE UNIVERSITY
OF CALIFORNIA
GIFT OF
George C* Linton
THE METALLURGY OF GOLD.
NEW METALLURGICAL SERIES
EDITED BY
W. C. ROBERTS-AUSTEN, C.B., F.R.S.,

st and Assayer of the Royal Mint; Professor of Metallurgy in the
College of Science.
In Large Svo, Handsome Cloth. With Illustrations.
1. INTRODUCTION to the STUDY of METALLURGY.

By the EDITOR. THIRD EDITION.
"No English text-book at all approaches this in the COMPLETENESS with which the
most modern views on the subject are dealt with. Professor Austen's volume will be
INVALUABLE." Chemical News.
2. GOLD (The Metallurgy of). By T. KIRKE ROSE, D.Sc.,

Assoc. R.S.M., F.C.S., Assist. -Assayer of the Royal Mint. SECOND
EDITION.
" The Four
chapters on Chlorination, written from the point of view alike of the
practical man and the chemist, TEEM WITH CONSIDERATIONS HITHERTO UNRECOGNISED,
and constitute an addition to the literature of Metallurgy, which will prove to be of
classical value." Nature.
3. COPPER (The Metallurgy of). By THOMAS GIBB, Assoc.

R.S.M.
4. IRON and STEEL (The Metallurgy of). By THOMAS

TURNER, Assoc. R.S.M., F.I.C. Vol. L, IRON.
"A MOST VALUABLE SUMMARY of useful knowledge relating to every method and
stage in the manufacture of cast and wrought iron down to the present moment . . .
particularly rich in chemical details. ...

An EXHAUSTIVE and REALLY NEEDED
compilation by a MOST CAPABLE and THOROUGHLY UP-TO-DATE metallurgical authority."
Bulletin of the American Iron and Steel A
ssociation.
5. METALLURGICAL MACHINERY : the application of

Engineering to Metallurgical Problems. By HENRY CHARLES
JENKINS, Wh.Sc., Assoc. R.S.M., Assoc. M.Inst.C.E., of the Royal
Mint.
6. ALLOYS. By the Editor.
%* Other Volumes in Preparation.

04
3
30
THE
METALLURGY OF GOLD.
BY
T. KIRKE BOSE, D.Sc.,

ASSOCIATK OP THE ROYAL SCHOOL OP MINES; FELLOW OF THE CHEMICAL SOCIETY;
ASSISTANT ASSAYER OF THE ROYAL MINT.
BEINO ONE OF A SERIES OF TREATISES ON METALLURGY,

WRITTEN BY ASSOCIATES
OF TIIE
ROYAL SCHOOL OF MINES.
EDITED BY
fcrot TO. (L 1Roberts*Busten, <r.., JMR.S.
WITH NUMEROUS ILLUSTRATIONS.
SECOND EDITION.
LONDON:
CHARLES GRIFFIN AND COMPANY, LIMITED.
PHILADELPHIA : J. B. LIPPINCOTT COMPANY.
1896.
[All Rights Reserved.]

/ W/oU
PREFACE
IN preparing this edition, the whole book has been care-
fully revised, due attention being paid to some useful
suggestions contained in the various reviews of the first
edition. Eight new . illustrations have been added, in-
cluding the frontispiece, which is from a photograph kindly
lent by Messrs. Fraser & Chalmers. The rapid progress

made in improving the cyanide process has rendered it
necessary to make considerable alterations in, and additions
to, the chapters devoted to that subject, and a new chapter
on Economic Considerations has also been added. The
author desires to express his thanks to the old students
of the Koyal School of Mines and others, who have kindly

sent information on work connected with the metallurgy
of gold, now in progress in various parts of the world.
ROYAL MINT,
August, 1896.
327
INTRODUCTION TO THE FIRST EDITION.
As this is the first of a series of treatises devoted to indi-
vidual metals, which is being prepared under ray guidance
it may be well to offer a few introductory remarks.

Associates of the Royal School of Mines have taken their
full share in conducting mining and metallurgical operations
in all parts of the world, but, notwithstanding the wide
experience they have gained, no treatise claiming to give a
general account of the Metallurgy of Gold and adequately

dealing with modern processes could hitherto have been
attributed to a Student of the School. It may be claimed
that in this country Dr. Percy founded the literature of
"
Metallurgy, but his volume on Silver and Gold," although
unrivalled in accuracy of detail, is only a splendid fragment,
and gold is alone dealt with in the sections devoted to the
refining of bullion and to assaying. A large amount of

valuable information concerning new methods and machinery
used in the treatment of gold ores has appeared in various
official publications during the last twenty years,
and much
is either scattered over the pages of the scientific press or
incorporated in the proceedings of various learned

societies.
Any attempt, however, to study the subject as

a whole from
Vlll INTRODUCTION.
these sources of information would be hopeless for those

whose time is limited, or who have not good libraries at
their disposal.
Mr. Rose gained his practical experience of gold and

silver extraction in the Western States of America, and has
written a volume which should prove to be very useful,

as its careful and conscientious preparation entitles it to
confidence.
W. C. ROBERTS-AUSTEN.
ROYAL MINT, March, 1894.

PKEFACE TO THE FIEST EDITION.
IN the present volume an effort has been made to supply

a succinct summary of the existing condition of the
metallurgy of gold, for the use of students and others

who are interested in the industries connected with the
precious metals. Ithas been said that mill-managers as

a class have little to learn from books but, even if this
;
be so, they will still need to keep themselves acquainted

with the progress of their art in far distant countries.
To them the bibliography appendedwill be useful.
Although, in this book, brief accounts are given of some

typical machines, attention has been directed rather to
methods of procedure than to details of machinery, which
will be dealt with in a separate volume in the series.
The practice in particular extraction works has been

described at some length, as such information, wherever
it is available, is among the most valuable that can be
given. Some recently devised methods of great importance

in the metallurgy of gold, such as the MacArthur-Forrest
cyanide process, the new barrel chlorination process, and

the improved Gutzkow parting process, are given for the
first time in a manual. Particular attention has been
paid to the assay of gold bullion, the system in use at

the Royal Mint and the precautions necessary to ensure
X PREFACE.
the highest attainable accuracy being described. Little
space is devoted to the geographical and geological dis-

tribution of gold ores, as this ground has been amply
covered by other works. Moreover the processes of
smelting, leaching, and pan-amalgamation used in the

treatment of silver ores, whether they contain gold or not,
have either been omitted or merely sketched in outline,
as belonging to the metallurgy of silver.

I am indebted to the President and Council of the
Institute of Civil Engineers for leave to reproduce Figs.
Nos. 14, 16, 17, 32, 33, 35, and 36, and to Mr. John
Murray, for Figs. 58, 60, 61, 62, 63, 64, 66, and 67, which
are copied from Dr. Percy's Metallurgy. I also take
this opportunity of expressing my thanks to my colleague,
Mr. F. W. Bayly, and to other friends for their kind
assistance while certain sections were going through the
press.
In conclusion, I desire gratefully to acknowledge the
kindness of Prof. Roberts - Austen in giving valuable
advice throughout 'the progress of the work. It was

at his suggestion that the task was undertaken, and it
ishoped that it will contribute to the realisation of his

wish that the experience of School of Mines men should
be collected in a number of works which will together
form a comprehensive metallurgical series.
T. K. ROSE.
KOYAL MINT, March, 1894.
CONTENTS.
CHAPTER I. THE PROPERTIES OF GOLD AND ITS ALLOYS.

Xll CONTENTS.
CHAPTER III. MODE OF OCCURRENCE AND DISTRIBUTION OF GOLD.
PAGE PAGE
Forms in which gold occurs in Petzite, . 37
nature, 34 Nagyagite, . 37
Vein gold, . 34 Composition of native gold, 33
Placer gold and nuggets, 36 Geographical distribution of
Calaverite, . 37 gold, 38
Sylvanite, . 37 Origin of gold ores, . 41
CHAPTER IV. PLACER MINING SHALLOW DEPOSITS.
.42 .53
Placer deposits,
Methods
. .
of obtaining gravel
Tail-race,
Ground-sluice,
.
...
....
.
53
from shallow placers,
....
Appliances used in washing the
. 43 Booming,
Dry diggings, ....
...
53
54
Pan,
gravel, 44
44
Cement gravels,
Tail sluices, ....
....
54
54
Batea,
Prospecting trough,
... . .
44
45
Fly catchers,
River mining, .... 55
55
Horn- spoons,
Cradle, ....
....
45
45
1. River mining proper,
2. Dredging, ... . 55
57
Long-torn,
Puddling-tub,
Siberian trough,
...
...
46
46
47
3. Deep bar mining,
....
Methods of working Siberian
placers,
. . 57
58
Sluice, 49 1. Siberian sluice, . . 59
Sluicing, 50 2. Trommel, 60
Use of drops, . . .51 3. Pan, . . . .60
.61
,,
Cleaning-up,
mercury, ...
undercurrents,
...
. . 51 Beach mining, . .
52 New method of working shallow

52 placer deposits, .
.
. .61
CHAPTER V. DEEP PLACER DEPOSITS.
^Nature and
deposits, ....
mode of origin of
62
Supply of water,
Breaking down the bank,
71
74
gravels, ....
Distribution of gold in the
Origin of gold in the gravels,

66
67
Sluicing gravel,
Cleaning-up,
Disposal of tailings,
76
79
80
gravels,
Methods of working,
....
Minerals occurring in the
69
70
Drift mining,
Shaft
"
Hydraulic elevator,
81
8i
83
" 70 Economic conditions, 85
Hydraulicking,"
Commencement
tions, .... of opera-
71
Shallow placer deposits,
Deep
85
87
CONTENTS. X11L
CHAPTER VI. QUARTZ CRUSHING IN THE STAMP BATTERY.
Primitive methods of crushing

and amalgamating, .
XIV CONTENTS.
CHAPTER IX. CONCENTRATION IN STAMP MILLS.

PAGE PAGE
Concentration, . . 166 Duncan's, . . 174
Settling boxes, 168 Percussion tables, . 175
Classification according to size 169 Gilpin County concentrator, 175
Screens, 169 True vanner, .. 176
Pointed boxes, 170 Method of working, 180
Early concentrating machinery 171 Riffle belted, . 183
Blanket strakes, . . 172 Embrey concentrator, 183
Riffled sluices, 173 Liihrig vanner, . 184
Raising-gate concentrator, 173 Hartz jigs, . 187
Round buddle, 173 Pneumatic jig, . 188
Centrifugal concentrators,
Hendy s,
174
174 centrator, ....
Clarkson and Standfield's con-
188
CHAPTER X. STAMP BATTERY PRACTICE IN PARTICULAR LOCALITIES.

In California,
In Colorado,
On
.
.
.
.
.
free-mflling ores in Australia

.
.195
1 90
Zealand,
In Dakota,
....
In the Thames Valley, New-
. . . .206
203
and New Zealand, . .199 In the Transvaal, . . .207
CHAPTER XI. CHLORINATION THE PREPARATION OF ORE FOR

:
TREATMENT.
The Plattner process, . . 215 Elimination of arsenic and
Origin, . . . . 215 antimony, 232
Method of working at Reich- Use of salt, . 233
eustein, . . . .216 Losses of gold, 235
Modern practice in chlorination, 217 Mechanical furnaces, 237
Crushing, . . . .218 1. With mechanical stirre s, 238
Krom's rolls, . . .219 (a) O'Hara, . 238
Rolls at Rapid City, Dakota, 222 (*) Spence, . 239
stamps, ....
Comparison between rolls a ad
223 2.
(c)
With
Pearce Turret,
rotating bed,
240
240
Drying the ore,
Roasting, ....
Reverberatory furnace,
. . 224
225
226
.
.
3. Revolving cylinders,
(a)
(&)
BrUckner,
Hofmann,
.
.
243
243
244
Chemistry of oxidising roasting, 229 (c) White, . . 246
minerals, ....
Decomposition of various
231 Use
ing,
(d) White-Howell,
of producer gas in roast-
247
248
CHAPTER XIT. CHLORINATION: THE VAT PROCESS.
Construction of vats,
Charging-in, .
Generation of chlorine,
249
251
252
Amount
Leaching,
Precipitation of gold,
....
of chlorine required,
.
.
.
257
258
259
Impregnation, . 254 Cost of working, . . .261
Reactions in vat, 255
CONTENTS. XV
CHAPTER XIII. CHLORINATION : THE BARREL PROCESS.
History, .... Organic substances,

Sulphuretted hydrogen
XVI CONTENTS.
CHAPTER XVI. CHEMISTRY or THE CYANIDE PROCESS.

PAGE
Action of potassium cyanide on Re-precipitation of gold and
gold and other metals, . 333 silver in leaching vata, . 345
Decomposition
cyanide, ....of potassium
338
Testing strength of solution,
Strength of solution required,
. 345
347
Decomposition
boxes, .... in the zinc
339
Consumption of cyanide,
Methods of increasing the
. 347
Action of potassium cyanide

On metallic salts and minerals,
On oxidised pyrites, . .
340
343
cyanide,
The Hood process,
....
speed of action of potassium
. . .349
348
The soda, solution, . . .344 The Sulman-Teed process, . 351

Ores suitable to process, . . 352
CHAPTER XVII.
Pyritic smelting, 353
CHAPTER XVIII. THE REFINING AND PARTING OF GOLD BULLION.

General considerations, . 356 5. Reduction of silver
.
Refining, . . . .357 chloride, . . . 374

Composition of bullion, . 357 Parting by sulphuric acid, . 375
Melting furnace,
Crucibles,
Melting,
....
.
. .
. .
.
.360
359
360
1. Mixing and granulating
2.
alloys, .
Dissolving silver,
.
.
.375
. 376
Refining, . . . .361 3. Melting gold residue, . 378
Toughening,
Casting, ....
Losses of bullion, .
. . .
.
363
364
366
4. Precipitation of silver,
5. Crystallisation of sul-
phate of copper, . 379
. 379
Refining by sulphur, . . 368 Combined process, . . 380

Osmiridium in gold bars, . 368 Gutzkow process, . .381
Parting processes, . . . 369 New Gutzkow process, . 383
Cementation, .. . 370 Miller's chlorine process, 386 .
Parting by sulphide of anti- Original method, . . 387

mony, . . . . 370 Later improvements, . 389
Parting by sulphur, . . 370 Modern practice at Mel-
Nitric acid process, . .371 bourne, . . .392
1. Granulation of alloys, . 371 Modern practice at Sydney, 401
2. Dissolving granulations, 372 Refining brittle gold by
3.
4.
dues, ....
Treatment of gold resi-
Treatment of silver solu-

373
chlorine,
Parting by electrolysis,
. . .
.
403
404
tion, . . A . .374
-THE ASSAY OF GOLD ORES.

General considerations, .
'
407 Fluxes, . 412

Blowpipe assay, 407 Methods of operation, 414
Metallics, ....
Sampling and crushing the ore, 408
409
Roasting be lore fusion,
Cleaning slag, .
416
417
Fusion,
Assay-ton weights, .
....
Crucible method of assay, 410
410
411
^Treatment of base ores,
Cupellation,
Influence of base metals,
417
417
420
General charges, 411
CONTENTS. XV11
CHAPTER XIX. Continued.
PAGE i PAGE
Inquartation and parting, . 421 2. Amalgamation, 428
Examination ef assay 3. Chlorination, . 428
materials, . . 423 4. Whitehead's method, 430
Examination of cupel, . 423 5. Assay of pyrites, 430
Assay by scorification, . 424 (a) Svvartz's method, 430
Detection of gold in minerals, 426 (b) Stapff's ,, 430
Estimation of gold in dilute 6. Assay of purple of Cassius, 431
solution, . . .427 7. a Mint sweep, . 431
Special methods of assay, 427
1. Mixed wet and dry method, 427
CHAPTER XX. THE ASSAY or GOLD BULLION.
Parting assay, . . . .431 Assay of alloys of gold, silver

1. Selection of sample, . 432 and copper, 452
2. Preparation of assay-piece Effects of other metals, . 453
for cupellation, . 433 Assay of various gold alloys, 453
3. Cupellation, . . 436 A. Alloys requiring scorifi
Assay furnace, . . 436 cation, 453
Cupels, . . .438 Arsenic and antimony, 453
Method of operation . 438 Iron and manganese, 454
Flashing, . .439 Cobalt and nickel, . 454
4.
Temperature of muffle, .
Preparation of cupelled
440 Zinc,
Tin, .... 454
454
5. Parting,
(a)
....
buttons for parting,
In parting flasks,
.
.
443
444
444
Aluminium,
B. Amalgams,
C. Platinum group,
454
455
455
(b)In platinum trays, . 445 (1) Platinum, . 455
Relative advantages of (2) Palladium, . 456
these, . . .446 (3) Rhodium and iridium, 456
6. Weighing the cornets, . 447 D. Tellurium compounds, 457
ing, .....
Losses of gold in bullion assay-
Silver retained in cornets, .

447
449
Wet methods
Other methods,
1. Touchstone,
of assay, . 457
458
458
Occluded gases, . . . 450 2. Colour and hardness, 459
Checks or proofs, . . . 450 3. Density, . 459-
Limits of accuracy in assay, . 451 4. Spectroscope, . . 459
Parting by sulphuric acid, . 451 5. Electrolysis, . 460
Preliminary assay, . . . 452 6. Induction balance, . 460-
Assay by cadmium, . . . 452
CHAPTER XXI. ECONOMIC CONSIDERATIONS.
Management of gold mills, 462 . Table of production in different

Cost of production of gold, . 462 countries, . . . 467
Annual production of gold, past Amount produced by different
and present, .. . 464 processes, . . . 469
Consumption of gold, . . 469
XV111 CONTENTS.
BIBLIOGRAPHY.
General,
...
....
Periodicals,
PAGE
471 Roasting,
472 Chlorination,
.
.
. .
<
.
.
.
.479
PAGE
478
Properties of gold and its alloys 473 Cyanide process, . . . 480

Distribution of gold, . 475 Pyritic smelting, . . . 480
Placer mining, .
Crushing and amalgamation,

Concentration, .
.
.
476
477
478
Assaying, ....
Refining and parting of bullion, 480
480
INDEX, 483
THE METALLURGY OF GOLD,
CHAPTER, L
THE PROPERTIES OF GOLD AND ITS ALLOYS.
PHYSICAL AND CHEMICAL PROPERTIES OF GOLD.
Introduction. Some of the more obvious physical properties of

gold were already well known and taken advantage of at a very
early period in the history of man, long before any exact science
could be said to exist, and it is of interest to the metallurgist
to remember that the earliest dawn of the science of chemistry
was heralded by the study of the properties of gold, and by the
efforts which were made to invest other matter with these
properties. From the fourth to the fifteenth century, chemistry,
which was first called " chemia," and then " alchemy," was
defined as the art of transmuting base metals into gold and
silver, almost all the labours of philosophers being intended to
aid directly or indirectly in solving this problem. At the end
of this period, while Paracelsus was giving to chemistry a new
aim that of investigating the composition of drugs, and their
effect on the human body Agricola was reducing to order the
numerous empirical facts which together made up the art of
metallurgy, and although alchemy died hard, its era of usefulness
may be said to have ended here. Gold has been accused, with
some justice, of having been the cause of many of the wars and
marauding expeditions from which the world has suffered, but
it may be asserted with equal truth that it has been instru-
mental, in a far greater degree than most other commodities, in

promoting the growth of civilisation, the efforts of the alchemists
having laid the foundations of the science of chemistry, and
those of the gold-seekers having resulted in the discovery of
new countries, and in the spread of knowledge of all kinds.
Colour. The lustre and fine colour of gold have given rise
to most of the words which are used to denote it in different
languages. The word "gold" is probably connected with the
Sanscrit word "jvalita" which is derived from the verb "jval"
to shine. It is the only metal which has a yellow colour when
1
2 THE METALLURGY OF GOLD.
in mass and in a state of purity. Impurities greatly modify this

colour, small quantities of silver lowering the tint, while copper
raises it. In a finely divided state, when prepared by volatilisa-
tion or precipitation, gold assumes various colours, such as deep
violet, ruby and reddish-purple, the tint varying to brownish-
purple and thence to dark brown and black. This purple colour
has been supposed by some experimenters (viz., Guy ton de
Morveau, Biichner, Desmarest, Creuzbourg and Berzelius) to be
due to the formation of a coloured oxide of gold of unknown com-
position, but Buisson, Proust, Figuier and, more recently, Kriiss
have shown that no oxygen can be obtained from this coloured
material, and that it probably consists of metallic gold. Similar
colours are seen in purple of Cassius, and in Roberts- Austen's
purple alloy of aluminium and gold, the colour in each case being
probably due to a particular form of finely divided gold. The
ruby coloured gold, suspended in a liquid in which it has been
precipitated by ether and phosphorus, is present in the most
finely divided state obtainable, and perhaps it then approaches
the condition of separate atoms. Such gold shows no tendency
to settle by gravity even after being kept undisturbed for several
years. Somewhat less finely divided gold gives a faint blue tinge
to light transmitted through the liquid in which it is suspended.
Gold precipitated from its solution as bromide is in a different
molecular condition from that formed from chloride, giving out
3 '2 calories in passing into the latter state.* The surface colour
of small particles of native gold is often apparently reddened by
being coated with translucent films of oxides of iron. Very thin
plates of gold are translucent, and appear green by transmitted
light, while remaining yellow by reflected light. On heating,
the green colour changes to ruby red, but is restored by the
pressure of a hard substance by which the state of aggregation
is again altered (Faraday). Molten gold is green, and its vapour
is alsoprobably greenish.
Malleability and Ductility. Malleability and ductility
are possessed by gold at all temperatures to a far higher
degree than by any other metal. Asingle grain of gold can
be drawn out into a wire over 500 feet long, and leaves of
not more than 8 X of, an inch in thickness can be obtained
by beating. Faraday has shown that the thickness of these

leaves may be still further reduced by floating them in a
dilute solution of potassic cyanide by which they are partly
dissolved.
Hardness. The hardness of gold lies between that of alu-
minium and that of silver, corresponding to the number 979 in
Bottone's scale, in which the diamond is 3,010.
Tenacity. The purest gold obtainable has a tenacity of 7 tons
*
Thomson, Thermochemische Untersuchungen, vol. iii., p. 412.
THE PROPERTIES OF GOLD. 3
per square inch and an elongation of 3O8 per cent., but the
presence of minute traces i.e., -g-^-g- of other elements, especially
those having high atomic volumes e.g., bismuth, tellurium, lead,
<fec., greatly lowers these constants, as well as the malleability and
ductility of the metal, while its hardness is increased.* Gold

containing înnr of bismuth can almost be crumbled in the
fingers.
Specific Gravity. The specific gravity of gold when precipi-
tated from solution by oxalic acid is 1949 (G. Rose) f when ;
cast it varies from 19-29 to 19-37, but this can be raised by

compression to over 19 '48. J Henry Louis has shown that the
" "
specific gravity of unannealed parted gold (i.e., the residue
left after boiling silver-gold alloys in nitric
acid) is 20-3, its
density being lowered by the process of annealing, a fact which
Louis considers is indicative of the existence of an allotropic
modification of the metal.
Cohesion. On heating, gold can be welded like iron below
the point of fusion, and finely divided gold agglomerates on
heating without being subjected to pressure. Pressure alone is
also sufficient to make gold dust cohere, while a true flow of the
particles of gold can be induced in the case of the pure metal
and some of its alloys.
Specific Heat. The specific heat of gold is -0324 (Regnault)
or -0316 (Violle).
Fusibility. Gold fuses, after passing through a pasty stage,
at a clear cherry-red heat, just below the fusing point of copper
and much above that of silver. The metal expands considerably
on fusing and contracts again on solidifying. Carnelley gives
the temperature of fusion as 1,037, Yiolle as 1,045, whilst
recent determinations have given it as 1,061-7 (Heycock and
Neville), 1,050 to 1,060 (Le Chatelier), and 1,072 (Holborn
and Wien).
Spectrum. In the gold spectrum Huggins saw 23 lines, the
wave lengths of the most important ones being 523-1, 583-5, and
627-6 respectively. 1 1
Latent Heat. The latent heat of fusion of gold is 16-3, and

the normal lowering of the freezing point for 1 atom of impurity
in 100 atoms of gold is 10-6, but the presence of from O'l to
4-0 per cent, silver does not cause any alteration in the freezing
point.
Magnetism. Gold is diamagnetic, its specific magnetism
being 3-47 (Becquerel), if that of iron is taken as 100.
* Roberts- 339.
Austen, Phil. Trans. Royal Soc., vol. clxxix. (1888), p.
t Pogg. Ann., vol. Ixxiii. (1848), p. 1, and vol. Ixxv. (1848), p. 403.
J Eighth Report of the Royal Mint, 1877, p. 43.
Trans. Am. Inttt. of Mng. Eng., Chicago Meeting, 1893.
||
For full information on the spectroscopic characteristics of gold, see
clxiv. (1874), part 11.,
Lockyer and Roberts, Phil. Trans. Royal Soc., vol.
40.
p. 495; and Fre"my, Ency. Chim., voL iii. (1888), L'or, p.
Conductivity and Expansion. Its electrical conductivity

is 76 '7, and its thermal conductivity 53-2
(or, according to Gray,
77), that of silver being 100 in each case.
3
Its coefficient of
linear expansion is 0-0000144 between O and 100.
Atomic Weight and Volume. Its atomic weight, formerly
believed to be about 196 '2, has been more recently given by
Kruss as 196-64, by Thorpe and Laurie as 196-85, and by Mallet
as 196-79. It is probable that 196-8 is not far from the truth.
The atomic volume of gold is 10 '2.*
Volatilisation of Gold. The boiling point of pure gold has
not been determined ; calculated according to Wiebe's formula it
would be about 2,240, or nearly 500 above the melting point of
platinum.f However, contrary to the belief of the older experi-
menters (Gasto Claveus, and others), it is sensibly volatile in air
at far lower temperatures. Robert Boyle was unaware of this
fact, but Homburg gilded a silver plate in 1709 by holding it over
gold strongly heated in the focus of a burning mirror (Encyclo-
pcedia ritannica, 1778, and Gmelin's-Handbuch), and St. Claire
Deville volatilised and again condensed gold when melting it
with platinum. The rapid volatilisation of gold, when heated by
an ordinary blowpipe, was first proved in 1802 by Dr. Robert
Hare, of Philadelphia (TillocKs Magazine), a purple stain being
thus produced on bone-ash in a few seconds. Lastly, a discharge
of high-tension electricity from gold points causes its volatilisa-
tion, and if the discharge is sent through a fine gold wire
stretched on paper, it converts it into a purple streak of finely
divided condensed particles of the metal. The rapid distillation
of gold caused by heating it in a current of air of considerable
velocity, such as that furnished by a blow-pipe, by which the
liquid is thrown into waves, may be shown at any time by
heating a fragment of the precious metal of the size of a pin's
head 011 a bone-ash cupel in the oxidising flame of a good mouth
blowpipe. Almost immediately after the fusion is complete, a
purple stain of condensed gold begins to form 011 the outer
margin of the cupel. The author has found that a piece of gold
weighing 0*5 gramme loses half its weight in an hour, if heated
on a cupel by a foot-blowpipe (the temperature attained being
probably less than 1,300), and only a few minute beads are
observable, detached from the main button. Alloys of copper
and gold disappear much more rapidly. No doubt most of the
gold passes off as spray, but perhaps part of the loss may be due
to rapid volatilisation, and could not be correctly described as
mechanical loss.
The both when pure and when alloyed with
volatility of gold,
silver and copper, has been investigated by Napier, \ who found
*
Introd. to the Study of Metallurgy, by Prof. Roberts- Austen, p. 58.
t L. Meyer, Mod. Theories of Chemistry, p. 134.
$ Chem. Hoc. Journ., vol. x. (1859), p. 229; and vol. xi., p. 168.
,
THE PROPERTIES OP GOLD. 5
that an alloy of 100 parts gold to 12 parts of copper, if

kept for
six hours at a temperature just high enough to
keep it melted,
lost 0-234 per cent, of its gold contents, and at the tem-
highest
perature attainable in an assay muffle, it lost 0'8 per cent, in six
hours. An increase in the amount of copper present caused an
increase in the loss of gold. In the simple operation of pouring
about 30 Ibs. of a gold-copper alloy from a graphite crucible into
moulds, fumes were given off, of which the part condensed in a
wet glass beaker held above the crucible contained 4 '5 grains of
gold. Napier also found that gold does not appear to volatilise
so readily when alloyed with silver only, as when copper is
also present.
Makins found that gold volatilises sensibly along with silver
and lead, when melted with these metals in a muffle in an
ordinary bullion assay. The loss of gold by volatilisation on
melting its copper alloy is the common experience of mints. At
the Sydney Mint it was estimated to be 0-017 per cent., or 170
per million sterling melted, and is probably seldom less than
0-01 per cent., or one part in ten thousand. At the same
Mint, Leibius found that the sweepings from the top coping-stone
of a chimney 70 feet high contained 1 -46 per cent, of gold and
6-06 per cent, of silver. Similar results have been obtained at
some other mints.
A number of experiments were recently made at the Royal
Mint by the author,* with the view of determining the effect of
variations in the temperature and other conditions on the vola-
tility of gold and some of its alloys. The test pieces were heated
in a muffle furnace, the temperature of which was determined by
the optical pyrometer, and by the Le Chatelier thermo-couple,
which consists of platinum and rhodio-platinum wires. The
results of some of the experiments on fine gold, and certain
alloys of gold and copper, are given in the table on the next page.
The following conclusions may be drawn from the table :
1. The loss of gold on heating the pure metal rises with the
temperature, being four times as great at 1,250 as at 1,100,

whilst it is insignificant at 1,075 and probably nil at 1,045, the
melting point (Violle).
2. An atmosphere consisting largely of carbonic oxide is
apparently favourable to the volatilisation of gold, the rate being

six times greater than in an atmosphere of coal gas at the same
temperature. The increased loss when the graphite crucible was
substituted for a cupel may be noticed in this connection ; clay
crucibles of similar shape to the graphite one have an opposite
effect, the loss being less than on cupels.
3. The increase of loss of gold alloyed with copper, when the
percentage of the latter metal is increased, which is observable

in one of Napier's experiments, is confirmed.
*
Chem. Soc. Journ., vol. Ixiii. (1893), p. 714.
6 THE METALLURGY OP GOLD.
as
la
O r- 1
CO CO
S 04
000
co o o
c<i
lO ô
Oo
oo
O biGOOO
S ? 8
OOO^H
<MIOOO
bbbb
r-iCO CO
So.,*
|b
o
fer
^ *
(N
' 52 *
the
of
gns -d
a
â ' -g
S fl
C3
"3
=6
'.
S
rf cS
^ -g _g g
sO wO^.O O
Nature
Atmosphere.
00^0 s
O 00 w s
, THE PROPERTIES OP GOLD. 7
The volatilisation of gold-copper alloys merits further in^

vestigation,as it is of great industrial importance. The high
percentage of the losses shown in the table is, no doubt, due to
the smallness of the masses treated, and to the comparatively
large surfaces exposed in consequence of this.
A number of experiments were also made at the same time on
various gold alloys to determine the relative effect of the presence
of other metals on the volatilisation of gold. The temperatures
and conditions used were similar to those described above. It
was found that the volatility of gold was increased by the pre-
sence of any metallic impurity, even by the non-volatile metals,
such as platinum. The tellurium alloys suffered the heaviest
losses, gold containing five per cent, of tellurium losing from
16-5 to 39*5 parts per 1,000 by volatilisation per hour at a
temperature of 1,245 (i.e., 200 above the melting point of pure
gold), a point of much importance in connection with the metal-
lurgy of telluric gold. Lead and platinum had a very slight
effect in increasing the volatility of gold; copper and zinc a more
marked effect, while five per cent, of antimony or mercury
caused losses amounting to about two parts per 1,000 of gold per
hour at 1,245. Further conclusions which were drawn were as
follows :
1.Although the temperatures employed were higher than

those at which pure zinc, cadmium and tellurium, and probably
antimony and bismuth, are distilled, yet these elements were
never completely volatilised. The " temperature of dissociation,
of the alloys," as in the case of the bismuth-arsenic alloy
investigated by Edward Matthey in a paper read before the
Royal Society, January 26, 1893, is much higher than the
ordinary distillation point of the more volatile constituent.
Thus zinc boils at 950, but gold alloy loses little or no zinc
its
at 1,120, and still retains part at 1,250, whilst antimony was
not driven off to any extent by the highest temperature attained.
Copper appears to pass off more easily than some of the so-called
volatile metals, perhaps because the dissociation of its alloy with
gold may not form an initial stage of the operation.
2. The amount of gold lost depends partly on the volatility of
the alloying metal, but perhaps too much stress has been laid on
this factor in the past, the results of heating alloys of mercury,
zinc, antimony and copper pointing to that conclusion. A
metal
with a strong attraction for gold, such as copper, may carry it
off, perhaps as a vaporised alloy, more easily
than one which
mixes with it less intimately.
3. It is observable that those impurities which reduce the
surface tension of a button of liquid gold appear to increase the
vapour pressure of the metal as indicated by the loss on heating.
This was to be expected.
4. A current of air or coal gas insufficient to disturb the
surface of the liquid metal does not appear to increase the

volatilisation.
It may be added that Hellot stated that if an alloy of one
in air, the whole
part of gold and seven parts of zinc is heated
of the gold comes off in the fumes.
Crystallisation of Gold. Gold crystallises in the cubic
in the form of cubes,
system, occurring frequently in nature
octahedra and rhombic dodecahedra. Cleavage is never exhibited.
Single detached crystals are comparatively rare,
and the crystals
are usually attached end to end, forming strings, and branching,
arborescent, or moss-like masses, which are composed of micro-
Fig. 3.
Fig. 2.
scopic crystals, usually octahedra. These forms occur frequently

in quartz veins, but the single crystals, which are usually of larger
size, viz., from ^ inch to 1J inch in diameter, are mainly found
in drift deposits. They are rarely perfect or of brilliant lustre,
although such crystals were found at the Princeton Gold Mine,
Mariposa County, California, but occur more frequently with
rounded angles, raised edges, and cavernous faces, which are often
marked with parallel striations, and possess little or no lustre
(Fig. 1). The octahedra found in California are usually flattened
parallel to two opposite faces, or elongated, or otherwise dis-
torted. Still more frequently they are only partially developed,
as in Figs. 2 and 3. In all these cases " the incomplete crystals
THE PROPERTIES OP GOLD. J)
have the appearance of a failure for lack of material"

(W. P.
Blake). Crystals of greater complexity, containing many modi-
fying faces, occur chiefly in Siberia, Transylvania and Brazil.
The most common forms occurring naturally in Australia are
the octahedron and the rhombic dodecahedron.*
Artificial crystals can be obtained in several
ways, but with
great difficulty. The slow cooling of an ingot of gold from fusion
usually gives faces and sometimes angles of octahedra on the
surface of the metal, fThe presence of small quantities of copper
prevents this crystallisation. Feathery crystalline plates are
precipitated in the electrolysis of a solution of chloride of gold
and ammonium. By keeping an amalgam containing 5 per cent,
of gold at a temperature of 80 for eight days, and then di-
gesting it at 80 with nitric acid of specific gravity 1'35,
and subsequently subjecting the residue to a red heat, bright
crystals of gold which are, however, usually microscopic can be
obtained. J In the Percy collection are some gold crystals found
in the mercury troughs at the foot of the " blanket strakes," in an
amalgamation mill. The troughs are placed so as to catch any
stray particles of gold that may pass the blankets. As the
amount of gold recovered in this way is very small, it is not
worth while to clean out the troughs frequently, and in this case
they had remained undisturbed for nine months, at the end of
which time all the amalgam was found to be crystallised. The
mercury has been dissolved off by nitric acid, and the gold
crystals remain. The smaller crystals are rather indefinite in
shape, but amongst the larger ones (which are about half the
size of a pea) are well-defined combinations of the octahedron,
rhombic dodecahedron, and cube.
Solubility of Gold. Gold is readily soluble in aqua regia,
or in any other mixture producing nascent chlorine, among such
mixtures being solutions of (1) nitrates, chlorides, and sulphates
e.g., bisulphate of soda, nitrate of soda, and common salt;
(2) chlorides and some sulphates e.g., ferric sulphate; (3) hy-
drochloric acid and potassium chlorate; (4) bleaching powder
and acids, or salts such as bicarbonate of soda. The action
is much more rapid if heat is applied or if the gold is alloyed
with one of the base inetals than if it is pure. The presence of
silver in the gold retards the process, a scale of insoluble chloride
of silver being formed over the metal, and the action may even-
tually be completely stopped if the percentage of silver present
is large. Gold is also dissolved by chlorine and bromine, but
the action of both of these is much slower than that of aqua
regia, and subject to the same difficulties if silver is present ;
*
For a full account of the crystalline forms of native gold, see a Paper
byYV. P. Blake, in Precious Metals of the U.S.A., 1884, p. 573.
t Chester in Am. Journ. of Science and Arts, vol. xvi., July, 1878, p. 29.
Krafft, Encyclopaedia Brit., article "Crystallisation/'
10 THE METALLURGY OF GOLD
heat assists the dissolution. Iodine only dissolves gold if it is

nascent, or if heated with gold and water in a sealed tube to 50.
Metallic gold dissolves in hot strong sulphuric acid, especially
if a little nitric acid is added (the precipitated metal dissolving
most readily), forming a yellow liquid, which, when diluted with
water, deposits the metal as a violet or brown powder. The
solution also becomes covered with a shining film of reduced
metal on exposure to moist air. On addition of hydrochloric
acid or a metallic chloride, auric chloride is formed, no longer
precipitable by water.* Gold is also attacked when used as the
positive pole of the battery in the electrolysis of strong sulphuric
acid, but is immediately reduced again by the evolved hydrogen, f
According to Nickles,j the easily decomposable metallic per-
chlorides, perbromides and periodides are capable of dissolving
gold, lower chlorides, &c., of the base metals being formed, and
gold chloride, &c., produced. The higher chlorides and bromides
of manganese and cobalt (Co 2 Cl 6 &c.) act well, and a hot strong
,
aqueous solution of ferric bromide or of

ferric chloride will also
readily dissolve gold. Ferric iodide is decomposed by gold

under ordinary conditions, aurous iodide being produced. Some
other haloid compounds only attack gold in the presence of
ether, in which case even hydriodic acid itself is decomposed
and aurous iodide formed. Selenic and iodic acids have also
been mentioned as solvents for gold, and the effect of a mixture
of nitric and nitrous acids is described in Chap. xx. Alkaline
sulphides attack gold slowly in the cold, and more rapidly if
heated, producing sulphide of gold which is subsequently dis-
solved. Ditte's observations, however, tend to show that alka-
line polysulphides have no action on metallic gold.
Spring has shown that gold is soluble in hydrochloric acid if
heated with it to 150 in a closed tube, and is subsequently
reduced by the liberated hydrogen and deposited as microscopic
crystals on the side of the tube. C. Lossen pointed out in 1895
that if a solution of potassium bromide is electrolysed, the re-
sulting alkaline solution, containing hypobromite and bromate
of potassium, is capable of dissolving gold. Gold is dissolved
by aqueous solutions of simple cyanides and by certain double
cyanides, such as sulphocyanides and ferrocyanides, which act
very slowly except in presence of oxidising agents and with the
aid of heat.
Preparation of Pure Gold. The purest gold obtainable is
required for use as standards or check pieces in the assay of

gold bullion. The following method of preparing it was adopted
by Roberts-Austen in the manufacture of the Trial-Plate, by
* Watt's Diet.
ofChem., Supplement, p. 652.
t Spiller, Chem. Neiv*, vol. x., p. 178.
$ Ann. Ch. Phys. [4], vol. x., p. 318.
Zeitschr. anorgan. Chem., vol. i. (1S93), p. 240.
PROPERTIES OF GOLD. H
which the imperial gold coinage is tested.* Gold
assay cornets,
from the purest gold which can be obtained, are dissolved
in nitrohydrochloric acid, the excess of acid
expelled, and
alcohol and chloride of potassium added to
precipitate traces of
platinum. The chloride of gold is then dissolved in distilled
water in the proportion of about half an ounce of the metal to
one gallon, and the solution allowed to stand for three weeks.
At the end of this time the whole of the precipitated silver
chloride will have subsided to the bottom, and the
supernatant
liquid is removed by a glass siphon. Crystals of oxalic acid
are then added from time to time, and the liquid
gently warmed
until it becomes colourless, when precipitation is
complete, a
point reached in three or four days if ten-gallon vessels are used.
The spongy and scaly gold so obtained is washed repeatedly
with hydrochloric acid, distilled water, ammonia, and distilled
water again, until no reaction for silver or chlorine can be ob-
tained, after which it is melted in a clay crucible with bisulphate
of potash and borax, and poured into a stone mould. Lack of
care in any one of the operations will result in gold containing
one or two parts of impurity in ten thousand.
With regard to the above method, it may be observed that
carefully purified sulphurous acid gas is a more convenient pre-
cipitant than oxalic acid, and may be substituted for it without
any ill effects, as any foreign metals that may be present are in
such small quantities that their sulphites, even if formed, would
remain dissolved. It should be added that, according to the
recent researches of Kohlrausch, Rose, and Holleman, silver
chloride is soluble in 600,000 to 700,000 parts of pure water at
the ordinary temperature. It follows that, under the condi-
tions given by Roberts- Austen, the solution contains at least
0'3 to 04 part of silver per 1,000 parts of gold, and this pro-
portion is doubtless higher in practice, owing to the greater
solubility of silver chloride in solutions of other chlorides than
in pure water. There can be no doubt that part of this silver
is precipitated with the
gold, and that, by redissolving and
reprecipitating, a purer product is obtained The amount of
silver remaining in solution can, moreover, be reduced to about
one-fifth of the amount noted above by adding a small quantity
of hydrobromic acid to the solution, silver bromide being far
less solublethan silver chloride. Another additional precaution
is remove the gases taken up by the gold during the process
to
of melting by heating it to redness in vacua.
Allotropic Forms of Gold. Little is known of these. The
marked influence of traces of other metals on the properties of
gold has already been touched on from this and from the
;
variations in colour and other properties the existence of

several allotropic modifications of gold might be inferred. la
*
Fourth Annual Report of the Mint, 1873, p. 46.
alloys containing appreciable quantities of other metals,

evidences of allotropy are not met with so frequently. The
potassium alloy, however, containing 10 per cent, of gold, on
being attacked by water, leaves a black finely-divided gold
powder, and there is reason to believe that this combines with
water to form a hydrate.*
Wilm f states that if gold is dissolved in dilute sodium
amalgam under water, the aqueous liquid becomes dark violet,
and when this is acidulated with hydrochloric acid, a black pre-
cipitate of pure gold is obtained. The black gold differs from
the ordinary modifications in its extreme lightness ; moreover,
it is soluble in alkaline solutions, and does not amalgamate with
mercury or with sodium amalgam. When heated, it yields the

ordinary modification as a violet red powder. This form of gold
appears, from Wilm's account, to resemble the black precipitate
obtained on digesting certain aluminium-gold alloys with hydro-
chloric acid.
ALLOYS OF GOLD.
Gold can be made to alloy with almost all other metals, but
most of the bodies thus formed are of little or no practical
importance. Tin, zinc, arsenic and antimony unite with gold
with contraction, and form pale yellow or grey coloured, hard,
brittle and easily fusible alloys, of which all, except those con-
taining zinc, are soluble with difficulty in aqua regia. The
arsenic and antimony alloys are slowly decomposed by mercury,
the base metal being separated as a black powder, which con-
sists in part of arsenide or antimonide of mercury. Lead and
iron alloy with gold with expansion, while in the case of copper
no change of volume takes place.
Gold alloyed with a small percentage of lead is a hard, brittle,
pale-yellow substance, which can be crumbled with the fingers.
If more than about 4 per cent, of lead is present, there is
marked segregation on solidification, and this also takes place
in the case of the zinc alloy and of some others.
Heycock and Neville have shown J that the freezing point of
lead is lowered by the addition of gold to it in accordance with
the general law. Thus, the freezing point of pure lead being
327, an addition of 3-8 per cent, of gold reduces it to 301, and
Roberts-Austen has recently found that the eutectic alloy of
gold and lead, which contains about 13 per cent, of the former
metal, melts somewhere between 190 and 198. Similarly, by
adding 6 '9 per cent, of gold to thallium, the freezing point of
the latter is lowered from 301 to 261.
*
Introd. to Study of Met., p. 91.
t Zcitschr. anorgan. Chem., vol. iv. (1893), p. 325.
J Chem. Soc. Journ., vol. Ixv. (1894), p. 72.
ALLOYS OP GOLD. 13
Several alloys of gold with other metals in molecular propor-

tions have been isolated, or their existence proved in various
ways. Thus, for example, the compound AuSn was recognised
by Mathiessen, from the curve of electric conductivity of the
gold-tin series of alloys, and this substance was more recently
detected by A. P. Laurie by observing the E. M. F. developed
by alloys of different compositions when dipped into a solution
of SnCl.2 .
Heycock and Neville succeeded in isolating the corn-
pound AuCd in 1892, after having suspected its existence for
some time. In the same year, Mylius and Fromm prepared a
number of gold alloys by precipitation from solution.* Thus
a gold-zinc alloy, containing equal weights of the two metals,
and approximating in composition to AuZn 3 was obtained in
,
the form of black, spongy flocks, by adding a solution of gold

sulphate to water in which a zinc plate was placed. The gold-
zinc slimes obtained in the cyanide process (q.v.} may be com-
pared with this. Gold-cadmium, similarly obtained, is a lead-
grey crystalline precipitate, having the composition AuCd 3 On .
heating, it is converted into gold-mono-cadmium, AuCd (see

above). If the gold-zinc alloy is shaken with a solution con-
taining a cadmium salt, the gold-cadmium alloy and a zinc salt
are obtained. Similarly, a copper plate, acting on solutions
containing gold, yields a black, spongy compound of gold and
copper ; and gold-lead and gold-tin alloys in the form of black
slimes are also readily prepared.
The aluminium alloys have been investigated by Roberts-
Austen, f These alloys are remarkable for their intense colour,
varying from yellowish-green to purple, and some of them appear
to present the characteristics of true chemical compounds. A
white alloy containing 10 per cent, of aluminium is very hard,
and has a melting point no less than 417 lower than that of
pure gold ;
but a deep purple alloy, containing 22 per cent, of
aluminium (corresponding to the formula AuAl 2 ), appears to
melt at a temperature between 1,065 and 1,070, or over 20
higher than the melting point of pure gold. It presents, there-
fore, the extremely rare case of an alloy, the fusing point of
which is higher than that of the least fusible of its constituents,
and this fact affords strong evidence that it is a true compound
of gold and aluminium. It is hardly necessary to point out that
the melting points of ordinary chemical compounds are often
much higher than the melting point of the least fusible con-
stituent. There is a strong tendency for this purple alloy to be
formed when gold is melted with an excess of aluminium, and
the result is that the alloys rich in aluminium usually show a
marked lack of homogeneity. Evidence of the existence of the
compound AuAl 2 was also obtained by Heycock and Neville I
*
Chem. Soc. Journ., vol. Ixvi., part 2 (1894), p. 236. Nature, vol. xliv.
(1891), p. 111.
t Proc. Roy. Soc., vol. 1. (1892), p. 367. + Loc. cit.
by noting the effect on the freezing point of tin caused by addi-

tions of gold and aluminium in various proportions. crystal- A
line alloy of gold and bismuth, containing gold 68'22 per cent.,
bismuth 31-78 per cent, has been prepared by Pearce by means
of liquation. By melting this with silver, a crystalline alloy
corresponding to the formula AuAg was produced, and crystal-
line alloys of gold, silver and copper were also obtained.*
Specimens of these products are in the Percy collection.
The diffusion of gold into other metals, both liquid and solid,
has lately been investigated by Roberts- A usten.f The rates of
diffusion in liquid metals are of the same order as those of
soluble salts in water, but the diffusion in solids is very slow.
If gold is placed at the base of a cylinder of solid lead 70 mm.
high, some is found to have reached the top in thirty days, the
temperature being kept at 251. The rate of diffusion is still
measurable at 100, but is almost inappreciable at ordinary
temperatures.
The alloys which are most important to the metallurgist are
those which gold forms with mercury, with copper, and with
silver.
Amalgams. Mercury rapidly dissolves gold at ordinary tem-
peratures, forming liquid, pasty, or solid amalgams, according
to the proportions of the metals present, and their purity. A
piece of gold rubbed with mercury is immediately penetrated by
it and becomes exceedingly brittle. The ductility is not always
restored when the mercury is removed by distillation, a crystal-
line structure being often induced. Some forms of precipitated
gold are not readily taken up by mercury, the particles tending
to float on the surface of the latter. An amalgam containing
90 per cent, of mercury is liquid, that containing 87 '5 per cent,
is pasty, and that containing 85 per cent, crystallises in yellowish-
white easily fusible prisms. On dissolving precipitated gold in
mercury heated to 120, and then cooling the mass, white
crystalline plates having a composition corresponding to the
formula AuHg4 separate out.| Amalgams with smaller propor-
tions of mercury can be obtained in various ways by heating
gold and mercury to different temperatures up to a low red heat,
and acting on the products with nitric acid. Gold amalgam
dissolves readily in excess of mercury, forming a liquid mass
from which it may be partially separated by straining through
chamois leather, when a white pasty amalgam containing about
33 per cent, gold remains behind, while the mercury which
filters through contains some gold, the amount varying with the
temperature, but not with the pressure applied. Kasentseff has

shown that this liquid amalgam contains O'll per cent, of gold
* Trans.
Am. Inst. Mng. Eng., 1885.
t Proc. Roy. Soc., vol. lix., p. 281. Bakerian Lecture, Phil. Trans., 1896,
Fremy, Ency. Chim., vol. L'or, p. 99.
iii.,
Bull. Soc. Chim. [2], vol. xxx., p. 20.
ALLOYS OF GOLD. 15
if it is filtered at0, 0*126 per cent, at 20, and 0-650 per cent,
at 100 C.; these amalgams, therefore, behave like aqueous
solutions.
When amalgams are gradually heated, the mercury is distilled
off by degrees, the action soon ceasing if the temperature is
allowed to become stationary, and distillation recommencing if
it is again raised. At 440 (somewhat below a red heat), an
amalgam containing about three parts of gold to one of mercury
is obtained, and at a bright red heat almost all the mercury is
expelled, and if the heating has not been pushed too rapidly the
vapours contain but little gold. The gold obstinately retains
about 0-1 per cent, of mercury, which is not driven off below the
melting point of gold.
Gold and Silver. Gold and silver unite in all proportions,
yielding alloys which are harder, more fusible, and more elastic
than either metal. The hardest is that containing two parts of
gold to one of silver. The colour of gold is sensibly lowered by
the addition of very small quantities of silver, and on increasing
the proportion of the latter, the colour changes by tints of a
greenish-yellow (when from 20 to 40 per cent, of silver is pre-
sent) to white, with a scarcely perceptible yellow tinge (when
50 per cent, of silver is present), and silver- white (when more
than 60 per cent, of silver is present). Pearce has obtained
in regular octahedra the alloys corresponding to the formulae
Au 8 Ag, Au 6 Ag, andAu 2 Ag by liquation, and Levol has
obtained AûAg, AuAg and AuAg5 in perfectly definite
2,
homogeneous forms. The alloys containing small quantities,

(lessthan 20 per cent.) of gold liquate readily if kept for some
time in a state of quiet fusion, an alloy containing one part of
gold to five parts of silver (AuAg 9 ) sinking to the bottom, and
slightly auriferous silver floating at the top. The silver-gold
alloys most used in jewellery are green gold (silver 25, gold 75),,
dead-leaf gold (silver 30, gold 70), arid the alloy containing 40
per cent, of silver, Triple alloys of gold, silver, and copper
are emploj-ed far more frequently by English jewellers than
those last mentioned; .of these the alloys, consisting of 22-, 18-,
15-, 12-, and 9-carat gold respectively, can be hall marked.*
Alloys of gold and silver were much used for coinage before the
methods of parting became well known and inexpensive.
JSlectrum, which includes pale yellow alloys with from 15
to 35 per cent, of silver, and which -occurs native, was much
used for ornaments and coins by the Greeks and Romans, and
by the nations which acquired their arts. The use of silver
in the gold-copper coinage alloys was not discontinued until
quite recently, all English guineas and the Australian sovereigns
manufactured at Sydney up to the year 1871 containing some
* For
22-carat gold contains || by weight of fine gold, and so on.
details of these alloys see Gee's Goldsmith's Handbook, pp. 41-52.
of it. Both nitric and sulphuric acids attack silver-gold alloys,

almost completely dissolving out the silver if it is present in
amounts variously stated at from 60 to 75 per cent, or over,
while, if the proportion falls below 60 per cent., some of the
silver is left undissolved with the gold. Hydrochloric acid
scarcely attacks these alloys, and the action of aqua regia is soon
arrested if the proportion of silver is considerable.
Gold and Copper. These metals dissolve in one another in
all proportions, forming a complete series of homogeneous alloys,
which are less malleable, harder, more elastic, and more fusible
than gold, and possess a reddish tint. Those with less than
12 per cent, of copper are fairly malleable; when more than
this is present they are more difficult to work owing to their
hardness. Since no change of volume occurs when these alloys
are formed, their densities may be calculated from those of gold
and copper. The density of standard gold is 1748, and that of
the alloy containing gold 900, copper 100 is 17*16. For the den-
sities of other gold-copper alloys, see Rigg in the Report of the
Mint, 1876, p. 46. Many of the alloys have been used for coinage
at various times. The Greeks and Romans, after electrum had
fallen into disuse, employed the purest gold they could procure,
viz., that from 990 to 997 fine. Under the Roman Emperors,
however, copper was intentionally added, and in the two cen-
turies preceding the fall of Rome very base alloys were used,
some containing only 2 per cent, of gold, or even less.* In the
"
middle ages these base alloys were discarded, and the " byzant
of Constantinople and the "florin" of Florence were both nearly
pure gold, while the first gold coins struck by the nations of
Western Europe were also intended to be absolutely fine. The
standard 916*6 or \ (i.e., 916*6 parts gold in 1,000) was adopted
by England in the year 1526, the standard of 994*8, which had
been introduced in 1343, being finally abandoned in 1637; the
900 standard was introduced in France in 1794, and subsequently
adopted in other countries. These two standards are now those
most commonly used, the English standard being employed by
Russia, Portugal, India, and Turkey, and the French standard
by most other civilised countries the Austrian ducat, however,
;
has a fineness of 986 and that of Holland a fineness of 983, while

the Egyptian standard is only 875. Of all these alloys the
900 and 91 6 -6 standards are those best adapted for coinage,
keeping their colour fairly well, and resisting wear better than
richer alloys. The 900-alloy is harder and wears better than
the 916'6-alloy, but the difference is not great, the rate of wear
depending less on such small differences of composition than on
the mechanical and thermal treatment of the alloys during the
operation of coining.! The alloys used in coinage generally
*
La Monnaie dans VAntiquite. Paris, 1878.
t Fourteenth Report of the Royal Mint, 1884, p. 45 and seq.
ALLOYS OP GOLD. 17
contain from eight to twelve parts of silver per 1000 in addition

to the gold.
Gold-copper alloys tarnish on exposure to air owing to oxida-
tion of the copper, and blacken on heating in air from the same
cause. This oxidised coating may be removed and the colour of
fine gold (not that of the original alloy) produced
by plunging
the metal into dilute acids or alkaline solutions, the operation
"
being technically known as blanching." The colour of alloys
may be improved without previous oxidation by dissolving out
some copper by acids, a film of pure gold being thus left on the
outside which can be burnished. French jewellers use a hot
solution of two parts of nitrate of potash, one part of alum,
and one of common salt for this purpose.
Nitric or sulphuric acid dissolves out the copper from gold-
copper alloys under conditions similar to those under which
it removes silver from silver-gold alloys. If the copper falls
below 6-5 per cent, the alloy is not attacked by these acids
(Pearce). Aqua regia dissolves all the alloys completely.
Liquation of Gold Alloys. The subject of liquation
generally, including that observed in gold alloys, has been dis-
cussed in the volume introductory to this series. * Experiments
of Roberts- Austen and of Peligot are there described which tend
to prove that liquation does not occur in the gold-silver and
gold-copper alloys rich in gold and free from all impurities.
Levol had previously come to the same conclusion as Peligot
with regard to gold-copper alloys, but gave no details of
his experiments, remarking, however, that the oxidatior. of
the copper made the research difficult. Roberts-Austen has
also cited t the evidence afforded in the preparation of the
standard gold trial plate made in the Royal Mint in 1873 as
conclusive in proving that considerable masses of standard gold
can be obtained of uniform composition. A
mass of standard
gold weighing 72 ozs. was cast in a suitable mould, and rolled
into a plate 37 inches long and 6 '5 inches wide ; portions of
metal were cut from different parts of it, and when these were
assayed it was found that the greatest variation between any
two assays was T7jlhr?F5 there being no evidence of concentration
of the precious metal anywhere. In this case the gold and copper
used to make the alloy were of exceptional purity. In 1889,
moreover, this chemist J examined the composition of two
large ingots, weighing 400 ozs. each, which had been sent to
the
Mint for coinage by the Bank of England. The fineness of these
ingots was 896-2 and 978-5 respectively, and the results showed
that no definite liquation had taken place in them. In the case
of many of the ordinary trade ingots, however, the discrepancies
between assays on piecestaken from opposite ends of the bars
* Roberts-
Austen, Introd. to the Study of Met., p. 72.
t Nineteenth Report of the Royal Mint, 1888, p. 54. $ Loc. cit.
2
j>rove that the composition is not uniform, but this lack of uni-
formity is doubtless due to the presence of some other element in
addition to gold, silver and copper.
In support of this view it may be mentioned that Louis
Janin Jr. instances the case of three ingots from an Idaho mine,
which were melted with borax in plumbago pots, and on cooling
showed evidences of liquation.* Dip samples were taken after
vigorous stirring, and granulated, and assay cuts were also taken
from the diagonally opposite corners of the top and bottom faces
of the bar after pouring. On assaying these samples the fol-
lowing results were obtained :

CHEMISTRY OF THE COMPOUNDS OF GOLD. 19
the centre of the sphere, and that this fusible alloy contains
much gold but very little copper. Arnold's micrographic re-
sults* tend to confirm the view which follows from this that
the brittleness of crystalline gold is due to the presence of films
composed of such eutectic alloys separating the crystals of gold
from each other.
It has been shown by Edward Matthey f that when gold ingots
containing members of the platinum group are cooled from a
state of fusion an alloy rich in the more fusible element (gold)
falls out first, driving the less fusible constituent to the centre.
Thus the assay of an outside cut of such an ingot gives a result
too high in gold, sometimes by several per cent. It has long
been known, moreover, that iridium and osmium become concen-
trated towards the bottom of the mass. The reason for this
is that, at the temperature of fusion of gold, these
refractory
elements, either free or alloyed with gold, sink in the molten
metal and are left in the state of small crystalline particles.
CHAPTER II.
CHEMISTRY OF THE COMPOUNDS OF GOLD.

Compounds of Gold. Gold is characterised chemically by an
extreme indifference to the action of all bodies usually met with
in nature. Its compounds are formed with difficulty, and decom-
pose readily, their heat of combination being in general small,
while some are formed by endothermic actions. The result of
this condition of things is that gold is found in nature chiefly
in the metallic form, and the mineralogist has, therefore, few
compounds to consider. Nevertheless, the laws governing the
formation of artificially-prepared compounds of gold are of great
importance to the metallurgist, as the processes of extraction
are all based on these laws, and in particular, some knowledge
of the reactions and general behaviour of gold in its compounds
is essential to those who are engaged in extracting gold by wet
processes. Ashort account of some of the compounds of gold is,

therefore, appended, special attention being paid to those bodies,
which are most likely to be of interest to the metallurgist.
Gold forms two series of compounds, having the general
formulae AuR and AuR 3 while doubtful compounds corre-
,
sponding to AuR.,, AuR4 and AuR 5 have been declared to

, ,
exist by Thomsen, Prat, Figuier, and others. These two series

are denominated aurous and auric respectively.
*
Engineering, vol. Ixi. (1896), p. 176.
t Proc. Roy. Soc., vol. li. (1892), p. 447, and Phil. Trans. Roy. Soc. t
vol. clxxxiii. (1892), A. pp. 629-652. See also Proc. Roy. Soc,, 1896.
Compounds of Gold with the Halogens. Gold forms two

series ofcompounds with the halogens, the general formulae
being AuR and AuR 3 respectively. Supposed compounds having
the formula AuR 2
have been described, but are probably mix-
AuR
tures of the series denoted by andAuR .
3
All these bodies
are very unstable, existing throughout only low ranges of tem-
perature, whether in the dry state or in aqueous solution. The
chlorides are the most easily formed and the least unstable, the
bromides coming next, as might be predicted from the heats of
combination which are given in the subjoined table in calories:
CHEMISTRY OF THE COMPOUNDS OF GOLD. 21
temperature AuCl 3 began to decompose at about 180 in spite of

the presence of an excess of chlorine, and protochloride of gold,
AuCl, of a greenish-yellow colour was formed. At 220 the
AuOl was completely decomposed, leaving metallic gold, while
up to this temperature a little AuCl 3 continued to be sublimed,
but towards 300 all action ceased, and the gold remained un-
attacked by the chlorine at all higher temperatures. On lowering
the temperature, Kriiss observed in succession the formation of
AuCl at 220, and of AuCl 3 at 180.
The exact point at which gold ceases to be attacked by chlorine
and the rate of volatilisation of the chloride are of great import-
ance in connection with the loss of gold on roasting auriferous
materials with salt.* The matter has, therefore, been lately
investigated by the author f with the following results. Gold
unites with chlorine if placed in the gas at atmospheric pressure
at all temperatures up to a white heat, but the subsequent
decomposition of the chloride is rapid above 300. The absorption
of chlorine by gold with the formation of chlorides at first in-
creases in rapidity as the temperature rises, and reaches its
maximum at about 225. The fact that gold is attacked by
chlorine, and that the chloride is subsequently volatilised at all
temperatures between 180 and 1,100, was proved by means of
Deville's hot and cold tubes, which enable part of the sublimed
chloride to be collected. The rate of volatilisation at various
temperatures is as follows :
and only about O02 per cent, in 30 minutes at 1,100,* or about

one-hundredth part of that volatilised in chlorine.
In the table the amounts " lost by volatilisation include the
"
amounts recovered from the water tube and the gold condensed
on the inside of the outer tube the latter sublimate was not
;
recovered separately after each experiment.

In every case the gold recovered from the water tube was
associated with a little less chlorine than was required to form
gold trichloride, and therefore presumably contained either some
metallic gold or AuCl, or both ; these doubtless resulted from
the dissociation of some of the trichloride when in the form of
vapour.
The amounts volatilised vary according to two different factors
(1) The vapour pressure of gold trichloride,
AuCl 3 which of ,
Fig. 3a.
course increases continuously as the temperature rises ; and (2)

the pressure of dissociation of the trichloride, which also rises
continuously with the temperature,
but not at the same rate as
the vapour pressure. The rise of vapour pressure tends to raise,
and that of the pressure of dissociation to reduce, the amount of
gold volatilised as chloride.
The vapour pressure increases more
dissociation at temperatures below
rapidly than the pressure of
tem-
300, and also above 900, but less rapidly at intermediate
Hence the curve 3a) showing the variation of
peratures. (Fig.
volatilisation with is irregular, passing through a
temperature
* vol. Ixiii. (1893), p. 717.
Chem. Soc. Journ.,
CHLORIDES OF GOLD. 23
maximum near 300, and a minimum at a point somewhere

below the melting point of gold. The first-named change in the
direction of the curve possibly occurs at the melting point of the
chloride,namely 288. The second change is perhaps caused by
the change of sign of the heat of formation of the trichloride
AuCl 3 when this becomes negative, the pressure of dissociation
;
of the compound would decrease, in accordance with the law of

van't Hofl and Le Chatelier. However this may be, it is certain
?
that when gold is heated in chlorine at atmospheric pressure,

trichloride of gold is formed and volatilised at all temperatures
above 180, up and probably far beyond, 1,100.

to,
The usual method adopted for the preparation of auric chloride
is to dissolve gold in aqua regia, and then to evaporate the
liquid to dry ness, keeping the temperature above 100 to pre-

vent the formation of the hydrochloride AuCl 3 HCl. A brownish-
red mass is thus formed, consisting of AuCl 3 mixed with more
or less of the protochloride and of hydrochloric acid. On taking
up with water the protochloride is decomposed into gold and
trichloride, but the hydrochloric acid can only be eliminated by
shaking with ether, which withdraws the trichloride from its solu-
tion in water. If an attempt is made to drive off the hydrochloric
acid by heat, a partial decomposition of the trichloride results.
Auric chloride exists both in the anhydrous state and in
combination with two equivalents of water, AuCl 3 2 2 O, when . H
it occurs in orange-red crystals. The anhydrous salt is of a
brilliant-red colour, crystallising in needles belonging to the
triclinicsystem and melting at 288 under a pressure of two
atmospheres of chlorine. It can be prepared by drying the
hydrated salt at 150. The anhydrous and hydrated
salts are
both hygroscopic, and dissolve readily in water with elevation of
and in
temperature ; they are also soluble in alcohol and ether,
some acid chlorides, such as AsCl 3 SbCl 5 SnCl 2 SiCl 4 &c.
, , , ,
Auric chloride is readily decomposed by heat. Lowe states*

that 4 grammes of the trichloride, when heated in a porcelain
basin on a boiling water bath, can be completely transformed
into the monochloride, although not until after the lapse of
several days. On the other hand, as has already been mentioned,
Kriiss states that the decomposition of auric chloride, in an
atmosphere of chlorine, begins at 180. According to the
is observed to suffer
experiments of the author,! auric chloride
slow decomposition at as low a temperature as 165 in an
1-6 per cent, being
atmosphere consisting of chlorine, about
converted into monochloride in four hours at this temperature ;
the decomposition is about five times more rapid at 190. The
observed at 100, although
decomposition in air can be readily
it does not seem to be so rapid as was indicated by
Lowe. In
* vol. cclxxix., p. 167.
Dingler's polyt. Jown., 1891,
t Ghem. Soc. Journ., vol. Ixvii. (1895), p. 902.
seven days only 6 '6 per cent, of the trichloride was decomposed,
the initial rate of decomposition being 0-041 per cent, per hour.
At 165, however, the initial rate of decomposition appeared to
be 3-2 per cent, per hour, and the conversion into monochloride
was complete in four or five days at 160 and in ten hours at 190.
The rate of decomposition of the trichloride in air at various
temperatures can be calculated from the above data by the help
of Harcourt and Esson's formula a a^
= ( Tl / r where
y )
, 15 2
are the rates of decomposition at the absolute temperatures

rv rg respectively. The value of the constant ra for this
chemical action is found to be about 27, and by substituting
this number for m in the equation, the rate of decomposition of
gold trichloride is calculated to be 0-365 per cent, in a year at
15. The decomposition begins to be observable at 70 in air,
when the monochloride is formed, about twenty- five years,
however, being required for the complete change at this tem-
perature. The observed rate of decomposition shows that a
similar change would require about 1,000 days at 100, while it
results from calculation, using Harcourt and Esson's formula,
that at 200, thirty-six hours, and at the melting point namely,
288 less than one minute suffices for the complete decomposi-
tion of AuCl 3 in air.
Whether dry or in aqueous solution, the trichloride is also
decomposed by light, gold being deposited in scales in the latter
case, but the presence of free hydrochloric acid prevents this
decomposition. Weak voltaic currents precipitate metallic gold
from the solution of the trichloride upon the negative pole. The
solution of trichloride of gold is also decomposed by many re-
ducing agents, such as most organic substances, metals and
protosalts; heating the solution in every case hastens the
decomposition. The reduction by organic matter is assisted by
the action of light, which is especially efficacious in the presence
of starchy and saccharine compounds, or of charcoal or ether.
In direct sunlight the last reagent deposits a bright mirror of
metallic gold, but under ordinary conditions excessively finely
divided gold (Faraday's gold) is precipitated. Alkalies also
quicken the action of organic matter, and it may be said that
all organic compounds reduce gold chloride on boiling in the
presence of potash or soda,* while Miiller states that a mixture

of glycerine and soda lye is one of the best precipitants for
gold chloride, separating the metal completely in highly dilute
solutions. According to Kriiss, if potash and soda are quite
free from organic matter, they have no action on solutions of
auric chloride, whether cold or hot. If a small quantity of
organic matter is present, sub-oxide of gold is precipitated if ;
larger quantities are present, both metallic gold and sub-oxide

* vol. 16e Cahier, p. 74.
Fr&ny, Ency. Chim., iii.,
CHLORIDES OP GOLD. 25
are precipitated in the cold, but gold alone at boiling point.

Alkaline carbonates are without action on cold solutions, but if
they are hot, then half the gold is precipitated as hydrate, while
the other half remains in solution in the form of a double
chloride of gold and the alkali. The precipitation by means
of charcoal is of especial importance in view of its adoption in
practice.
It has been stated that a current of hydrogen gas will preci-
pitate gold completely, especially on boiling the solution, but
Kriiss has proved that, if the hydrogen is quite pure, it has no-
effect either on cold or hot solutions. Sulphur, selenium, phos-
phorus and arsenic all precipitate gold on boiling the solution
of the trichloride. Many metals reduce chloride of gold, the
action being, of course, most rapid in the case of the most highly
electro-positive metals, such as zinc and iron. It seems probable
from the author's experiments * that turnings of these metals
would make good precipitants for use on a commercial scale in
the chlorination process. Lead sometimes gives fine dendritic
plates of gold. Sulphuretted hydrogen precipitates sulphide of
gold from both neutral and acid solutions, all traces of gold being
readily removed from a solution by this reagent, whilst phos-
phoretted, arseniuretted, and antimoniuretted hydrogen all
precipitate metallic gold. The lower oxides of nitrogen,
"
nitrous acid and many other " ous acids and oxides effect the
same decomposition. Sulphur dioxide is a convenient reagent,
and is often used in the laboratory, being almost equally effica-
cious in cold and hot solutions. The reaction is
2AuCl 3 + 3S0 2 + 6H 2 = 2Au + 6HC1 + 3H 2 S0 4 .
Various protosalts also reduce trichloride of gold. Ferrous

sulphate is often used to detect the presence of gold in solution
as chloride this reagent gives the solution a pale blue colour
;
by transmitted light, and brown by reflected light, owing to the

formation of finely divided precipitated gold. The reaction is.
represented by the following equation
2AuCl 3 + 6FeS0 4 = 2Au + 2Fe 2 (S0 4 ) 3 + Fe 2 Cl 6 .
To a dilute solution for gold, a test tube filled with the

test
liquid held in the hand side by side with a test tube filled
is
with distilled water and a few drops of a clear solution of ferrous
sulphate are added to each. On looking down through the
length of the test tubes from above, with a white surface as
background, any slight changes of colour may be detected by
comparison and the liquids may also be compared with the
original solution in a test tube. In this way, by a little prac-
tice, the presence of gold in the proportion of only T^Vinr
(1 dwt. per ton of water), or even less
can be detected. The
* See Section on " of Gold," xiii.
Precipitation Chap.
method is often used in the chloriiiation process, but it is better

to use protochloride of tin, SnCl 2 This substance gives a brown
.
precipitate of variable composition in concentrated solutions,

but if mixed with the tetrachloride, SnCl 4 it gives a precipi-
,
tate of purple of Cassius. The reaction is very sensitive, and

by its means a violet coloration by transmitted light can be
obtained in a solution containing 1 part of gold in 500,000 parts
of water, while by special means the presence of 1 part of
gold in 100,000,000 parts of water can be detected, as de-
scribed below * : The liquid supposed to contain gold is raised
to boiling, and poured suddenly into a large beaker contain-
ing 5 to 10 c.c. of saturated solution of stannous chloride,
and the liquids agitated so as to effect complete mixture. A
yellowish-white precipitate of tin hydrate forms, which settles
rapidly, and can be readily separated from the bulk of the liquid
by decantation. If the solution originally contained at least
1 part of gold in 5,000,000 of water (3 grs. per ton), the
precipitate is coloured purplish-red or blackish-purple, accord-
ing to the nature of the solution, and the condition of the
precipitant. The colour can be seen without comparing it with
other precipitates. If less gold than this is present it is better
to compare the precipitate with one obtained by the use of
boiling distilled water, and to increase the quantity of liquid
used while adhering to the same amount of stannous chloride.
In this way the presence of 1 part of gold in 100,000,000
parts of water (1 grain of gold in 6 tons of water) can be
detected, the amount of liquid required in this case being about
3 litres. The gold is concentrated in the precipitate in which
a distinct colour is caused by less than OO5 per cent, of the
precious metal.
Chlor-auric Acid, H AuCl 4 Gold trichloride in the presence
.
of free hydrochloric acid is supposed to form this compound,

which crystallises out on evaporation in vacuo in long, yellow
needles, having the composition HAuCl 4 + 4H 2 O, and since
gold chloride unites with many other soluble chlorides to form
double chlorides, this hydrochloric acid compound is regarded as
an acid. It is more stable than gold trichloride. The chlor-
aurates, having the general formula M'AuCl 4 or AuCl 3 M'Cl,
,
.
are readily soluble bodies which can be crystallised, and which

decompose with about the same readiness as chlor-auric acid.
BROMIDES OF GOLD.
Gold Protobromide, AuBr, is a yellowish-green powder ob-

tained by heating the tribromide to about 140. It is insoluble
in water, but is decomposed by it, metallic gold and the tribro-
*
T. K. Rose, in Chem. News, 1892, vol. Ixvi., p. 271.
CYANIDES OF GOLD. ..
27
mide being formed ; the change is especially rapid on boiling,

and is hastened by the presence of hydrobrornic acid.
Auro-auric Bromide, Au 2 Br4 is produced by the action of
,
bromine on finely-divided gold in the cold, some tribromide

being simultaneously formed. Water breaks up this bromide
into AuBr and AuBr 3 and, according to some authorities, it is
,
only a mixture of these bodies.'

G-old Tribromide, AuBr3 is produced by the action of a
,
mixture of bromine and water 011 gold, particularly on the

application of heat. Auric tribromide resembles the trichloride
in most of its properties. It crystallises in blackish needles or
scarlet plates. It is deliquescent, and very soluble in water, and
suffers decompositions similar to those noted in describing AuCl 3,
itssolutions being still less stable than those of the chloride. A

solution of gold tribromide is gradually decolorised by sulphur
dioxide, being completely reduced to the state of monobromide

before any precipitate of metallic gold is formed. It is prepared
in a pure state by heating finely-divided gold in sealed tubes
with bromine and arsenic bromide, AsBr 3 to 126. Gold tri-
,
bromide forms intensely coloured brownish-red aqueous solutions,

the presence of a mere trace of the salt in a solution being
observable in this way. Double bromides exist analogous to
the chlor-aurates. The bromides are not volatile.
The Iodides of Gold are of little interest to the metallurgist.
They are prepared with difficulty, and decompose more readily
than the chlorides and bromides and are not likely to become of
practical importance. The tri-iodide is formed if gold is heated
with water and iodine to 50, particularly in direct sunshine.
CYANIDES OP GOLD.*
Cyanogen and gold unite in two proportions forming aurous

and auric cyanides, but the latter is only known with certainty
in combination.
Aurous Cyanide, AuCy, is obtained by heating aurocyanide
of potassium, KAuCy 2 with hydrochloric or nitric acid.
,
It is
a lemon-yellow crystalline powder, insoluble in water, and
unaltered by exposure to air. It is decomposed by heat, yield-
ing metallic gold and cyanogen, and is soluble in ammonia, in
alkaline cyanides, and in hyposulphite of soda. It is unattacked
by the mineral acids, except by aqua regia, but

is decomposed
when boiled with potash, metallic gold being thrown down.

Aurocyanide of Potassium, KAuCy 2 is obtained by crystal-
,
lisation from its solution, which

prepared by dissolving
is
metallic gold, auric oxide or aurous cyanide in a solution of

potassium cyanide. It is slightly soluble in water and the
*
See also chapter xvi.
2C THE METALLURGY OP GOLD.
aqueous solution, especially if hot, gilds copper or silver without

the agency of a battery, the gold being replaced in solution by
the other metal. Precipitates are also formed on the addition
of salts of zinc, tin, iron, or silver, no precipitates being formed
if potassium cyanide is present in excess.
Auricyanide of Potassium, AuCy 3 KCy, is formed by adding
.
potassium cyanide to a solution of trichloride of gold, the pre-

cipitate first formed being redissolved. The solution is com-
pletely decolorised, and on cooling deposits colourless crystals of
AuCy3 KCy + 3H 2 O.
. These effloresce in air, giving up two
molecules of water ; and, on heating, the third molecule of water
and some cyanogen are given off, aurocyanide of potassium being
formed, and this in its turn is decomposed at a slightly higher
temperature.
OXIDES OF GOLD.
Aurous Oxide, Au 2 O. This oxide is prepared by decomposing

aurous chloride, AuCl, or the corresponding bromide by potash
in the cold (Berzelius), when a violet precipitate forms which
is blackish when moist, but greyish when dry. When freshly
precipitated it is soluble both in alkalies and in cold water,
forming an indigo blue solution, with brownish fluorescence, and
on warming the solution slightly the corresponding hydrate is pre-
cipitated. It is also prepared by the action of nitrate of mercury
on the trichloride, and by boiling aurate of potash with organic
compounds, such as citrates or tartrates, or by boiling a solution
of the trichloride with the potassium salts of these acids. When
prepared according to these methods, aurous oxide always
contains a certain proportion of metallic gold. Kriiss obtained
the oxide pure by reducing brom-aurate of potassium at by
SO 2 passing in the gas only until the solution became colourless,
,
after which an excess of gas would have precipitated metallic

gold. Aurous hydrate is then precipitated by potash, and, after
being agglomerated by boiling, it is filtered, washed with cold
water, dried, and heated to 200 to expel the water of hydration.
At 250 it is resolved into gold and oxygen. Hydrochloric acid
decomposes aurous oxide into metallic gold and auric salts, slowly
in the cold, quickly at a boiling temperature; aqua regia dis-
solves the oxide, but sulphuric and nitric acids are without action
on it, while weak bases at once decompose it.
An intermediate oxide, AuO, is prepared as a black powder
by dissolving metallic gold in aqua regia containing an excess of
hydrochloric acid, then adding an excess of carbonate of potash,
and afterwards filtering and drying the precipitate. It has been
little studied, but the temperature at which it decomposes has
been fixed at 205 and its hydrate has been prepared.
Auric Oxide, Au 2 O 3 This, the best known oxide, is a black
.
OXIDES OF GOLD. 29
powder when anhydrous, and is precipitated from solutions of
auric chloride in the form of a hydrate by the caustic alkalies, the

carbonates of the alkalies, and hydrates of the alkaline earths or
zinc. The readiest method of preparation of this compound is
to add caustic potash, little by little, to a hot solution of gold
chloride, until the yellow precipitate first formed is dissolved to
a brown liquid. Then a slight excess of sulphuric acid or some
Glauber's salt is added, the precipitate filtered off, washed and
purified from potash by being redissolved in concentrated
nitric
acid, and reprecipitated by dilution with water. On drying this

precipitate in vacuo, the hydrate . H
Au 2 O 3 2 O, an ochreous
powder, results. If it is heated to 110, oxygen begins to be
given off. At 160, AuO remains, and on heating for some time
;it 250, metallic gold remains. Trioxide of gold dissolves in
concentrated sulphuric and nitric acids, from which it is partly
reprecipitated on boiling or on dilution, and these solutions are
supposed to contain sulphates and nitrates of gold respectively.
Double nitrates of gold and the alkalies have been obtained as
crystals. Hydrochloric and hydrobromic acids dissolve the
trioxide forming the haloid salts, but hydriodic acid decomposes
it on boiling, giving iodine and metallic gold. Gold trioxide
dissolves in boiling solutions of alkaline chlorides, giving aurates
and chlor-aurates, while it also combines with metallic oxides to
form aurates.
It is easily reduced by hydrogen, carbon and carbonic oxide,
with the aid of very gentle heat. Boiling alcohol reduces it,
yielding minute spangles of gold which were formerly used in
miniature painting.
Aurates. The aurates of potash and soda have the general
formula Au 2 O 3 R'2 O or K' 2 Au 2 O 4 assigned to them.
.
They
are readily soluble, crystallisable compounds, and are formed
when alkalies are added in excess to solutions of gold chloride.
The aurates of calcium, magnesium and zinc are insoluble in
water, but soluble in hydrochloric acid.
Fulminating Gold a compound of auric oxide with am-
is
monia, Au 2 O 3 (NH 3 ) 4,
is formed by precipitating gold
which
chloride with ammonia or its carbonate, or by the action of
ammonia on gold trioxide. When prepared by the former
method its composition is variable, but the fulminate is always
a fearful explosive, decomposing with violence at 145, or 011
It is de-
being struck, and sometimes even spontaneously.
composed without explosion by sulphuretted hydrogen, and
by protochloride of tin. It is a grey pulverulent powder,
insoluble in water, but soluble in potassium cyanide, auricyanide
of potassium being formed.
Sulphites of Gold. Alkaline sulphites, or sulphur dioxide,
which reduce gold trichloride easily, do not produce the same
effect on a solution of an alkaline aurate. If sodic bisulphite is
added to a boiling solution of sodic aurate (NaAuO 2 ) a yellowish

precipitate is formed, soluble in excess of sodic bisulphite, and
consisting of a double sulphite of gold and sodium, or sodic
auro-sulphite, having the composition 3Na 2 SO 3 2
SO 3 + 3H 2 O. . Au
It is obtained pure by precipitating the corresponding baric salt
with Bad,,, and decomposing the precipitate with the minimum
quantity of sodic carbonate. Double sulphites of potassium and
ammonium with gold also exist. These salts are decomposed by
acids, sulphite of gold being deposited, and also on boiling their
aqueous solutions, but the addition of sulphuretted hydrogen or
alkaline sulphides has no effect on them.
Hyposulphites of Gold. These compounds are now called
thiosulphates by chemists, but the old name is retained as being
universally employed by metallurgists. They are especially
interesting to the metallurgist, as on their formation depends the
extraction of gold from auriferous silver ores, when these are
treated by the ordinary hyposulphite, or by the Russell process.
The soluble double hyposulphites of gold with the alkalies and
alkaline earths have the general formula 3R"S 2 O 3 Au 2 S 2 O 3 + .
4H O. The double compounds of gold with sodium, potassium,

calcium, magnesium and barium, are all known. The sodic
salt is prepared by adding a dilute solution of gold trichloride
little by little to a concentrated solution of sodic hyposulphite,
when the following reaction occurs :
8Na 2 S 2 3 + 2AuCl 3 = Au 2 S 2 3 . 3Na 2 S 2 3 + 2Na2 S 4 O 6 + GNaCl.
The double hyposulphite may be separated by precipitation with

strong alcohol, with which it is also washed, or it may be
purified by repeated solution in water and precipitation with
alcohol. Thus prepared, it consists of colourless crystalline
needles, highly soluble in water, but almost insoluble in alcohol.
The solution, which possesses a sweetish taste, decomposes under
the influence of heat, the action being much more rapid when
nitric acid is present; metallic gold and sulphate of soda are
formed. Gold, however, is not reduced from its solution as double
hyposulphite by either stannous chloride, ferrous sulphate or
oxalic acid, although sulphuretted hydrogen and alkaline sulphides
give a black precipitate of S
2 3
Au
The addition of hydrochloric
.
acid or of dilute sulphuric acid does not immediately cause an

evolution of sulphur dioxide and a deposit of sulphur, as in the
case of ordinary hyposulphites. Since, therefore, the double
sulphite of soda and gold does not present the characteristics of
either aurous salts or of hyposulphurous acid, it has been sug-
gested that it contains a compound radical,
and has a composition
expressed by either Na S
3 4
O G
Au or Na,S 4
O 6Au + 2H2 O. The
addition of any dilute acid soon effects the decomposition of this
body in solution, gold sulphide being precipitated j the reaction
is accelerated by heat.
OXIDES OF GOLD. 31
The double hyposulphites of potassium, calcium, barium and

magnesium present similar characteristics. If the barium salt
is treated with the amount of sulphuric acid required
by theory,
a solution of the acid auro-hyposulphite, 3H 2 S 2 O 3 Au 2 S 2 O 3 is
.
,
obtained, but it cannot be crystallised. It has been supposed

that the calcium salt is more easily formed than the sodic salt,,
and, therefore, that calcium hyposulphite is more suitable than,
sodium hyposulphite for use in the leaching process, whenever
gold is present in perceptible quantities. According to a series
of experiments conducted by Russell,* this is not the case, very
little difference existing in the case of formation of the calcic,
sodic, magnesic and potassic salts.

Russell has demonstrated f that finely divided gold is soluble-
to a limited extent (i.e., OO02 gramme in 1,000 c.c. in 48 hours),,
in solutions of sodic hyposulphite of all degrees of concentra-
tion. The action depends on the oxidation of the gold by the
air present in the solution, the soluble double hyposulphite,
Au 2 S 2 O 3 3Na2 S 2 O 3 + 4H 2 O, and caustic soda being formed.
.
The formation of this hyposulphite by the action of the sodic-

salton gold sulphide is far more complete and rapid. In 24
hours in the cold, 0*066 gramme of gold, and in 2 hours at
65, 0*117 gramme of gold were dissolved in dilute solutions.
Since an alkaline sulphide will again precipitate Au 2 S 3 from
the solution, these facts seem to be at variance, and Stetefeldt
suggests that the gold sulphide originally contained a small
quantity of metallic gold in an excessively fine state of division,
and that on heating more gold was set free. He instances Level's-
statement that sulphuretted hydrogen precipitates metallic gold,,
and not gold sulphide, from a boiling solution of the trichloride.
He believes J that it is only the free gold which is attacked, and
that the results obtained on attacking the sulphide were better
than those in which metallic gold was used, owing to the much
finer state of division in which the gold existed in the former
case. Another reason for the inverse reactions may of course
exist in the influence of mass, the small quantity of sodic
sulphide which is formed by the reaction being insufficient to
precipitate any gold, which comes down on further additions of
the same reagent.
In some further experiments which Russell conducted, it was
proved that gold sulphide is far more soluble in a solution
containing the molecular quantities, 4Na.,S 2 O 3 3Cu 2 S 2 O 3 (hypo-
.
sulphite of copper and sodium), than in sodic hyposulphite alone,

especially if the solutions are cold. If the solution is heated
the reduction of the gold sulphide to metallic gold renders it less
soluble, and less rapid dissolution occurs. The composition of
"
the double salts obtained by the use of this " extra solution
has not been worked out.
*
Stetefeldt's Lixiviation of Silver Ores, New York, 1888, p. 90.
t/6/rf., p. 19. ilbid., p. 22.
SILICATES OF GOLD.
The existence of auro-silicates is now admitted without dis-
pute, and gold has for centuries been used to impart colour to
glasses, the method used being as follows A solution of chloride
:
of gold is added to a mixture of sand with alkalies and alkaline

earths or lead, and the whole is then fused, and colourless or
yellow transparent silicates of gold thus formed. These are
decomposed by being reheated gently to low redness, oxides of
gold, or more probably metallic gold, being set free, and red or
purple colorations thus obtained. The occurrence of silicates of
gold in nature seems to be doubtful.
Experiments conducted by E. Cumenge* tend to show that
the alkaline auro-silicates, obtained in the wet way, may have
played an important part in the formation of auriferous quartz.
The following conclusions have been established by these
investigations:
1. If an alkaline aurate, obtained

by dissolving auric ses-
quioxide in caustic soda, is mixed with an alkaline solution of
silicate of soda (soluble glass), the mixture may be concentrated
by evaporation until it has attained a syrupy consistency with-

out being decomposed. Auro-silicate of soda is, therefore, fairly
stable, so long as there is an excess of alkali present.
2. The decomposition of this auro-silicate is effected by the
addition of hydrochloric acid to it, by which gelatinous silica is
precipitated. This carries down a certain proportion of gold
which gives a rose colour to the white magma.
3. This decomposition may also be completely effected by the
action of an aqueous solution of carbonic acid under pressure.
Thus, if the syrupy, alkaline auro-silicate is introduced into a
bottle of seltzer water, which is then hermetically closed, the
decomposition can be seen to be gradually going on without
the semi-fluid mass being dissolved, and the latter is replaced
at the end of some days by coherent silica, which, on exposure
to the air, assumes a white opaline appearance tinged with rose
colour.
4. When gelatinous silica, obtained by the decomposition of
an alkaline auro-silicate, is heated to redness in a current of
steam, it assumes either a beautiful, unalterable rose colour, or a
reddish tint with visible grains of gold, according to the pro-
portion of precious metal present, and the conditions under
which the precipitation has been effected.
SULPHIDES OF GOLD.
These compounds are prepared as brown or black precipitates
by passing sulphuretted hydrogen through a solution of gold
*
Frgmy, Ency. Chim., vol. iii., L'or, p. 62.
SULPHIDES OF GOLD. 33
chloride. The exact composition of the precipitate varies with

the temperature and degree of concentration of the solution, and
the amount of free acid present. Levol and Kriiss state that
Au.7 S is precipitated in the cold, but that only metallic gold and
free sulphur are thrown down from boiling solutions. According
to others, Au 2 S 3 is precipitated from cold and Au 2 S from boiling
solutions. It seems probable, however, that free sulphur is always
formed in considerable quantities, and whether the solutions are
hot or cold, dilute or concentrated, definite compounds are not
precipitated, variable mixtures of the two sulphides with free
sulphur and metallic gold being formed. Similar precipitates
are formed by alkaline sulphides and by sulphides of most of the
heavy metals. The sulphides are soluble to some extent in a
saturated solution of sulphuretted hydrogen, and are easily
soluble in hot solutions of alkaline sulphides or alkalies, forming
double salts so that precipitation from alkaline solutions is
never complete. The sulphides are readily decomposed into
gold and sulphur by the action of heat, the decomposition
being complete below 270. Sulphide of gold is also dissolved at
ordinary temperatures by potassic cyanide, and is slowly attacked
by mercury with formation of mercury sulphide.
Purple of Cassius. This body was discovered by Cassius
of Ley den in 1683. It contains gold and oxide of tin, and is
used to colour glass and glazes, various shades of violet, red
and purple being thus obtainable. Several methods of prepara-
tion are used, of which the following is that employed at the
factory at Sevres""" : Half a gramme of gold is dissolved in aqua
regia composed of 16*8 grammes of hydrochloric and 10*2 grammes
of nitric acid, and the solution is then diluted with 14 litres of
water. To this solution is added, drop by drop, a solution of a
mixture of protochloride and tetrachloride of tin, prepared as
follows : 3 grammes of finely divided tin is dropped, little by
little, into 18 grammes of aqua regia (constituted as above, with
the addition of 5 c.c. water), the reaction is checked by cooling
if it is too violent, and the solution of chloride of tin formed
is allowed to cool. The precipitate of purple oxide thus obtained
is finely coloured when it has been washed with boiling water.
The purple precipitate obtained by Miiller, by reducing chloride
of gold with glucose in an alkaline solution containing tin oxide
in suspension, differs from that prepared by the foregoing method
in losing its colour at a red heat, while the true purple of Cassius
becomes brick-red under such conditions. The true colour is
seen when metallic tin acts on trichloride of gold, or when alloys
of gold and tin are attacked with nitric acid.
The composition of purple of Cassius has been the subject of
much discussion. Some chemists have considered it to be a
compound of aurous oxide with the oxides of tin. Debray
* vol. v., L'or, p. 63.
Fr^my's Ency. Chim.,
3
regards it as a lake of stannic acid coloured by finely divided

gold. If this latter view is correct then the gold may be present
in an allotropic modification, since
(1) the purple of Cassius is
completely dissolved by ammonia, a purple solution being formed,
and the solution may be kept unaltered for weeks although it
is decomposed by light, or on heating, becoming bluish, and
finally depositing metallic gold; the ammonia may be removed
by dialysis, leaving a purple aqueous solution which contains
both gold and stannic oxide;* (2) the purple of Cassius does
not yield any gold on treatment with mercury. Miiller confirms
Debray's views, showing that fine purple compounds can be made
with gold and magnesia, lime, baryta, sulphate of barium, &c.,
the colour depending on the presence of finely divided gold and
not on the other constituent. The purple colour possessed by
(possibly allotropic) gold, when in a finely divided condition,
is further attested
by the purple stain given to the fingers by
a solution of gold trichloride, and by the colour of Roberts-
Austen's aluminium alloy, AuAl 2 .
CHAPTER III.
MODE OF OCCURRENCE AND DISTRIBUTION

OF GOLD.
Forms in which Gold occurs in Nature. Gold is obtained
from two very different sources, viz., (1) from veins in rock for-
mations, and (2) from placers, or the alluvial deposits of ancient
and modern streams.
Vein Gold. In this case the metal, whenever it is present
in visible grains or masses, has sharp angular edges, and although
usually not distinctly crystalline, it frequently penetrates the
rock irregularly in various directions, and is completely inter-
woven with, and attached to the matrix, usually quartz, so that
the metal cannot be separated from the rock without crushing
the latter.
The gold in lodes is sometimes in the form of crystallisations,
which are, however, exceedingly rare, and crystals of gold are still
probably unknown to most miners, although they occur more fre-
quently in placer deposits. Arborescent branching and dendritic
masses of crystalline gold are more common than single crystals
in both quartz lodes and placer deposits. The crystalline forms
met with have already been described, p. 8. In the Transyl-
vanian lodes, gold occurs chiefly in thin sheets or plates, often as
*
Chem. Soc. Journ., vol. Ixiv. (1893) p. 575.
OCCURRENCE AND DISTRIBUTION OF GOLD. 35
much as from half an inch to two or more inches in breadth.

Such plates are rarely thicker than- a visiting card, and are
generally covered with crystalline lines and markings, revealing
a distinct geometrical structure. Gold also occurs in wire-like
forms, sometimes penetrating crystals of other minerals, such
as calcite and dolomite.
The matrix in which the gold is contained is usually quartz,
intersecting as veins or interlaminated with sub-crystalline, slaty,
or schistose rocks, especially hydromica and chloritic slates.
Gold also occurs sparingly in similar veins in granite and gneiss,
and has been detected in the trachytes of Colorado, and in
silurian and carboniferous trachytes, as well as in some lime-
stones.
With regard to the distribution of gold, W, P. Blake ob-
serves* " There is a much greater dissemination of gold in a
ragged granular condition, in situ, in fine particles in the midst

of rock formations, and without any obvious connection with
veins, than is generally supposed. Prominent examples are
found in the belts or zones of layers of soft slate in Georgia, and
in North Carolina. ... The Boly- Fields gold vein, Lumpkin
County, Georgia, is an example of the occurrence of coarse ragged
gold in the midst of a mass of slate, without any defined quartz
vein. The gold is closely associated with bornite, pyrites, and
dolomite." The dissemination of gold in the schistose rocks of
North Carolina has also been noted by Professor Kerr,f and by
Dr. Emmons, and similar occurrences have been observed in Texas,
Nova Scotia, and in other parts of the world. At the Contention
Mine, Tombstone, Arizona, free metallic gold is found in the thin
cracks and cleavage surfaces of partially decayed porphyry, and
appears to have been deposited there from solution, and not
mechanically. It occurs in thin subcrystalline flakes and scales,
and may have been derived from the decomposition of the iron
pyrites with which the adjoining sedimentary formations are
charged. Gold also occurs in small quantities (1 part in
1,124,000) in the bed of clay on which the city of Philadelphia
is built. Its occurrence in solution in sea water has been proved
by Sonstadt.J
The wide distribution of gold in minute quantities throughout
the world was pointed out by W. E. Dubois, an Assayer in the
United States Mint, in 1861, and is further attested by a large
number of specimens now in the Percy collection. These con-
sistof small specks of gold of different sizes which have been
obtained from the most varied sources. Thus, samples of Pattin-
son's crystallised and uncrystallised lead, pig lead from all
*
Prod. Gold and Silver in United States, 1884, p. 581.
t Trans. Am. Inst. Mng. Eng., vol. x., p. 475.
J Chem. News, vol. xxvi., p. 159.
Journ. Am. Phil. Soc., June 1861.
countries, lead fume, red lead, litharge, white lead, precipitated

carbonate of lead and acetate of lead were all found to contain
gold, which seems to be invariably present in galena. Moveover,
it appears to be impossible to procure samples of copper in
which gold cannot be detected, although the Lake Superior
copper contains less than 1 part in 1,000,000; the bronze and
copper coins of all nations are usually found to contain much
greater quantities of gold than this. Similar evidence has been
adduced which tends to show that all ores of silver, antimony,
and bismuth contain gold. *
Placer Gold and Nuggets. Placer gold is usually in the
form of small scales, but pellets or rounded grains also occur,
while larger masses or nuggets are usually of a rounded mamrnil-
lated form. The chief difference between the appearance of
placer and vein gold lies in the fact that the former is always
rounded, showing no sharp edges, even the crystals having their
angles smoothed and rounded off. This has been pointed to by
the advocates of the erosion theory of the origin of placer gold, as
evidence in favour of their views, the roundness of the fragments
being taken to prove that abrasion of the gold has been effected
by attrition with water and grains of sand. The largest masses
of gold yet discovered have been found in auriferous gravel.
The "Blanch Barkley" nugget, found in South Australia,
weighed 146 pounds, and only six ounces of it were gangue;
and one still larger, the " Welcome " nugget, from Victoria,
weighed 2,195 ounces, or 183 pounds, and yielded gold to the
value of 8,376 10s. 6d. In Eussia a mass was found in 1842
near Miask, weighing 96 pounds troy. The largest mass from
California is given in the State Mineralogist's report as weigh-
ing 2,340 ounces, or 195 pounds, but no authentic cases seem
to be on record of nuggets from this State weighing more than
20 pounds.
The minerals most common in auriferous quartz lodes or in
the placer deposits are platinum, iridosmine, magnetite, iron
pyrites, galena, ilmenite, copper pyrites, blende, tetradymite,
zircon, garnets, rutile and barytes wolfram, scheelite, brookite
;
and diamonds are less common. Diamonds are associated with

gold in Brazil, and also occasionally in the Urals and in the
United States. The sulphides present in auriferous quartz
frequently contain gold the gold in such an ore is usually in
;
part quite free, disseminated through the quartz, in which visible

grains of the metal often occur, and in part locked up in the
pyrites, whence but little can in general be extracted by mercury.
It is, however, in all probability in the metallic state in pyrites,
although this is not completely established. The subject is
discussed in Section on " Gold in Pyrites," Chap. vii.
*
Loc. cit., and E. A. Smith on Bismuth, &c., Journ. Soc. Chem. Ind. t
vol. xii. (1893), No. 1.
Among minerals other than sulphides which contain gold,

the following may be mentioned, although none are of great
importance in metallurgy : Calaverite is a bronze - yellow
gold telluride, usually containing a little silver, occurring in
certain mines in California and Colorado. One analysis gave
tellurium 55 -5 per cent., gold 44'5 per cent., corresponding to
the formula AuTe.2 .
Sylvanite, called also graphic tellurium,
is a telluride of gold and silver, supposed to correspond to
(AuAg)Te 3 It sometimes contains antimony and lead in ad-

.
dition. It is from white to brass-yellow in colour, and the

arrangement of the crystals sometimes bears a resemblance to
writing characters, whence the name graphic. It occurs in
Transylvania, in Calaveras County, California, and in Colorado.
Petzite is a telluride of silver, Ag 2 Te, in which the silver is
partly replaced by gold. Aspecimen from the Golden Rule
Mine, according to Genth, contained tellurium 32*68 per cent.,
silver 41-86 per cent., and gold 25-60 per cent. It occurs in
Transylvania, Chili, California, Colorado, and Utah.
Nagyagite or Foliated Tellurium is remarkable for being
foliated like graphite, which it also resembles in its colour, a
blackish lead-grey, and in having a hardness of from 1 to 1-5 only
Its density, however, is above 7. It occurs in Transylvania, and
contains tellurium 32*2, lead 54'0, gold 9-0 to 13-0 per cent., some-
times with silver, copper and sulphur in addition. Other gold
tellurides and some native gold amalgams are occasionally met
with, but these minerals in which gold is an essential part are
rarely of much importance as ores, as they seldom occur
in
sufficient abundance to be regarded as anything but specimens
for collectors. In some few mines, however, notably at the
Cripple Creek district, Colorado, in Transylvania, in Boulder
County, Colorado, and at the Bassick mine in the same State,
the value of the ore depends on the tellurides of gold con-
tained in it.
One of the most striking differences between the ores of gold
and those of all other metals lies in the extremely small propor-
tion which the desired material bears to the worthless gangue
with which it is accompanied. Occasionally hand specimens
of vein stuff are found containing several per cent, of the
occurrence
precious metal, but these are of quite exceptional
and have not the slightest economic importance. The greater
part of the vein gold now being produced is derived
from ores
or eighty
containing only about one part of gold in seventy
thousand, whilst, under exceptional circumstances, a yield of
one
part in half a million parts of gangue may give handsome profits.
Placer deposits are usually much less rich than this; the average
amount of gold contained in, those now worked does not exceed
one part in one million, and in California deposits of gravel with
only one part of gold in fifteen millions have proved
"
susceptible
of successful treatment by the " hydraulicking method on a
large scale.
Composition of Native Gold. Native gold usually contains

silver, which occurs in varying proportions, the colour becoming
paler with the increase of silver. The finest native gold yet
found is that from the Mount Morgan Mine, Queensland, which
is 997 fine. The finest Russian gold was that formerly obtained
at Katerineburg in the Urals, and yielded gold 989'6, silver
1*6, copper 3-5, and iron 0*5 (G. Rose). Gold dust from West
Africa has been found to contain 978 '1 of fine gold, the remainder
being silver. The gold found in the British Isles varies from
800 to 900 fine, the remainder being silver the specimens from
;
the district around Dolgelly are sometimes a little over 900 fine.
Particulars of the fineness of gold from Australia, California,
&c., are given in the chapter on Refining, Chap, xviii. Gold is
occasionally found alloyed with copper, and sometimes also with
iron, bismuth, palladium, or rhodium. Rhodic-gold from Mexico
was found to be of the specific gravity 15*5 to 16*8 and contained
34 to 43 per cent, of rhodium. Bismuthic-gold has been called
Maldonite.*
Geographical Distribution of Gold. Gold occurs in lodes
in many districts composed of partially metamorphosed rocks
such as slates or schists, while its occurrence in holocrystalline-
metamorphic, or igneous rocks is comparativaly rare. Among
sedimentary rocks, its occurrence is almost confined to the sands
of rivers which run for a part of their course through crystalline
formations, or more particularly through districts in which gold
occurs in quartz veins. Such river sands are rarely quite free
from gold. The beds of ancient rivers no longer existent are
also frequently auriferous. In spite of the fact that the sea
contains gold in solution, the aggregate amount perhaps
exceeding that contained in the accessible portion of the earth's
crust, nevertheless, unaltered marine deposits seldom or never
contain a perceptible quantity of the metal, except in one or two
cases of beach deposits formed by the erosion of auriferous land-
formed gravels.
In the British Isles, gold is found in some of the streams of
Cornwall and in lodes and river gravels near Dolgelly and in
other parts of Wales, in Sutherlandshire and near Leadhills in
Scotland, and in the County of Wicklow. The total amount which
has been obtained from these localities is small, probably not
exceeding 40,000 ounces, and little is now being produced. On
the Continent of Europe, gold is most abundant in Hungary and
Transylvania, where the gold occurs in quartz lodes contained in
eruptive rocks of tertiary age, chiefly propylite, porphyry, diorite
and granite. The minerals occurring with the gold are galena,
blende and pyrites. In the German Empire, the gold obtained
is chiefly derived from the smelting of argentiferous galena in
which small quantities of the more precious metal are contained.
*
Dana's Mineralogy, p. 108.
In Italy the only important mines are those of Pestarena and

Val Toppa in North Piedmont, near Monte Rosa. The pyritic
ores from these mines are treated by amalgamation. Gold is
also found in the sands of the Rhine, the Reuss, the Aar, and
other rivers, and in small quantities in Sweden.
The gold-bearing districts of R-ussia are (1 the Urals, (2) Siberia,
)
Eastern and Western, whilst an insignificant amount is' also

derived from Finland and from the Caucasus. The gold was for-
merly derived chiefly from lodes both in the Urals and in Western
Siberia. In the Urals, quartz-mining began in 1745, and the
output from this source continually increased up to the year
1810, when it began to fall off, and has been trifling since 1838.
The placers, which are mainly situated on the eastern slope of
the range, were discovered in 1774, and the yield has continually
increased since then up to the present time. The Siberian placers
were discovered in 1829, although quartz-mining had been prose-
cuted since 1704.* The output from these continually increases,
in spite of the exhaustion of the old placers and the reduction in
the percentage of gold contained in those still worked ;
this
increase of output in consequence of the discovery of new
is
placers further east. The auriferous gravels are all thin, shallow
deposits, ranging from 3 to 20 feet in thickness, and as they are
worked out, other gravels are opened-up further east, so that
operations are being gradually transferred from west to east.
When the exhaustion of these placers has proceeded further it
may be expected that more attention will again be paid to the
quartz lodes. The relative amounts produced by the different
districts are given as follows :
f
chiefly by various primitive methods little is known of the

;
methods used and of the amount produced in the former country.

From the United States a large percentage of the total gold
production of the world is obtained. The chief producing States
are California, Colorado, Dakota, Montana, Nevada, Idaho and
Oregon, but smaller amounts come from many other States. The
produce is now far more from lodes than from placer deposits,
and in the treatment of auriferous quartz and pyritic ores almost
all the known methods of treatment are
applied in different
localities.* Mexico produces considerable quantities of gold,
and gold ores are also found in various parts of Canada, Colombia,
Bolivia, Chili, Venezuela, Brazil, Peru, and the small States of
Central America. The production of several of these countries
was formerly much larger than it is at the present day, the
reduction being especially marked in the cases of Brazil and
Venezuela. On the other hand, the gold districts of British
and French Guiana, of the Argentine Republic and of Uruquay
are now being developed, and may eventually prove to be of
considerable importance.
Gold is somewhat widely distributed in Africa, the chief
sources of production in former times being the placer deposits
of the Gold Coast and Abyssinia. The discoveries of auriferous
quartz deposits in the Transvaal since 1884 have converted that
region into one of the most important among gold producing
countries. The richest deposits are in the Banket formation on
the Witwatersrand, whence about nine-tenths of the gold
obtained in the Transvaal is derived, the De Kaap field being
next in importance. No placer deposits of value have been
discovered, and the gold is mainly obtained by stamp-battery
amalgamation, followed by treatment of the tailings by means
of the cyanide process. The produce from both sources is being
continually increased. The production by means of treatment
with cyanide is now about 30 per cent, of the whole.
Gold is found in all the Colonies of Australia, together with
Tasmania and New Zealand; the largest amount is now pro-
duced by Victoria, and Queensland, New South Wales, and
New Zealand follow in the order named. The chief gold pro-
ducing districts of Queensland are Charters-Towers, Rockhampton
(where the Mount Morgan mine is situated), Croydon, and
Gyinpie. Almost all the gold is produced from the quartz
mines, the placers having been practically exhausted thus in ;
1891, out of a total production of 576,439 ozs., 560,418 ozs.

were produced from quartz, and only 16,021 ozs. from alluvial
deposits, while in 1877 the amounts were Placer gold, 164,778
ozs.; quartz gold, 188,488 ozs. total, 353,266 ozs.f
*
For full details as to the geographical and geological distribution and
the nature of the gold ores found in the United States, the student is
referred to the Annual Reports of the Director of the United States Mint,
and of the Californian State Mineralogist.
t Mineral Jndu*ti-y, 1892, p. 192.
OCCURRENCE AND DISTRIBUTION OF GOLD. 4}
The production of gold in Victoria is now increasing, but is far

less than formerly, owing to the exhaustion of the alluvial
deposits,
although the yield of the quartz mines has continuously increased.
In 1894 the production was divided as follows Alluvial
:
gold, 254,308 ozs.; reef gold, 419,371 ozs. total, 673,680 ozs.
The chief producing districts are Ballarat, Sandhurst (Bendigo),
Beech worth, Maryborough, Castlemaine, Gippsland and Ararat.
In 1894 in Victoria the average yield per ton of quartz crushed
was 8 dwts. 8 grains of gold, this being slightly less than the
mean yield during the last decade.
In New South Wales about half the gold is obtained from
quartz lodes the pyritic ores are not yet effectively treated, and
;
the river bank deposits have not up to the present been exploited
on a large scale.* In New Zealand, only quartz lodes are worked
in the North Island, alluvial workings being confined to the
Middle Island. The most important districts are those of
Kuaotuna, Thames, Coromandel, Waihi and Reefton in the
North Island, and Ross and Kuraara in the Middle Island. f
Origin of G-old Ores. The origin of mineral veins, includ-
ing those in which gold is contained, has long been discussed
by geologists. The old theory that the quartz of veins was
originally in a molten condition and was ejected from below into
fissures is no longer maintained, although in 1860 H. Rosales
brought forward evidence in its favour as far as the Victorian
lodes are concerned. It is now admitted by all that the
materials forming the veins have been transported in aqueous
solution and precipitated where they occur. In a few excep-
tional cases, sublimation may have played a part. The view
that the solutions found their way downwards from above has
been abandoned, but the ascensional theory and the lateral
secretion theory both have many adherents. The last-named
theory has found its principal supporter during the last twenty
years in Prof. F. von Sandberger, who pointed out that the:
gangue of many lodes varies in composition if the nature of
the rocks through which they pass is changed, and claimed to
have proved by analysis that the materials forming vein-stone
are derived from the adjacent country rocks. He stated, more-
over, that such minerals as augite, hornblende, mica, and
olivine, which are essential constituents of crystalline rocks,
contain small quantities of the heavy metals occurring in veins. |
Although Sandberger did not try to detect gold in the silicates,
this metal is not likely to be an exception. Prof. A. Stelzner
objected to these conclusions, urging that small quantities of the
sulphides of the heavy metals were probably mechanically mixed
*
Report of the Dept. of Mines and Agriculture, N.S. W., 1894.
t Annual'Report of the Mining Commissioner of New Zealand^ 1891.
J Untersvchungen uber Erzgange. Wiesbaden, 1882 and 1885. Useful
abstracts are given in Phillips' Ore Deposits and in Le Neve Foster's Ore
and stone Mining.
with the crystals of minerals which Sandberger analysed in the

belief that they were pure. Stelzner advocated the retention of
the ascensional theory, which alone affords a satisfactory expla-
nation of the difference in composition observable in neighbour-
ing lodes passing through the same rocks, and apparently formed
at different periods. The two theories are, however, not con-
tradictory, and perhaps neither need be entirely rejected, the
solutions being supposed to pass more or less freely in the plane
of the lode, after they have been impregnated. For the origin
of placer gold, see p. 67.
CHAPTER IV.
TREATMENT OF SHALLOW PLACER DEPOSITS.

" "
THE deposits grouped together under the name of placers
comprise sands, gravels, or any loosely coherent or non-co-
herent alluvial beds containing gold. They have accumulated,
owing to the action of running water, in the beds of rivers, or on
the adjojiiing inundation plains, or on sea beaches. They fall
naturally into two groups, between which no strict line of
demarcation exists. These are
(1) Shallow or modern placers, which are in or near existing
rivers, and have not yet been covered by other deposits.
(2) Deep level or ancient placers, which now lie buried beneath
an accumulation of debris or coherent rock, the rivers by which
they were formed having often been deflected into other channels
by more or less extensive changes in the physical geography of
the district in which they existed. Beach deposits occur in each
subdivision.
] n this chapter, the first of these groups will be considered.
From the earliest times to the present day, shallow placer
deposits have probably yielded more gold in the aggregate than
has been derived from all other sources put together.
Shallow placers consist of loose aggregations of sand, gravel,
loam and clay, accumulated by the action of existing rivers and
streams, and not extending to a greater depth than 10 or 15 feet
from the surface. They contain metallic gold in fragments of
all sizes, ranging from the finest dust to nuggets weighing
thousands of ounces. Auriferous sands are found in the beds of
most rivers which flow during any part of their course through a
region composed of crystalline rocks. If the rivers have rocky
beds, gold may be found in the crevices, caught in natural riffles,
and the whole may subsequently be covered by beds of sand.
SHALLOW PLACER DEPOSITS. 43
Gold also occurs in river bars and banks, in river "

flats," or
inundation plains, in the dry beds of streams which only flow
after heavy rains (" gulch diggings "), in terrace gravels on the
sides of valleys high above the present level of the water ("bench
diggings "), and 011 the sides and tops of hills (" hill diggings ").
The last two subdivisions are evidently ancient rather than
modern deposits. The gravels may contain boulders of any
size, up to several feet in diameter, or may shade off into fine
sand, while sandy clays, especially if on the bed-rock, are
frequently very rich. In the Urals, the placer deposits often
consist of heavy clays, while others are formed of waterworn
fragments of auriferous quartz, talcose and chloritic schists,
serpentine, greenstone, &c. Gold occurs under very various
conditions in these deposits. It may occur in the grass roots on
inundation plains, or near the surface of the gravels in river
beds, or dispersed through the whole thickness of a stratum.
More commonly, however, the lowest part of the superficial beds,
" bed-rock "
just above (the country rock of the district), is
richest. In hollows, cracks, and crevices of the bed-rock, or,
if it is soft and decomposed, in the substance of the upper part
of the rock itself, to the depth of 1 or 2 feet, gold occurs in
the greatest quantities, vln pipeclays just above bed-rock, in
Victoria, it was not uncommon to find 12 ozs. of gold or more in
a single tubful of " dirt," and similar rich bed-rock deposits
occur in California. The depth at which bed-rock is found
varies greatly ; it may crop out at the surface, or it may be
buried beneath hundreds of feet of gravel, and 'great variations
occur even in a single district. In the Urals, however, the
thickness of the gravel is usually less than 3 feet thick, and is
rarely more than 12 feet.
Methods of obtaining Gravel from Shallow Placers.
The gravel obtained from any placer deposit is, with exceptions,
ultimately all treated alike, but the mode of winning the dirt
varies with the necessities of the case. On flats and bars, the
surface gravel, if rich enough, is loosened with pick and shovel,
and then washed. If only the part just above the bed-rock will
pay for treatment, it is reached by "stripping," or, if covered Iby
too great a thickness of barren material, shafts are sunk, and
short levels run from the bottoms of them in all directions. If
the nature of the ground permits of it, tunnels are run without
shafts, and the rich gravel is then followed from the surface,
wherever it is found. This system was much practised in the
early days in California, although now seldom to
be seen in
" which lives
operation it was called coyoting," from the coyote,
;
in holes in the ground. When water was encountered in the

shaft, it was drawn out by a bucket, until it came in too fast,
when the claim was abandoned. In California, in somewhat
later times, efforts were made to reach the gold in the river beds
by deflecting the streams of water from their courses, and in

other ways. These methods will be briefly described under the
head " River Mining."
Methods of Washing the Gravel. The Pan. When the
existence of gold in the placers of California and Australia first
became known, the diggers were not acquainted with any appar-
atus which was well adapted to extract the metal. The house-
hold pan was used everywhere to wash the gravel, and though,
in its original form, it was a difficult implement to use efficiently,
it has retained its place in both countries for prospecting and also
for washing small quantities of rich material, however they may
have been obtained. The pan is usually made of stiff sheet-iron,
is flat-bottomed and circular, and at bottom about 14 inches
in diameter. The sides slope outwards at an angle of about 30
to the bottom, and are about 5 inches wide. A riffle is a useful
addition, formed by the thickening or bulging inwards of the

side, situated about half-way up the latter and running about
half-way round the pan. The method of using this pan embraces
several operations. First, it is filled to about two-thirds of its
capacity with pay-dirt, of which it then contains from one-
fifth to one-quarter of a cubic foot. It is then placed at the
bottom of a water-hole or convenient stream, and the dirt is
thoroughly broken up with both hands, care being taken not to
leave any lumps of clay. As soon as the contents of the pan are
reduced to the consistency of soft mud, the pan is grasped with
both hands a little behind its greater diameter, inclined away
from the operator, raised until the dirt is only just covered with
water, and shaken from side to side, while a slight oscillatory
circular motion is also imparted to it. The mud and fine sand is
soon obtained in suspension in the water, and gradually passes
over the far edge, which is lowered more and more, until little
but the stones, coarse particles of sand, black sand and gold is
left. The larger stones lie on the top and are removed by hand.
The final stage consists in lifting the pan with a little water in
it, and by a movement of the wrist, something like that used in
vanning, causing the material to be spread out by the water, in

a comet shape, in the angle of the pan. The "colours" i.e.,
yellow specks of gold are seen at the extreme head of the comet,
and also occur in the succeeding inch or two, mixed with the
black sand, while the quartz-sand forms the remainder of the
tail and is scraped or washed off. The gold is separated from
the black sand by (a] amalgamation with mercury, or (6) drying
and blowing away the black sand, a wasteful process. Liquid
amalgam is readily separated from sand, and the mercury is
then driven off by heat.
The Batea differs from the miner's pan in not having a flat
bottom. It is of wood turned in a lathe, about 20 inches in
diameter, conical, or more rarely basin shaped, and about 3
inches deep in the centre, so that the angle at the apex is about
160. The gold collects at the lowest point and clings to the
wooden surface under conditions when it would slide over iron.
The batea consequently is more rapid and effective in obtaining
a " prospect" than the pan, especial 'y when the gold is fine, but is
less frequently used in the United States and Australia. It had
its origin in South A merica. It is usually now made of enam-
elled iron with a hole in the centre fitted with a cork, when it
is for use in countries other than South America. It is con-
sidered by MM. Oumenge and Fuchs to be especially favoured
by the negro race.
Prospecting Trough. This instrument is used in the far east,
especially by the Chinese, Malays, Annamites, <fec. It is made
of wood, and is shaped in the form of a very flat reversed roof-
top, the angle between the long sides being about 150. In
place of a circular movement of the water an alternating rocking
motion is used, the water flowing up and down. The instrument
is easily handled, but is very slow.
Horn Spoons cut out of black ox-horns have been used by
prospectors, especially to finish the work begun by the pan.
The surface holds the gold well and shows " colour " very
readily.
Cradle or Rocker was introduced in California soon after
Tlie
the rush to the diggings took place in 1849. It consists of
first
a rectangular wooden box, about 3 feet long and 18 inches wide,
resting on two rockers (D, Fig. 4) similar to those used for
infants' cradles. The
shape of the walls is
shown in Fig. 4, which
is a section of the ap-
paratus. The method of
using it is as follows:
The gravel is shovel-

led into the riddle-box,
A, the bottom of which
consists of J inch mesh Fig. 4.
=
18 inches,
screen; theworkman sits
^ca le 1 in
-
by the side of the machine and rocks it with one hand, while he
pours on water by means of a dipper filled from a water-hole with
the other. The dirt is disintegrated and carried through the riddle,
and falls on the apron, B, which consists of blanketing. Here
the fine gold is caught, and the dirt then passes out from back
to front over the bottom, which is slightly inclined towards the
front, and the coarse gold, black sand, &c., is caught in two or
three rifiles, C, each of about 1 inch in height, to which mercury
is sometimes added to assist in retaining the gold. The rocking
motion not only assists in the disintegration of the dirt, which
is effected the water, aided by the stones, but also prevents
by
the sand from packing behind the riffles in the event of this
;
happening, gold would pass over the surface of the sand and be
lost. Consequently the rocking should be quite continuous,
since, after every pause, the sand in the riffles must be stirred
up before recommencing. It is, therefore, desirable for two men
to work together at the cradle, one to carry the gravel and
charge it into the hopper, and to remove the large stones from
the latter by hand, while the other man rocks the cradle and
pours on water. It requires three or four parts of water to wash
one part of gravel, and it is, therefore, better to carry the ore to
water than to carry water to the ore. When a clean-up of the
cradle is desirable, the riddle is removed, the apron is taken out
and washed in a bucket, and the accumulations behind the riffles
are scraped out and panned. Most of the fine gold in the dirt
is lost by the cradle, and two men working together can only
wash from 3 to 5 cubic yards per day, according to the nature of
the dirt.
The Long Tom replaced the cradle in California after a short
time, and was used there for some years, while it is still in
operation in parts of Australia and Dutch Guiana. It consists
of a sluice-box or trough (A, Fig. 5) about 12 feet long, 20 inches
Fig. 5.
wide at the upper end, and 30 inches at the lower end, and 9 inches
deep, with an inclination of about 1 inch to the foot. The lower
end of the trough is cut off at an angle of about 45 and closed
by a screen of punched sheet iron, B, which prevents large stones
from passing through it. Below the screen is the upper end of
the riffle-box, C, which is about 12 feet long, 3 feet wide, and at
about the same inclination as the upper trough. It is fitted
with several riffles, which are sometimes supplied with mercury.
In working, a stream of water enters at the upper end of the
sluice-box, into which gravel is continually shovelled, while a
man breaks up the lumps with a fork, removes the large stones,
and puddles the lumps of clay. Two to four men can work at
one torn, and wash about five times as much in a day as can be
done by one or two men with the cradle. Only the coarse gold
is caught, and the machine is only suitable for washing small
quantities of rich dirt where there is a plentiful supply of water.

The material caught by the riffles is scraped out occasionally and
panned, but the riffle-box is too short for close saving of the gold.
The Puddling-tub. When water is scarce, as was the case in
many places in Australia where rich gravels were found, the
long-torn is inadmissible, and the puddling-tub is resorted to.

This is particularly well adapted for washing clays, and is still
used to disintegrate lumps of clay encountered in sluicing opera-
tions. It consists of one half of a barrel which has been sawn
in two ; into this the dirt is dumped and stirred up with water
by means of a rake, until all the clay is held in suspension in the
water, when a plug a few inches from the bottom is removed,
and the slime run off.The operation is repeated until the tub
is filled with gravel and sand to the level of the plug-hole, and
this residue is then shovelled out and washed by the pan, the
cradle, or by sluicing. Large boxes were used in Australia in
this way in early days, the rakes being worked by horse- or
steam-power; in 1860 no less than 3,958 boxes, worked by
horses, were in use in Victoria alone.*
The and puddling-tub are now little used
cradle, long-torn
except by the Chinese, who gain a precarious livelihood with
their help by washing over the heaps of tailings accumulated from
sluicing or hydraulicking operations in Australia and California.
TJie Siberian Trough.^ In Siberia, in the Urals and in the
valleys of the Obi, the Yenisei and the Lena, individual workers
Fig. 6.
stillexclusively use a trough, which differs from the long-torn

mainly in requiring more constant attention on the part of the
operator, and which resembles the old German buddle.
The
trough consists of a rectangular box open above and at one end.
When sandy gravels are being treated, the bottom of the disinte-
gration box (A, Fig. 6) which is about 40 inches square,
is
made of a perforated screen of wood or sheet-iron, having holes

of from ^ inch to 1 inch in diameter. The dirt is shovelled
into this box, and. contrary to cradle-practice (see p. 45) if
the gold is present in fine flakes, mercury is added here also,
the amount depending on the richness of the auriferous ma-
terial as determined by assay, the proportion used, however,
being never more than 10 of mercury to 1 of gold. Water is
directed upon the charge in the box, either by pipes from a
*
Philips' Met. of Gold and Silver, 1867, p. 139.
t For further particulars see the account given by Cumenge and Fuchs
in Fre"my's Ency. Chim., vol. v., L'or, 1st Section, p. 12.
reservoir or more often by

pumping, and the fine material is
carried through the screen and falls on to the
table, B, while the
pebbles are collected by hand and thrown away. If clay is being
treated, no screen is used; the lumps are puddled in the box, and
the mud carried over by an overflow of water. The table is
slightly inclined, about 20 feet long, and, for the greater part of
its length, is about 20 inches wide. It is furnished with five
riffles, of which two (E, C) near the top are about 2 inches
high,
while the others (D, E)are of less height. The
disintegration of
the sand is completed on the table with the aid of a small rake con-
tinually used by the workman. When disintegration is complete,
the stream of water is diminished in amount, and the workman
continues to rabble the sands which have accumulated above the
riffles, pushing the contents of the lower riffles up the table again,
until the water runs clear, and little except pyrites is left
behind the riffles where the so-called " grey concentrates "
accumulate. These are either concentrated further on the same
table, or removed and worked on a smaller table. In either
case the stream of water is still further reduced, being graduated
so as to carry away the last particles of quartz, together with all
materials of moderate weight, such as garnets, rutile, tourmaline,
<fcc., and even all the fine pyrites. If mercury has not been
added previously, it is sprinkled on before this last operation,
unless the gold is very coarse, when no mercury is added at any
stage of the proceedings.
The " black concentrates," thus obtained, consist entirely of
amalgam, magnetite, and the large grains of pyrites. The final
operation, by which the amalgam is separated, is the most difficult,
and requires the greatest amount of skill on the part of the
operator. The material is worked on the same table with very
little water, with the aid of a small rake, or more often with the
hand of the workman, who kneels down by the trough for the
purpose. Finally, all the pyrites having been washed away, the
magnetite is removed with a magnet, and the amalgam collected.
The tailings from the black concentrates are treated over again,
together with the grey concentrates.
The apparatus just described treats about 500 Ibs. of sand at
one time, and can be worked by one man, but usually gives
employment to four people (of whom three are frequently women),
who can treat about 5 tons of sand per day. The degree of
success attained depends largely on the skill of the workman ;
in Siberia and Russia the art is handed down from father to son,
certain families devoting their whole lives to the work during
many generations. These workmen attain such a degree of
dexterity in the use of the trough, that practically the whole of
the valuable contents of the gravels treated are extracted by
them, but the work is only suited to those who are content with
small earnings.
The Sluice. The sluice has replaced all these implements for
washing the gravel from shallow placers, where water is abundant.
Sluices are constructed of " boxes," each of which resembles the
upper part of the long-torn. The bottom of each box is made of
rough boards, about 12 feet long and 1J inches thick, cut 4 inches
wider at one end than at the other the total width is usually
;
from 16 to 18 inches, while the sides are 8 or 10 inches high.

It is considered by some New Zealand experts that sluice boxes
are usually made too narrow, so that the current is too deep and
fine gold is lost, while the flow is unduly checked by the sides,
thus creating undesirable eddies. The box is held together with
nails, and no attempt is made to render it watertight, as the
swelling, caused by the absorption of water, and the filling up of
chinks with sand and clay soon stops all leakages. The narrow
end of the box fits into the wide end of that next below it, and
so a sluice, made up of hundreds of boxes, can be rapidly put up
or taken down and moved to another locality.
In all but the smallest and cheapest sluice boxes, extra strips
of wood are affixed to the sides so as to protect them from the wear
Fig. 7.
Scale = -BV.
caused by the attrition of the stones and gravel carried through
by the current. When worn thin, these strips are replaced.
The bottom is similarly protected by the riffle bars, whose main
function is to catch the gold. These riffle bars are usually placed
longitudinally, and are strips of rough wood
from 2 to 4 inches
are
thick, from 3 to 7 inches wide, and about 5J feet long. They
wedged in the boxes at a distance of 1 or 2 inches apart by
means of transverse bars, so that two sets of riffles are placed in
each box in the manner shown in Fig. 7, which represents the
whole of one box and parts of two others. The rectangular
are well adapted to
depressions thus formed between the bars
intercept heavy particles that pass down the sluice, such as

all
gold, mercury, amalgam, pyrites, &c., which gravitate to the

bottom of the stream. Sometimes the rimes are placed trans-
versely, and sometimes for a short distance in zig-zag fashion.
This arrangement does not retain anything, but affords a better
chance of amalgamating the gold, which, together with the
mercury (in this case fed in constantly), slides down the inclined
riffles from side to side of the sluice, having sunk to the bottom
by virtue of its high specific gravity.

When all is ready a stream of water is turned into the head
of the sluice, where the dirt is shovelled in also. The amount
of dirt shovelled in per man depends on the height of the
lift and the nature of the soil, as well as on the labourer.
It varies from 3 to 7 cubic yards per diem, and should average
5 or 6. The first gravel sluiced fills up most of the riffle depres-
sions, leaving enough inequalities of surface, however, to inter-
cept and retain the mercury, &c.
The length of the sluice varies with the consistency of the
gravel, the fineness of the gold, the capital available, and the
fall of the ground. It must be sufficient to complete the dis-
integration and then to catch the gold. The length may be
adjusted by experiment ; if, in the clean-up, the lowest boxes
yield much amalgam, an addition to the length is necessary,
while if they yield none, the sluice may be shortened by the
removal of one or more boxes. Even in the latter case, however,
the tailings would almost certainly contain some fine gold. The
grade of the sluice is measured by inches per box, so that a grade
of " 12 inches" means one of 12 inches in 12 feet. The usual
grades are from 8 inches to 20 inches per box, depending
(1) On tJie fall of the ground, since the sluice cannot be raised
far above it, nor sunk deep into it, owing to the increased
expense thereby occasioned.
(2) On the nature of the dirt to be washed. Tough, tenacious,
clayey, or cemented gravels require higher grades to effect their
disintegration than loose material. Instead of being disintegrated,
clay sometimes becomes aggregated into balls, which roll down
the sluices, picking up particles of gold previously caught in the
riffles, and these lumps of clay must be removed by hand and
puddled. There must be sufficient grade to enable the water to

carry away all but the largest stones, so as to avoid unnecessary
hand-picking, but on the other hand, while coarse gold is readily
caught, fine particles are lost if the current is too rapid.
(3) On the quantity of tvater available. The reduction of the
grade lessens the duty of the water, so that if the supply of the
latter is short or costly, the grade is made as steep as possible,
consistent with saving a fair proportion of the gold. A steep
grade reduces the necessary length of the sluice, as disintegration
takes place sooner. Since a steep grade, a rapid flow, and deep
currents are best suited to effect speedy and thorough

disintegra-
tion of the gravel, while a low grade, and slow and shallow currents
are best adapted for saving the gold, the upper part of a sluice,
for a sufficient distance to effect the complete
disintegration of
the gravel, is sometimes made of higher grades, or with narrower
boxes than the lower part, which is occupied solely in
catching
the gold. When this is done additional supplies of water should
be introduced at the point where the change is made, otherwise,
the duty of the water being reduced, the sand packs in the
angle
where the grade is altered, and constant attention is required to
prevent the stream from overflowing.
The opposite requirements of disintegration and gold-saving
are more often supplied by "drops" and "undercurrents." A
vertical fall of the pulp constitutes a drop, which is
arranged as
follows: The sluice terminates in a "grizzly," or inclined grating
made of parallel iron bars placed longitudinally to the stream
and from 1 to 6 inches or more apart, according to the exigencies
of the case. All the water and fine stuff pass through the
grizzly and fall a distance of from 1 to 10 feet into a sluice
below. The larger stones or boulders roll down the inclined
bars, and are shot over a precipice (if possible) or on to a steep
slope outside the sluice, as, unless some arrangement for remov-
ing these rocks is made, they will accumulate until they can no
longer roll off the grizzly. The higher the fall, the more
effectively it acts in causing disintegration. Sometimes, near
the head of a sluice, the grizzly is omitted from a fall, and the
boulders are retained to help in breaking up the gravel. The
chief disadvantage in permitting them to remain with the rest of
the gravel lies in the fact that they wear out the sluice, and that
much water is required to wash them down.
The undercurrent is often used in conjunction with a drop. A
grizzly with bars placed close together allows most of the water
and fine material to pass through, while the coarse stuff is carried
over and falls into the main sluice below. The fine material is
carried off by a short sluice placed at right angles to the general
direction of the main sluice, and is discharged into the upper end
of a large broad box from three to ten times as wide as the
sluice, and with its long diameter parallel to the main sluice.
A number of check-boards help to distribute the stream evenly
over the whole width of the box. This box, to which the name
undercurrent is often given, although it properly belongs to the
whole arrangement, is usually of lower grade than the sluice,
and is in that case supplied with additional water. A broad,
shallow stream is thus made to flow with reduced velocity over
the surface of the box, and, as the latter is plentifully supplied
with riffles and mercury, considerable quantities of fine and
"
"rusty gold and fine particles of amalgam, which would
otherwise
be swept away and lost, are retained by it. The tailings from the
undercurrent are discharged into the main sluice below the

drop.
Both grizzlies and undercurrents are used more frequently in
hydraulicking than in shallow placer sluicing, in which the large
stones are usually removed by a man with a blunt pronged fork,
who also either breaks up the lumps of clay or removes them
and puddles them in tubs.
The Use of Mercury in Sluicing. Mercury is added at the
head of the sluice after washing has been in progress for a suffi-
ciently long time for all leakages to have been stopped, and for
the lowest depressions to have been filled in with sand. The
mercury is broken up into small particles either by pressing it
through chamois leather above the sluice, or by letting a thin
stream fall on to a man's hand, by which it is scattered, or by
some similar method. The amount added varies with the rich-
ness of the dirt and the magnitude of the operations, enough
being added to dissolve the amalgam formed. It is carried down
the sluice and lodges in the riffles, the greater part being retained
in the first few boxes. Fresh supplies are introduced every few
hours at the head of the sluice, and sometimes at various points
lower down the sluice also ; in particular, mercury is added to
the undercurrents, as it is especially valuable in catching the
finer particles of gold which would otherwise be lost, whilst coarse
gold can in great part be saved without mercury. Sometimes
the latter is forced into the substance of the wooden riffles by
driving an iron gas pipe into the wood, and filling it up with
mercury, which is forced by the pressure of the column through
the pores of the wood. The amalgam then forms on the surface
and in the interstices of the blocks, and in cleaning-up this is
scraped off. A better plan is to use amalgamated copper plates,
which is now often done. These resemble the plates used in
stamp batteries, described on p. 117.
The Clean-up. The length of each " run," at the end of which
the boxes are cleaned-up, varies, according to the richness of the
gravel, from a day to a whole season, and is usually a week.
The upper part of the sluice, which retains most of the gold, is
usually cleaned-up more frequently than the remainder. A
clean-up is begun by discontinuing the supply of gravel, and
letting the water continue to flow until it passes through the
sluice quite clear. The first six or eight sets of riffle bars are
then taken up, and the sand, mercury and amalgam washed
down, all the latter being caught by the first riffle left in. It is
scooped out thence by a wooden ladle or iron spoon into a
bucket, and the rich sand is collected and panned. The next
few riffle bars are now taken up, and so on, or alternatively
the work may be begun on several sections at the same time.
Lastly, the whole sluice is carefully searched over, and particles
of amalgam or mercury picked out of every crevice where they
have lodged by spoons, penknives, &c.
SHALLOW PLACER DEPOSITS.' 53
The mercury thus collected is quite liquid, and is strained

through chamois leather to separate the solid amalgam, which
is then retorted. The retorts used in well conducted enterprises
are similar to those in use in stamp mills, described on pp. 133-135.
Amalgam obtained as the result of operations on a small scale,
however, is often merely heated on a shovel over an ordinary
fire, the mercury being driven off and lost.
Care must, of course, be taken to prevent partial clean-ups by
unauthorised persons. A man with a shot-gun stationed by the
sluice is a frequent preventive measure.
Tail Race. The tailings from sluicing operations on low
ground which has not much fallare removed through a covered-in
wooden sluice, or, better still, through a large iron pipe. The work
proceeds in the up-stream direction, and the worthless material
stripped from above the pay-gravel is thrown on the top of the
tail race, which thus passes through a mound of earth and dis-
charges into the open air lower down the valley. As the digging
and sluicing progresses up-stream, the tail race is lengthened and
the sluice boxes proper are conveyed further up the valley so as
always to be near the auriferous material last uncovered. This
method originated with the Chinese.
Ground Sluice. In some cases boarded sluice boxes are not
used, but a stream of water is conducted to a little trench cut in
the pay-dirt, which is soon enlarged by the action of the water,
while the banks are at the same time shovelled or prised by the pick
or crowbar into the sluice. The gold is caught in the natural
riffles afforded by the uneven wearing of the bed, or rocks may
be added to arrest the gold, no mercury being used. Ground
sluicing is only adopted where the water supply is precarious, or
the season very short, so that violent rains cause floods that
would sweep away sluice boxes, and then are succeeded by dry
intervals during which the boxes would warp and crack. Only
the coarse gold is saved, while the duty of the water is usually
much less than in wooden sluices. After a time, usually when
the water gives out, the auriferous material is collected from the
sluice and washed in a long-torn or cradle.
Booming. This method of sluicing, which originated in the
United States, is adopted when the water supply is insufficient
for continuous operations. A
dam with a light gate, capable of
being easily lifted, is built just above the part of the valley
where the auriferous gravel is situated. The water trickling
down the valley accumulates behind the dam, and finally over-
flows at one point into a small rectangular box fastened to the
end of a long lever. When full of water this box depresses its
end of the lever and raises the dam-gate, so that all the
accumulated water rushes out at once and scours the valley
bottom. As the box falls it empties itself of water, and the
dam-gate returns to its original position by its own weight.
This device is usually employed in connection with ground-

sluicing, but a line of sluice-boxes might be used through which
the sudden flood could carry gravel piled just above the head of
the series.
Dry Diggings. If no water can be obtained, it is sometimes
profitable to concentrate very rich pay-dirt by winnowing,
tossing it in a blanket until the lighter particles have been
blown away, and finishing on a batea with or without mouth-
blowing. It is of course a wasteful method of concentration.
Several machines have been invented for dry concentration,
some of which have attained a fair measure of success. These
machines all use a blast of air by which the sand is kept partly
in suspension, while it is moved by gravity down an inclined
table which is furnished with riffles. The auriferous material must
be quite dry or perfect disintegration cannot be accomplished.
In a typical dry washer, the gravel is first made to pass through
shaking screens, the mesh of which is adapted to the character
of the material. The object of the screen is to eliminate the
larger fragments, which are usually barren. The shaking screen
delivers the material on to an inclined table formed of a wire-
screen, covered with light canvas or some similar material
through which are forced pulsating blasts of air. These sudden
puffs throw up the sand and let it settle again alternately, and
as a result the light material works down the table, while the
gold is retained by the riffles, being too heavy to be tossed over
them by the air.
Cement Gravels. When gravels are cemented by iron oxides
or clay so as to be too hard for disintegration in the sluice,
except with the aid of a great number of drops, they are either
sent to the stamp-mill or, with more advantage, treated by a
cement-mill before being washed in the sluice. The treatment
of cemented gravel is best preceded by exposure to the air
for a time, frost, sun and rain having a very rapid effect on it.
The cement-mill is a large iron pan fitted with coarse screens ;
it has revolving arms carrying plough-shares, which puddle
the clay and break up lumps of gravel, but leave boulders
untouched. Water, fed in with the dirt, carries the fine stufi
through the screens, and at intervals the boulders are removed
by various contrivances.
Tail Sluices are sometimes erected to intercept the tailings
from one or several sluices with the object of collecting a further
percentage of gold from the waste material. These tail sluices
are made of much greater size than those described above, and
in some cases pay for construction. The Kumara sludge channel,
erected by the New Zealand Government to carry the tailings
from the sluicing works of the district into the river, caught
957 ozs. of gold in the four years ending in 1891. This sluice is
3 feet 6 inches wide, and has a grade of 1 in 28, while the boxes
which discharge into it are only 18 inches to 22 inches wide.

Usually, in good work, the sluices are long enough to make the
tailings too poor to be worked over again at a profit, except by
the Chinese, until after they have been enriched by natural
concentration in. the rivers.
I/'ly Catchers
were invented in Australia for the purpose of
catching the fine particles of gold, which, successfully evading
,the riffles of all sluices, float down on the surface of the rivers.
These devices consist of weirs constructed on piles driven into
the riverbed, and stretching across from bank to bank of the
river. Boards covered with blanketing or coarse gunny -sacking
are attached to the weirs and collect all particles floating on the
surface of the water. At intervals the blankets are taken up
.
and washed in a tank. These fly catchers soon pay for their
cost of construction on many rivers, but are liable to be
damaged by floods, and by being used as bridges by men and
animals.
River Mining. River mining consists in the working of
auriferous gravels in the channels and beds of existing rivers,
and may with convenience be made to include the exploitation
of deep bars below the level of the water. It is conducted in
three different ways :
1. Aportion of the river bed is laid dry by damming and

fluming.
2. The gravel is raised by dredges or similar contrivances,
operated from a boat.
3. on the bank, and the gravels below water
Pits are sunk
level excavated and raised to the surface by ordinary mining
methods, or by the hydraulic elevator. The gravel won by
either of these three methods is washed in the usual way by
sluicing.
1. River Mining Proper. This method is carried on chiefly
in California. In the case of the larger, rivers, or where only a
small capital is available (from 100 to 1,000), wing-dams
are
built out from the bank, above and below the part of the river
which it is desired to work, and a third dam, built parallel to
the direction of the current, connects their mid-stream ends.
The dams are usually constructed of boulders, with their inter-
stices filled with gravel, and are only slightly higher than the
river surface at low-water level.
(summer) The water-flow is
forced to the other side of the river channel, its level being little
altered. The space thus cut off is pumped dry by a Chinese
pump, worked byan undershot wheel placed in the current by
the side of the mid-stream dam. The pay-dirt is usually covered
and is exposed by
by a mass of barren boulders and gravel,
shafts and running drifts. The dirt is
stripping or by sinking
taken out down to the level of the bed-rock, which is cleaned
searched. The pay-dirt is

thoroughly, all crevices being well
washed by sluicing, the water being supplied directly from the

river, or, if necessary, raised to the required height by a dip-
wheel (an undershot water wheel with small buckets placed near
its periphery, which empty their contents into a flume). The
boulders are removed by a derrick worked by water-power. As
the construction of the dam can only be begun when the water
is low, the season is usually very short sometimes only a few
;
days or hours are left, after the draining and stripping have been
finished, for the actual collection and washing of the pay-dirt.
Operations are stopped by the autumnal rise of the river, which
overtops the dam and fills the pit, often coming so suddenly as
to leave no time to remove any of the tools and machinery.
Sometimes large returns of from 100 to 1,000 per day are
obtained in this short time, and these may be sufficient to yield
handsome profits on the. undertaking, but the dams are always
swept away in winter.
"When enough capital is available, or the river to be operated
on is small, head- and foot-dams are made, stretching across from
bank to bank, and the water is carried off in a wooden flume and
delivered into the river bed below the foot-dam. These dams
are usually built of timber, faced with planks, and supported by
earth and stones. The portion of the bed thus isolated is drained
and worked as in wing-dam practice, the source of power being
the water-race in the flume. The work is also in this case often
cut short by the rising waters, which carry away the dams,
flume, sluices, wheels, &c.
Of late years efforts have been made in two or three places in
California to prevent the winter floods from stopping work. On
the Feather River a permanent head-dam and flume has been
constructed by an English company to drain a part of the bed>
and tunnels have been made on the American and Feather
Rivers to permanently drain large reaches and deliver the water
at a lower point. None of these undertakings have as yet been
strikingly successful, from various causes.
River mining is probably subject to more uncertainty than any
other branch of gold mining. The whole capital invested may be
lost, and all works and machinery swept away by a flood before
the pay-dirt is sighted, while numerous instances are on record in
which the alluvium on the river-bed, after having been laid bare
at great expense, was not rich enough to pay for sluicing.
Between the years 1857 and 1880 the Californian river-beds
were covered so deep by the tailings from hydraulicking that
they could not be worked with advantage. Since the suspension
of hydraulicking, however, and the gradual working down of the
debris, some places have again become worthy of attention.*
*
For a account of river mining, with details of dam-construction,
full
&c., the studentis referred to the article 011 the subject by R. L. Dunn in
the Ninth Annual Report of the Cali/ornifin State Mineralogist, 1889, pp.
262-281. See also the Eleventh Report, 1892, pp. 150-153.
2. Dredging. This method of winning the gravel has seldom

met with any success, for the reasons that the gold, even when it
exists in the river bed, cannot be got at in such a simple way.
The bed-rock cannot by this means be cleaned and creviced, and
if the poor gravels on the surface are thick, they cave in and
slide down into the pit made by the dredge, so that the pay-dirt
is never reached. Moreover, boulders impede or prevent the
work. The favourite implements are various forms of vacuum,
lifts and pumps, the power being sometimes supplied by a hy-
draulic elevator. Water, gravel and stones are brought into the
barge together, and the stuff is usually then and there washed and
dumped into the river again. The system has made somewhat
more progress in New Zealand than elsewhere, and the following
details of recent work there are taken from a Government report
on mines.* On the Molyneux River in New Zealand the centre
bucket dredger has done good service, the difficulty of raising
big stones being that most severely felt. Here it is found that
little of the fine scaly gold of the rivers is caught in ordinary
sluices, and the washing is done by separating all the coarser

material by means of trommels, and passing the fine sand by
itself over wide tables in a very shallow stream. The volume of
water must be just enough to keep the tables free from sand, as if
the latter begins to collect, it is believed that the fine gold is not
being caught. The tables are covered with cocoa-nut matting,
baize, blanketing, or even plush, each material having its advo-
cates. On the Shotover River, the Sand Hills Dredging Company
separates the stones by passing the gravel through a revolving

cylinder 3 feet in diameter and 10 feet long, set with fine holes ;
30 tons per hour are treated by this trommel. The washing is
done on inclined tables which are 64 inches wide, and about
27 feet long, having a grade of 18 inches in 12 feet. These
tables have a superficial area of 144 square feet only, which is
insufficient for the treatment of 30 tons per hour. At the
Waipapa Creek Dredging Company's works, which are situated
on the sea coast, a Welman dredge is used, which acts on the
centrifugal suction system. The pump is 3 feet 6 inches in
diameter, and the delivery tube 13 inches in diameter, and the
pump raises gravel from an area having a radius of 40 feet. All
large stones are caught and separated from the fine stuff by
a
riddled hopper-plate. The gold is very finely divided, and is
caught on plush mats which are washed every eight hours. The
water for washing is supplied from a reservoir by means of an
18-inch pipe. Stones of 56 Ibs. weight are lifted by this pump,
which may be used with advantage on all low wet ground.
In Northern Italy a dredging plant has also been successfully
at work for some years.
3. Deep Bar Mining. Deep bar mining, as conducted by
*
New Zealand Mng. Comm. Report, 1891.
the old methods, consisted in sinking shafts in the bank and

running drifts below water level, either under the river bed
itself or under deep bars, for the purpose of digging out the pay-
dirt from the surface of the bed-rock and taking it to the surface
to wash. It was mostly unsuccessful owing to the amount of
water, quicksands, &c., encountered, although to-day these diffi-
culties could be often overcome by the present improved practice
in quicksand digging. As, however, this kind of digging is
better suited to the efforts of a few individuals than to under-
takings on a large scale, it is not likely to be largely reverted to.
In particular, working in loose detrital matter under the river
bed itself must always be hazardous to the workers, and the
chances of success very dubious, although profits have been
obtained where the pay-dirt was only worth 70 cents per cubic
yard. Deep bars, however, are now being worked profitably
by the hydraulic elevator, which has become of great import-
ance in placer mining. This machine may be more conveniently
described after hydraulicking has been considered.
Method of Working Siberian Placers. The methods and
apparatus employed in Siberia differ so markedly from those
which have been adopted elsewhere, that they are well worth
a special description, although they cannot usually be applied to
placers found in other parts of the world, owing to the difference
in economic conditions. In California the valleys are narrow
and the grade steep, so that watercourses are usually close to
the auriferous deposits, and the sluices can be made of almost
any length, while still conforming to the general slope of the soil.
In Siberia the slope of the valleys is so gradual that the flow of
the water is almost imperceptible, and takes place through wide
marshy tracts. The result of this is that the sluices must be
short, being usually less than 100 feet long, and their upper ends
must be raised on trestles. As a further consequence, also, the
gravel must be excavated by hand and carried in waggons to the
sluice, and the tailings removed in a similar way, the flow of
water acting by gravity not being available for these purposes.
The excavation is made in benches or terraces, working up the
valley. The height of each bench above the lower one is about
5 feet, and the gravel is picked down from the face of each bank
and shovelled into carts by which it is conveyed to the washing
establishments. The additional expense entailed by these causes
is balanced by the low cost of labour in the
country, and an inci-
dental advantage lies in the fact that the washing apparatus can
be placed outside the limits of the river during flood time, and
so may remain for a number of years undisturbed, while all the
gravel in the district is being washed. Only surface deposits
are exploited, no deep workings existing in the country.
There are three types of apparatus employed, each designed
for washing a particular kind of deposit. They are :
The Siberian sluice, which is used to wash light sand.

1.
The Trommel, used for loamy sands.
2.
3. The Pan, used for gravel which is cemented together
by
means of compact clay.
1. The Siberian Sluice. The apparatus at Voltchanka, which
may be taken as a type, consists of a head sluice and three
secondary sluices, which are placed at right angles to the head
sluice, and which leave it at different points and converge to a
common centre, where the tailings are discharged. The head sluice
begins at a height of 1 3 feet from the ground it is about 90 feet
;
long by 2 feet wide, and has a fall of about 1 in 14. The sands
are dumped from the waggons on to a wooden platform situated
above the sluice-head, and shovelled into the latter, a stream of
water being turned in at the same time. After passing through
a grizzly, the gravel runs over a series of cast-iron cross-bar
riffles, which form a number of rectangular depressions (or pigeon-
holes) in the bed of the sluice, by which the disintegration is

favoured. The stream then flows over an iron screen, through
which a part of it falls into the first secondary sluice, while the
remainder continues its course over more pigeon-hole riffles.
This arrangement resembles the Californian undercurrent. A
second and a third screen open on to the other secondary sluices,
and the part of the gravel (consisting chiefly of small stones)
which has resisted disintegration, and has not passed through the
screens, is then let fall into a hopper, whence it is removed to
the tailings waggons.
The secondary sluices are wider than the head sluice, and have
a steeper grade, and the amount of water and auriferous
material passed is of course much less in each one than on the
principal sluice. The sands first pass over a number of trans-
verse riffles, and then over about 30 feet of blanketing, the sluice
being widened at the same time, and subdivided by longitudinal
wooden partitions, and the grade being as much as 1 in 6, while
small drops are introduced at intervals of a few feet. The third
sluice has a more gentle inclination than the others. The tail-
ings fall into a shallow sump, and are instantly
"
lifted out of it
"
by a bucket elevator or tailings-wheel operated by water
power, and stored in a hopper, whence they fall into waggons,
by which they are removed to the dumping ground.
No mercury is used in these sluices, and only the production
of " grey concentrates is attempted, this work being continued
"
from 6 a.m. to 7.30 p.m. every day, after which the concentrates
are collected from the riffles. They consist of gold in scales and
plates, magnetic iron oxide, pyrites, rutile, together with some
quartz, tfec. They are treated with mercury in the Siberian
trough or on inclined tables, the method being that described
on p. 47.
The apparatus described above treats 500 tons of gravel per
day, the labour required being furnished by twenty men and ten
horses. The gravel treated contains an average of from 12 to 15
grains of gold per ton, rarely falling below 6 grains per ton ; the
exceptional richness of 3| dwts. per ton has been observed. The
gold is chiefly found in the head sluice, where 70 per cent, is
caught, 30 per cent, being caught on the secondary sluices. At
a similar establishment at Tchernaia-Retchka, however, where
the gold is less finely divided, 97 per cent, was caught on the
head sluice, and only 3 per cent, on the secondary sluices. The
amount of water used at Yoltchanka is about six times the
weight of the gravel. The cost of construction of the works
was 70,000 roubles, or about 7,000.
2. The Trommel. Gravels which are too compact for satisfac-
tory disintegration in the short sluices described above are sub-
jected to a preliminary treatment by a trommel. At Berezovsk
the trommel of sheet iron of 9 mm. thick, having holes in it of
is
about 1 mm.
in diameter. The trommel is about 12 feet long,
3^ feet in diameter at one end, and 4i feet at the other, and is
set inside with denticulated plates of iron to assist in the disin-
tegration effected by the water. The machine is driven by a
water wheel, and is sufficient for the disintegration of from 400
to 500 tons of gravel per day, requiring the expenditure of about
3 horse-power to drive it. The amount of water used in the
trommel and on the tables is 67 '5 litres per second, or about
seven and a-half times the weight of the ore. The washing is
effected on inclined tables only 30 feet long and 12 feet wide,
and with a grade of about 1 in 4, placed with the incline at right
angles to the length of the trommel. Near the head of the
table, and stretching across it, is a deep trough-like depression,
and below this there are a number of transverse riffles, in which
grey concentrates are caught and treated as usual. The Bere-
zovsk establishment employs twenty-five men and fourteen horses,
constantly the trommel usually lasts for two seasons.
;
3. Pan
Washings. Sandy clays cannot be economically
disintegrated in a trommel, and are, therefore, treated in a
washing pan, which bears a strong resemblance to the cement
pans employed in California. The pan usually consists of cast
iron, and is from 8 to 16 feet in diameter, with vertical sides
from 1 to 5 feet high. The bottom is of cast iron or sheet iron,
and has numerous holes in it of about TV inch in diameter,
widening downwards. The bottom is divided into 25 sectors,
between which are deep grooves for the collection of the pebbles.
Through a circular opening in the centre of the pan there
passes a revolving axis to which are suspended eight horizontal
arms studded with vertical iron teeth, some of these being shaped
like plough-shares. The revolution of these arms effects the
disintegration of the sandy clays, which are fed into the pan
together with water and puddled until fine enough to pass
through the holes in the bottom, and the stones are removed
at intervals by opening little gates placed opposite the radial
grooves. The disintegrated gravel falls from the pan on to
concentration tables, similar to those used after disintegration
in the trommel. At Berezovsk the pan is 11 J feet in diameter
and 5 feet deep, and the arms revolve at the rate of 25 turns per
minute. From 50 to 55 tons of material are treated in twelve
hours, the water consumed, including that required for power,
being ten times the volume of the sand.
Beach Mining. Beach mining is a comparatively unimport-
ant form of shallow placer mining. The sea beaches on parts
of the coasts of California, Australia and New Zealand contain
small quantities of gold, which has been proved, in all cases in
which the matter has been investigated, to be derived from the
cliffs, which mostly contain a still smaller quantity. Some
streaks of black sand, however, in the " Gold Bluff," California,
have yielded $135 or 6| ozs. per ton by actual working.* The
waves of the sea wash down and partially concentrate the poor
sands, and, under certain rather exceptional circumstances, as
the tide goes out the surface of the beach is left covered with
black sand, in which numerous specks of gold occur. This is
carefully scraped up and transported inland to be washed, as sea
water is not well adapted for the purpose, although it is used by
one Californian company. The next tide usually washes away all
the valuable material which has not been collected, or else covers
it with barren sand. There is great difficulty in washing the
black sand in California, as it consists largely of rounded grains
of magnetite, the density of which is about 5-0, while the gold
is in minute flakes and scales, which can be seen under the
microscope to be oblong in shape, and thicker at the sides than

in the middle (a shape due to continued pounding of a malleable
'
material). This form is so easily moved and buoyed up by water

that it is difficult to get a "colour" with the pan, and the
amount caught by the mercury in sluices or long-toms is
usually an insignificant proportion of the total assay value of
the sand. The industry is generally a languishing one.
Treatment of Shallow Placer Gravels by Steam Shovels
and Amalgamated Plates. A plant for the treatment of placer
deposits, which are similar in nature to those worked in Siberia,
has been devised in America, where labour-saving contrivances are
indispensable in all such cases, owing to the high rate of wages.
The apparatus consists essentially of a combination of a steam
shovel or navvy, and an amalgamator consisting of a wide waggon-
shaped wrought-iron trough loosely lined with silver-plated amal-
gamated copper plates, which form a series of steps or riffles at
the sides of the trough. The material is elevated into a hopper
by the excavator, and thence is charged into a revolving trommel
*
Prod. Free. Met., U.S.A., 1884, p. 557.
placed inside the trough. Here the disintegration of the gravel

is effected, and the fine material falls through into the trough,
while the stones are discharged outside at the end of the
trommel. At the bottom of the trough there is a water pipe,
carrying water at high pressure, and in that pipe a series of jets
pointing alternately forwards and backwards. The result is to
give a series of eddies or whirlpools in the water in the trough,
and the sand and fine gold is continually carried up to the
surface near the middle line of the trough, and in descending
again near the sides it comes in successive contact with several
of the plates, which form a series of steps. The fine sand is
eventually discharged at the end of the trough. It is stated by
the Bucyrus Steam Shovel and Dredge Company, by which the
machinery is manufactured, that such a machine erected in
Montana has a capacity of from 600 to 800 cubic yards of gravel
per day, all the water required being supplied by a pump raising
500 gallons per minute, the proportion being only three of water
to one of gravel, which seems much too small. It is stated that
gravel containing only 12 cents of gold (i.e., about 3 grains) to
the ton has been treated successfully by this machine, but exact
records of continuous work done by it are wanting. It will be
noted that special means must be adopted in each case for the
handling of the tailings.
CHAPTER V.
DEEP PLACER DEPOSITS.
Nature and Mode of Origin of Deposits. This discussion is
necessary in order that the description of the methods of treating
the deep placer gravels may be intelligible. Both in Australia
and California, besides the superficial placer deposits situated in
or near the existing rivers, which in the deep canons of the
Klamath and other rivers in the extreme north of California
attain a thickness of 250 feet, there exist auriferous gravels
which bear no apparent relation to the present drainage of the
country. These gravels often attain enormous thickness, and
are in many places covered by volcanic rocks, consisting of
basaltic lavas and tuffs, which are sometimes interbedded with
gravel and loam. This latter circumstance shows that inter-
mittent action of the volcanic vents, with long intervals of
repose, has taken place. There has been some difficulty in
accounting for the origin of these deep placers, and it has been
ascribed in succession to the agency of the sea, of ice, and (for
California) of a huge river flowing from north to south at right
angles to the direction of flow of the existing rivers. None of
"
these views are now entertained, and the " fluviatile theory is
DEEP PLACER DEPOSITS. 63
generally accepted, the origin of the gravel being ascribed to the

depositions of ancient rivers flowing in courses roughly parallel
to those of existing rivers. The geological age of these ancient
rivers has not yet been determined with certainty, but though
they may be Pleistocene, the balance of palseontological evidence
is perhaps in favour of Whitney's view that the deposits were
formed in the Pliocene period.
The ancient Californian rivers probably had their sources at
somewhat higher altitudes than those now existing, and had
more uniform general grades, the slope of their beds corre-
sponding more nearly to the general slope of the country. The
existing rivers, on the other hand, have steep grades in the upper
parts of their courses, followed by comparatively level stretches
below. The old rivers, however, like their successors, had rapids,
falls and level stretches, the grade varying from 5 feet to
250 feet or more per mile. The Pliocene rivers ran in valleys
which were broad and shallow in comparison with the present
deep precipitous canons, and the volume of water was in general
much greater than that delivered by their representatives of
to-day. The width of the valleys varied from 100 feet to fully
1J miles (which is the width at Columbia Hill), and the depth
must have been often over 1,000 feet. These valleys were
already partly filled up by accumulations of gravel, when the
outbreak of volcanic activity in many cases filled up the remain-
ier, and the streams were deflected into other channels, which
Dften lie close alongside the old cailons. These new channels
have been excavated by the running water until they now lie
much below the level of the beds of the Pliocene rivers, and
consequently the gravels which were deposited in the old valleys-
now sometimes crown the highest ground in the district, the
general level of the country having been greatly reduced in
height. The new channels have been cut partly in the old
country rock and partly in the Pliocene auriferous gravels and
their covering of volcanic rocks. Sometimes the course of the
present cailons cuts that of the
old at several points owing to
the sinuosity of both, see Fig.
8, in which A represents the
modern river, and B the
ancient one. The result is Fig. g.
that sections of the old valley
from bed-rock to surface are exposed in the sides of the
level of the
canons, usually at some height above the present
water, and it was at such points as these that the discovery
of the existence of the deep placers was first made. The hard
covering of basalt has served to protect the more
friable gravels,
which have been for the most part removed in those places
where the lava has been worn away or has never existed, so that
the largest tracts of gravels still existent lie beneath the volcanic
rocks.
Fig. 9 represents a section across two ancient channels (B, B)
and a modern canon, that of the American river. Here, A is the
volcanic capping, which is 800 feet thick above the Red Point
channel ; B, B are the auriferous gravel channels C, C are
;
deposits of gravel on the

"
rims," containing gold in places is ;
D
the bed-rock, consisting of dark-blue slates ; E is a barren deposit
of angular debris and boulders F F are prospecting tunnels,
;
which were put in at too high an altitude F' is the tunnel;
bored with the object of reaching the bottom of the gravel de-
posit ;
H
are prospecting winzes sunk in order to discover the
position of the gravel. The space included within the dotted
lines N" MY
M' N' has been obviously denuded since the deposi-
tion of the volcanic cappings, the soft slate rims NM
II and N' M'
Fig. 9.
having been worn away, while the hard lava has resisted erosion.
The vertical depth from M
to the American river is about 1,800
or 2,000 feet.
This condition of things is that prevailing in California, but in
Victoria the structure closely resembles that just described, with
the exceptions that the old valleys were smaller and that the
erosive action of the rivers since the deposition of the basalt has
been comparatively slight, owing to the slight grades of the
streams caused by the low elevation of the country and to the
small amount of the rainfall. In consequence of this the basalt
has usually not been worn through, and the " deep leads " or old
river bottoms are often below the level of the present streams,
so that although a larger proportion of the Pliocene gravel
remains, it is more difficult and expensive to mine.
The shallow placers, at any rate in California, have resulted in
the main from the erosion of these deep placers, the materials of
which, having undergone a natural concentration in the ground
sluices afforded by the river beds, furnished the
wonderfully
rich river-bed and bar deposits, which yielded so much gold
between 1848 and 1860. The deep level gravels vary greatly in
thickness, as has already been stated, being only 2 feet thick at
Table Mountain, Tuolumne County, and over 600 feet thick at
Columbia Hill, Nevada County, California, and averaging from
100 to 300 feet, the thickness varying with the nature of the
river bed, and the subsequent erosion. The gravel consists in slaty
districts chiefly of quartzose sand, the fine materials furnished
by the disintegration of the slate having been for the most part
swept away, and the products of the quartz veins contained in
the slate being left. Near bed-rock, but at no higher level,
there is often a collection of large boulders, varying in size up
to 10 feet in diameter, consisting mainly of quartz, but sub-
ordinate to these are others usually similar in character to the
bed-rock. These boulders, though rounded, are too large to have
been transported far by running water, and have probably been
polished by the attrition of the sand carried over them. At the
Forest Hill Divide, Placer County, California, the gravels consist
almost entirely of pure white quartz, but at other places such
quartz is rare. It is to be noted that it is only in slaty districts,
where the gravels are mainly quartzose, that rich auriferous de-
posits occur. In granite districts, where the gravels are composed
of more heterogeneous materials, and in cases where they consist
of volcanic boulders and detritus, little or no gold is found. The
lower parts of the gravels are often cemented into a conglomerate,
called " cement," by infiltration of silica, oxides or sulphides of
iron, or, rarely, carbonate of lime ] when the gravels are covered
with lava, the whole thickness is in some cases converted into
cement. The upper parts of the gravels often contain pipe-clay
in greater or less quantity, either in pure beds or mixed with
sand. Fossil leaves occur in the clays, and drift wood occurs
throughout the whole of the deposits in extraordinary abund-
ance, particularly in Australia ; this wood is for the most part
silicified or replaced by sulphide of iron. The higher portions of
the gravels are often altered by the action of air and water on
the iron sulphide, thus forming ferrous salts, haematite and
hydrous sesquioxides, which colour the gravels red and brown
" red
respectively. The upper gravels are hence called gravels."
The lowest layers, being protected from alteration from above,
are coloured dark blue-grey by the ferrous sulphide contained in
them, and are hence called "blue gravels," their occurrence
giving origin to the old "blue lead" theory, owing to their
uniformity of colour over wide areas in California. The sul-
phide of iron which incrusts fossil bones and teeth found in the
gravel, and replaces the substance of drift wood,
was formerly
5
believed to be derived from below, as the result of metamorphic

action going on in the country rock.
Distribution of Gold in the Gravels. The gold is found
chiefly either in contact with or just above bed-rock. If this con-
sists of soft slate, and especially if the planes of cleavage are at a
high angle to the horizon, particles of gold are often found in the
natural riffles thus formed, and are disseminated through the
rock to the depth of a foot or two. If depressions, pot-holes, or
fissures exist in the old river bottom, they are usually very
rich in gold. Where, as often happens, there is a channel, or
"
gutter," to adopt the Australian expression, cut by the stream
in the lowest part of the valley, the gravel filling it is usually
much richer than that found elsewhere. Such rich portions,
often only a few feet wide, and of insignificant depth, but extend-
ing to considerable distances in the direction of the stream, are
called "leads." As a result of these circumstances, the "blue
" "
gravels happen to be richer in general than the red gravels,"
from which arose the old theory that only blue gravel pays to
work. Coarse gold and nuggets chiefly occur near bed-rock
"
in the deeper parts of the channel, but the " rim-rock gravels
"
are also often rich. The " rim-rock is that portion of the bed-
rock which forms the sides of the old valley, thus lying consider-
ably higher than the central channel. The richness of gravels here
doubtless arises from the existence of old bench or terrace gravels,
which are consequently the oldest of the whole series, being
formed before even the gutter gravels. Rich streaks also occur
at various levels in the gravels, often resting on " false bottoms,"
which consist of impermeable beds of clay or some similar
material. Sometimes these streaks are richer than those encoun-
tered at bed-rock, as, for example, at the Paragon Mine, Placer
County, California. Although it is concentrated in this manner
at various points, gold nevertheless occurs disseminated through
the greater portion of the red gravel, where, however, it is in a
finer state of division and less abundant. Besides existing as
free particles, gold may occur in quartz boulders, although this
is rare. For instance at the Polar Star Mine, Dutch Flat,
California, a white quartz boulder was found, which contained
288 ozs. of gold. Gold may also occur, together with pyrites,
replacing the substance of drift wood.
The amount and position of the gold varies, as in the case of
the present rivers, with the grade, the shape of the valley, the
volume of water, the amount of gravel being carried down, <fec.
"An underloaded current i.e., a current charged with less
detritus than it is well able to carry is apt to cut its bed, and
prevent the accumulation of gravel. A
greatly overloaded
current will deposit too rapidly to admit of the concentration of
"
the gold dust * Under conditions intermediate between these
* Ross
E. Browne, Tenth Report of the Col. State Mineralogist, p. 448.
extreme states, the current may be just strong enough to keep its
bed clear from all accumulations except a small quantity of coarse
gravel and the coarse gold, which is caught in the natural riffles,
and thus all the conditions necessary to form a rich bed of pay-dirt
may be present. If, however, the bed consist of granite or other
rock which wears in smooth and rounded shapes, little gold will
be caught. Slates, consisting of layers of uneven hardness, wear
irregularly, and afford a good gold catching surface. The condi-
tions noted above as necessary to form rich gravels cannot be
expected to have been prevalent over great distances. "An
increase of grade or narrowing of the channel will cause an
increase of velocity, and the same stream may be underloaded
in a narrow steep section, and overloaded in a broad flat
section." * The difference of velocity between the middle and
sides of a stream, and between the inside and outside of a bend,
may give the right conditions in one part of a river bed and not
in another. Thus with high grades, rich gravels should occur
in the less rapid, and with low grades, in the more rapid parts
of a stream. Having regard to such considerations, the richer
parts of existing rivers can be pointed out with little trouble.
When, as in the Pliocene rivers, the beds are buried to a depth
of hundreds of feet, the richer parts are more difficult to find.
The history of the Pliocene rivers began with a period when ex-
cavation exceeded deposition, when the rivers were underloaded
for at least a portion of each year, and the channel was constantly
being deepened. Some bench or terrace gravels were formed at
this time, and being at the sides of an underloaded river tended
to be rich. The river bed, although rocky and comparatively free
from sand, would perhaps accumulate some coarse gold, which as
the channel deepened was no doubt in part ground up into fine
particles and carried off, but at the time when the excavation had
reached its lowest point, some of this coarse gold would certainly
be present. When the underloading of the stream ceased, what-
ever caused the cessation, a pause must in many cases have occurred
before the gravel proved too much for the stream to carry. During
this pause the conditions for gold catching were favourable, and
hence rich gravels were formed on bed-rock in the gutters or
channels. Then, as the streams became overloaded, sand and
gravel accumulated rapidly, so that little concentration of the
gold in them could take place. The rivers flowed over thick
sand banks and, in consequence, frequently changed their courses.
The sands, being deposited by overloaded rivers, of course con-
tained fewer and smaller boulders, and the thick masses of poor
sand thus went on accumulating until the volcanic outbursts
put an end to the process.
Origin of the Gold in the Placers. The origin of the
gold in deep placers has long been a vexed question. It was
*
Ross E. Browne, Tenth Report of the Col. State Mineralogist, p. 448.
formerly accepted without question that the erosion of auriferous

quartz lodes existing at higher altitudes furnished both gravel
and gold. In support of this it was urged that the same districts
which furnished auriferous gravels abounded in quartz veins at
higher levels, while Whitney pointed out that numerous lodes
were intersected by the valleys and were still to be seen in the
bed-rock. On the other hand, the fact that the nuggets found
in the drift are much larger than any masses of gold encoun-
tered in veins, and that the placer gold is of superior fineness,
are difficulties in the way of accepting this theory. Moreover,
Egleston states that nuggets as large as a man's fist have been
found embedded in tliQ midst of fine sand, whither they could
not have been carried by the action of running water, but authentic
instances of such finds seem to be lacking, nuggets usually occur-
ring in coarse gravel, among boulders. It is further declared by
the opponents of the erosion theory that if a small quantity of
soft material like gold, mixed with lumps of hard quartz, were
washed down by water, then, long before the quartz could be
reduced by grinding to the condition of grains of sand, the gold
would be worn down to such a fine state of division that none of
it could lodge in the river bed at all. In opposition to this con-
tention, it may be urged that the extreme malleability of fine
gold would make this comminution very slow, and scales of the
metal are said to have their edges blunted and thickened by the
pounding action of dry sand moved by the wind, instead of having
them worn away. However, it is certain that the form of placer
gold is from what might have been expected if it con-
different
sisted of water-worn vein gold. Nuggets are mammillary in
form, and generally appear more like concretions than water-
worn fragments, in spite of the fact that they often include
more or less quartz.
In 1864, in order to account, for these and other facts, Mr. A.
C. Selwyn, of Victoria, suggested a theory of solution in which it
is supposed that the gold disseminated through the rocks and
drifts is dissolved by percolating waters which contain acids and
salts in solution, and is reprecipitated around certain centres.
Selwyn considered that the waters capable of dissolving gold must
have acquired this property by passing through the beds of basalt,
<fec., overlying the drifts, inasmuch as large nuggets occur in
districts where basaltic eruptions have taken place, while, where
these are absent, the gold is very fine, and nuggets can scarcely
be said to exist. The fact has long been known that gold is
soluble in certain dilute solutions of salts, likely to be met with
in nature, such as a mixture of nitrates with chlorides, bromides
or iodides, or as the haloid ferric salts. This has been firmly
established by the researches of Skey, Daintree, Egleston and
others. Also, the precipitation of gold from these solutions around
nuclei consisting of particles of gold, pyrites, &c., by organic matter
present in the liquid, has been studied, and efforts made to form
nuggets similar to those found in nature, without much success.
This, however, is not surprising since the conditions in nature,
including almost unlimited time and immense quantities of ex-
ceedingly dilute solutions, cannot be reproduced in the laboratory.
Among other pieces of evidence against the erosion theory which
have been cited, may be mentioned the fact that some gold placers
occur at higher levels than any quartz veins yet discovered or
likely to be discovered; also that nuggets have been found
embedded in decomposed rocks in positions to which they could
not possibly have been carried by running water, so that these
nuggets at least must have been formed by accretion. Some
regard must also be paid to the prevalent belief among diggers
" "
that if a little seed gold is left in the tailings from sluicing
operations, the deposit will grow in richness so as to be worth
working over again after a few years.
Some of these arguments have been met by the exponents of
the erosion theory. The fineness of the placer gold has been
accounted for by supposing that the impurities (silver, copper,
&c.) formerly present in the native gold have been dissolved
away by meteoric water, in which they are much more soluble
than gold is. The existence of large masses of gold in placer
deposits was accounted for by Whitney by assuming that the
upper portions of the lodes, now washed away, were richer,
and contained larger masses of gold than the remains of the
lodes now left, but Liversedge has shown* that this assumption
is not necessary. Some nuggets too have been found showing
undoubted signs of erosion by water, but these are rare. Liver-
sedge has recently adduced evidence (loc. cit.) that, even if the
small particles of gold found in placers have grown by accretion,
nuggets cannot have appreciably increased in size. The sug-
gestions made to account for the great richness at bed-rock
viz., that gold has "settled" through the quicksands, or that the
gold solution has remained longest in contact with the sand
nearest bed-rock are not wholly satisfactory, and must be
supplemented by some such explanation as that given above,
p. 67.
Minerals occurring in the Placer Deposits. In Califor-
nia, if quartz grainsand silicified wood are excepted, the most
abundant mineral is black iron-sand, which usually consists of
magnetite, although menaccanite, a form of hematite in which part
of the iron is replaced by titanium, also occurs. These minerals
must have been derived from the lavas, as neither of them are
known to occur in the quartz veins of the country. Platinum and
present in more or less abundance; thus
its allies are usually iridos-
mine occurred to the extent of 1 in 100,000 of the gold in the
early days, and increased afterwards to 15 or 16 times that pro-
*
Proc. of the Royal Soc. of N.8. Wales, Sept., 1893.
portion. Grains of native copper, nickel, and perhaps lead have

also been detected, and a few diamonds occur, while
garnets,
small crystals of zircon and cinnabar are very abundant.
METHODS OF TREATING DEEP PLACER GRAVELS.
Both in California and Australia, when the first gold dis-

coveries were made, the river beds and bars were at once
explored, and soon afterwards the flats closely adjoining them.
Subsequently, the bench gravels situated in the same valleys,
and the side ravines and gulches, which remained dry during
most of the year, were prospected and worked, owing to the
rapid growth of the mining population and to the fact that
the exhaustion of the shallow placers was already beginning
to make itself felt. The result was that the exposed edges
of the outcrop of some of the deep leads were found, and
the pay-dirt followed into the hill-side by drifting. Then, as in
many cases it was found that the gravels overlying the pay-dirt
on hill-sides, although poor by comparison with the earth below
them, nevertheless contained a small quantity of gold, the idea
was evolved of breaking down the whole bank by jets of water,
and passing all the material through the sluices. Hence arose
the practices of drift mining and hydraulicking,* of which the
former is largely used in California, while the latter, though
much used in California prior to 1884 and in New Zealand,
cannot be applied in Australia or Siberia owing to the general
flatness of the country. In Siberia, only shallow placer deposits
are worked. In Australia, the deep leads are usually reached by
shafts, since the surface of the country is not intersected by deep
canons, as in California.
Hydraulicking. This method of working consists, as has
been already stated, in breaking down banks of gravel by the
impact of powerful jets of water, and passing the disintegrated
material through a line of sluices, without the agency of hand
labour. The chief requisites for the successful application of
hydraulicking are
1. Large quantities of auriferous gravel, not less than 30 feet
in thickness, and not overlaid by any appreciable thickness
of barren material, which would necessarily be passed through
the sluices with the pay-dirt. The gravel treated need not be
rich, a mean yield of less than 1 grain of gold per cubic yard
being often enough to furnish profits if the operations are 011 a
sufficiently large scale.
2. A plentiful and uninterrupted supply of water throughout
considerable portions of the year.
* Edward
The invention of hydraulicking is ascribed to Mattison, of
Sterling, Connecticut, who used the method in 1851, on a small scale.
3. Sufficient fall in the ground so that

(a) the water may be
delivered under the pressure of a head of from 100 to 300 feet,
(b) the tailings can be easily carried away to a large dumping
ground, which is most conveniently either the sea, or a large
and rapid river.
Commencement of Operations. In California, the naturally
occurring banks or cliffs in the gravels in the sides of the gulches
were first selected for attack. Later, some of the deposits occur-
ring in those channels which are not intersected at favourable
points by the present system of drainage were operated on. It is
necessary in such cases to run a tunnel from the nearest canon in
the bed-rock to the lowest point in the gravel, this point being
found or guessed at by prospecting operations. The tunnels are
often of great length, that at the North Bloomfield mine, Nevada
Co., Cal., for example, being 7,874 feet or 1| miles long. One or
more shafts are then sunk from the surface through the gravel
to the tunnel, and washing operations are begun by ground-
sluicing, letting the water and gravel fall down the shaft and
run through the tunnel, in which the sluices are sometimes laid.
The surface near the shaft is thus gradually lowered, or it may
be terraced by hand labour, until an excavation is made of
sufficient size to enable the ground to be attacked with the hose.
Washing then proceeds regularly in this manner, the bank being
broken down by jets of water, and the products being allowed to
fall down the shaft and pass through the tunnel. Care must,
of course, be taken in the initial stages of the work to prevent
the blocking of the shaft, either by the caving in of its walls,
which are held apart by heavy timbering, or by runs of ground,
which may occur when the upper terrace is undermined by the
jets. After work has proceeded for some time, however, the
gravel immediately round the shaft is all washed away down to
bed-rock, and the shaft is then almost or entirely obliterated
(according to the location of the tunnel), giving place to a huge
open cut or funnel-shaped excavation, at the bottom of which
is the upper end of the tunnel. Having thus briefly indicated
the usual sequence of events in opening a hydraulic mine, the
plant and the ordinary method of working may be considered
under four heads.
1. The supply of water.
2. Breaking down the banks.
3. The washing proper, or sluicing.
4. Disposal of the tailings.
1. The Supply of Water. The amount of water required by

large undertakings is more than can be obtained from the
far
rainfall on the hills immediately round the mine. Thus the
North Bloomfield Mine, in the season of 1877-8, used between
for the
sixty and seventy millions of gallons per day, or enough
total supply of a city of three million inhabitants. The workings,
moreover, must necessarily be considerably above the level of

any large rivers in the neighbourhood, so that it is often neces-
sary to build huge reservoirs at convenient spots to store up the
rainfall and melted snows of large districts, and to convey the
water thence to the mine by ditches, flumes, or pipes. In Cali-
fornia the reservoirs have an aggregate capacity of 8,000,000,000
cubic feet, the largest, that of the South Yuba Water Company,
having a capacity of 1,800,000,000 cubic feet.
The ditches conveying the water from these reservoirs to the
mines pursue a course determined by the contour of the country.
They are usually cut in the sides of hills, and are given as far
as possible a uniform grade, which in different ditches varies
from 5 feet to 40 feet per mile. With the higher grades more
care in the lining of the ditch is necessary to prevent erosion of
the banks. The ditch may be lined by stones or boards, or left
unlined, but the loss from leakage amounts to several per cent,
of the delivery of unlined ditches. The ditch is sometimes
buried in the ground to prevent damage by avalanches, to keep
the water from freezing, and to reduce the loss by evaporation,
which, in the case of one ditch in California, was found to be
about 1 2 per cent, of the amount delivered from the reservoir.
Wooden flumes are also used instead of ditches, but though
cheaper, are more liable to be damaged. The dimensions of the
ditches vary enormously ; they range up to 100 miles in length in
California, and are from 2 to 15 feet wide, and from 1 to 5 feet deep.
In New Zealand ditches have been used less than wooden flumes,
but latterly, the use of sheet-iron pipes has been greatly extended
there, and many flumes have already been replaced by them.
These pipes are made as much as 30 inches in diameter, and are
cheap, durable and easily repaired they entirely prevent losses by
;
leakage and evaporation, and deliver more water under similar

circumstances than flumes, as they offer less frictional resist-
ance to the flow.
Where it is necessary to cross side cailons or gulches the
flume or iron pipe is carried over on a light tressle bridge, or
the water is passed through an inverted siphon formed by a
wrought-iron pipe which passes down one side of the valley and
up the other, being filled from a head-box or reservoir, and
delivering the water at a lower level on the other side. At
Cherokee, Butte County, an inverted siphon pipe is used to
carry the water across a ravine 873 feet deep. The diameter of
the pipe is from 30 to 34 inches, and its greatest thickness (where
there is a pressure of 384 Ibs. per square inch) is 0-375 inch. *
The Miocene Ditch Company, operating in the same County,
carried their flume, of 4 feet wide and 3 feet deep, round the face
of a bluff 350 feet high, supported on L-shaped, iron brackets
made of bent T rails, soldered into holes previously drilled by
*
Ninth Report, Oal. State Min. (1889), p. 125.
men let down the face of the cliff by ropes. Successful ditch
construction often depends on such engineering feats.*
The water is delivered at a convenient height above the work-
ings into a head-box consisting of a small wooden reservoir from
which the pipes take their origin. In many cases the reservoirs
and ditches are owned by separate companies, or, as in New
Zealand, by Government, and the water is sold to the miners by
measure. The unit in New Zealand is a " government head,"
and in the United States a "miner's inch," following the system
in vogue in Spain and Italy. The amount of water that will
flow through an orifice 1 inch square, cut in a board of 1 inch
thick, under a head of water that varies with the custom of the
locality but is usually from 4 to 8 inches, is called a miner's
inch. The amount of flow in twenty-four hours is called a
"
twenty-four hour inch," and similarly there are ten-hour and
twelve- hour inches. The quantity of water in a miner's inch
varies with the head of water used and the form and size of the
orifice for delivery. Thus the amount delivered from an orifice
25 inches long and 2 inches wide is reckoned as 50 inches,
although it will be more than fifty times as much as the delivery
from an orifice 1 inch square. The twenty-four hour inch under
a head of 7 inches amounts to about 2,230 cubic feet.f
The water is conveyed from the head-box by pipes, which were
formerly made of canvas hose, to which iron rings 3 inches apart
were added for pressures of over 100-feet head. These latter
are called " crinoline hose," and were generally made of from
6 to 8 inches in diameter. They were used for some time, but
were found to suffer from rapid rotting, and to be liable to burst
at pressures of above a 200-feet head, and are now completely
replaced in the "United States by sheet iron. In New Zealand
the canvas hose still lingers, but is now being rapidly replaced.
The iron feed-pipes are made from 10 to 15 inches in diameter,
and the thickness varies according to the pressure which they
may be called on to withstand. Sharp bends in them are avoided
as the flow of water is checked thereby. They are liable to
collapse if the level ot the water in them is reduced, and a partial
vacuum formed inside ; hence, as in the case of all other sheet-
iron pipes used in hydraulicking, they are fitted with valves,
which are constructed so as to freely admit air from without.
The best and cheapest form of these is that used in New Zealand,
which has a 2-inch hole and a rubber clack like that used in
pumps. The valve is always open until the rising water lifts it
up on its surface, and closes the orifice. The water is discharged
*For details of the coat of construction, &c., of ditches, the student is
referred to Egleston's Mining and Metallurgy of Silver, Gold and Mercury
in the United States, vol. ii., pp. 120-180.
t For further details concerning miner's inch vide Art. by P. M. Randall
in Precious Metals in the United States, 1884, pp. 558-572.
through a nozzle called a "giant" or "monitor." The nozzle

was at first a sheet-iron tube, having an aperture 1 inch in
diameter, and was held in the hand. The size of the nozzle was
gradually increased, until it has now reached a diameter of
11 inches. Such a stream, under a head of 200 feet, requires
special appliances to control it, deflect it at will, and prevent the
nozzle from " bucking." A number of ingenious contrivances
have been devised both in America and New Zealand, but all the
monitors bear a general resemblance to the one shown in Fig. 10.
2. Breaking Down tlie Bank, When the jet is first directed
against the bank, the water spatters in all directions, then"buries
" cave takes
itself a little, and after a time, in loose ground, a
place, the undermined bank falling down. By the method of
undermining, the power of the giant is much increased, especially
where hard and soft layers alternate. When large caves are
about to take place the water is turned off, as otherwise the
Fig. 10.
= 4 ft.
Scale, 1 in.
ground may run so far as to overwhelm the monitor and the

workman directing it. The nozzle is placed as near to the bank
as possible, consistent with the safety of the workers, so as not
to waste too much of the initial velocity of the stream of water.
Consequently, lofty banks are not advantageous, and if they
exceed 200 feet they are usually worked in terraces of 100 feeb
or so in height. In some parts of the Spring Valley Mine, how-
ever (see Fig. 10a), a bank of 450 feet high was worked in a
single bench, and it was then not unusual for the runs of ground
to bury pipes which were throwing 7-inch streams fronx a dis-
tance of 400 feet from the face of the bank.
The jet is, if possible, delivered unbroken against the face of
the bank, as its disintegrating power is thus kept at its maximum.
However, in some cases, where it is cemented, the gravel is too
hard to be economically broken down by the water alone, and
blasting is then resorted to, a drift being run into the bank, and
cross-cuts made at the end in which the powder is placed ; the
drift is then and the charge exploded by electricity.

filled up,
It is more economical blow out the base of the bank, as the
to
upper part then falls by its own weight and can be broken up
by the water. Sometimes arrangements are made to explode
very large blasts thus, at the Blue Point Gravel Mine, a
:
charge of 50,000 Ibs. of powder was exploded at the end of a

drift 275 feet long in the year 1870, and 150,000 cubic yards of
gravel were brought down, while at another mine, 3,500 Ibs. of
dynamite were exploded in 1872, and 200,000 cubic yards of
gravel disintegrated at once.
3. Washing the Gravel in the Sluices. The sluices in which the
gold is caught are constructed on exactly the same principles
as those already described, but are larger and, though usually
made of wood, are of more massive construction, in accordance
with the great quantities of gravel to be handled and the con-
tinuous nature of the work. The sluices are commonly called
"
flumes," but it is better to restrict the use of this word to a
conduit for carrying water only. The sluice boxes used in
hydraulicking, though, as usual, only 12 feet long, are as much
as from 3 to 6 feet wide and from 2 to 3 feet deep ; they are
lined with heavy planks on the sides, and the pavements are
made of more durable materials than is usual in shallow placer
sluicing, wooden blocks, rocks, or T
railroad iron being most
usually employed. The wooden blocks are from 12 to 30 inches
square, and from 8 to 18 inches deep. They are usually
made of one of the softer varieties of pine (e.g., the " digger "
"
pine, Pinus sabiniana, and others), which " brooms up under
friction, and thus presents a better catching surface. The blocks
are cut across the grain of the wood, and are set side by side
across the sluice, each row separated from the next by strips of
wood to which they are nailed, while they are also kept in posi-
tion by the side lining which is placed upon them. The inter-
stices in the block pavement act as gold catchers, and are filled
with small stones, or, with less advantage, allowed to fill up
with gravel when washing begins. On account of the rapid
wearing away of the wood, much of the gold and amalgam caught
is scooped out and carried off again. Wooden block rimes only
last from two to four weeks when in heavy work, but are easy to
take up and put down again in cleaning up they are discarded
;
when worn so as to be only 4 to 6 inches thick.

Rock pavements are made of those boulders which are most
easily obtained in the particular district. Basalt is generally
used, oval stones of 15 or 18 inches long and from 9 to 12 inches
thick being selected and placed on end, with a slight slant in
the direction in which the current flows. They are held in place
by wooden planks which divide the sluice into compartments, so
that if one stone works loose the pavement as a whole is not
affected. The interstices as before are filled with gravel. Rock
pavements are very durable, lasting from three to six months,

but require more grade to the sluice, and occasion loss of time
in cleaning-up and re-paving the sluices. Consequently, they
are never used near the head of a sluice, where cleaning-up is a
frequent operation, but are ofton used for the lower parts
of sluices, where they sometimes alternate with block riffles, and
are especially suited for tail-sluices which are only cleaned-up
once a year. Rock pavements cost less than other forms of
riffles.
Iron riffles, which usually consist of T-iron rails, are placed
longitudinally in the sluice, closely packed side by side. They
present a large amount of space available for catching the gold
and amalgam, last well, present little resistance to the current (so
that the grade may be low while the duty of the water remains
high), and are easily taken up and put down. They are, therefore,
generally used at the head of the sluices. Though their first
cost is higher than that of wooden blocks, they are more
economical in the end, owing to the saving of time in cleaning-
*
up and to their longer life. Egleston instances the results of
experiments made at the Morning Star Claim, California, where
three sections of sluice, each 65 feet long, were laid at a distance
of 300 feet from the face of the bank which was being worked.
The first section was as usual laid with wooden blocks, the
second with old rails, and the third with rocks. When the
clean-up was made the middle section gave 9 ozs. more of gold
bullion than both the others combined. If old rails cannot be had,
strips of wood bound with iron are used, but are less durable
and satisfactory.
At the Blue Spur Consolidated Gold Company's plant, Gabriel's
Gully, New Zealand,! where the sluice is necessarily very short,
most of the stones are first separated from the gravel, and the
finer material is then passed over a sluice paved with trans-
11.
verse angle-iron placed with the hollow side facing down

riffles,
stream (Fig. which the arrow shows the direction of the
11, in
stream); these iron riffles are placed 2 or 3 inches apart. Below
the section containing these riffles, there is a false bottom to the
sluice, formed by an iron plate perforated with small round holes,,
*
Gold, Silver and Mercury in the United States, vol. ii., p. 218.
t Report of the Mining Commissioner of New Zealand, 1891.
through which some of the water and the finest particles of the
gravel fall on to cocoa-nut fibre matting, laid on the true bottom
of the sluice. Here the fine gold is caught, the principle being
similar to that used in undercurrents.
The sluice is often divided into two by a median longitudinal
partition, so that one side may be at work while the other is
being cleaned-up or repaired, both sides being sometimes worked
when water is very plentiful. There are usually unpaved rock-
cuts above the sluice, leading to it from the places undergoing
the process of piping. These rock-cuts are rarely supplied with
mercury, and very little gold is usually caught there. The sluice
may be placed above the tunnel, or in the tunnel itself (one way
of preventing unlawful cleaning-up), or below it. In the case of
the North Bloomfield Mine, the irregular slates, dipping at a high
angle, forming the floor of the tunnel, were used as natural riffles
for the lower part of the sluice, and thus all cost of wooden
frames, pavements, &c., was saved, but the floor of the tunnel was
lowered by 3 feet, and deep holes worn in it, after 22,000,000
cubic yards of gravel had passed through it.
The length of the sluice, if capital is not lacking, depends on
the cost of construction and of the maintenance, as compared
with the value of the gold saved owing to the increased length of
the system. The length may be diminished by a plentiful use of
drops, grizzlies and undercurrents, all of which are described above
under the head of shallow placer sluicing they are made of pro-
;
portionately large size in hydraulicking. Coarse gold is of course

soon caught, but fine gold may successfully evade all the riffles of a
long sluice. The Spring Valley Mine has three parallel lines of
sluices, each 2 miles in length, and it is estimated that 95 per cent,
of the gold contents of the gravel is caught.* This length is un-
usual, the average not exceeding about 1,000 feet. The cost of
construction is from $25 to $35 per box (of 12 feet long) at the
larger Californian mines, and less in New Zealand. Here, by a
number of contrivances, especially by eliminating all but the
smallest stones by means of grizzlies or trommels, the length of
the sluices has been greatly diminished, but American engineers
regard the retention of the stones as desirable, owing to their
action in assisting disintegration.
Grade of the Sluices. The grade depends on the available fall
of the ground and on the character of the material to be washed.
The minimum is from 2 to 4 inches per box, such low grades
being sometimes enforced by the nature of the ground, some-
times adopted from choice if the gravel is light, the gold fine,
and water plentiful. With these low grades, however, disin-
tegration is slow and incomplete ; stones, unless they are small,
cannot be sluiced ; large ones block the sluices and must be re-
moved by hand, and the " duty " of the water, as regards sand, is
* Ninth
fieport Gal. State Min. (1889), p. 129.
greatly decreased. The 6-inch grade is that most generally used,

but as much as 12 inches per box, where it can be obtained,
is regarded as desirable in good American practice.
Higher
grades than this are unusual. Steep grades effect disintegration
rapidly, thus shortening the length of the sluice, and enable
all but the largest rocks to be sluiced, but less gold is then
caught and a more plentiful use of undercurrents is necessary.

In New Zealand an idea is gaining ground that sluices have been
made too narrow, and that if the width is greatly increased and
the grades diminished, and the depth of the current thus reduced
to a minimum, the losses of gold will be reduced. Under such
conditions all rocks must be removed by grizzlies, and in any case
the cemented gravels, so common in the Californian blue leads,
could not be treated in this way. It is considered necessary to
have a sufficient depth of water to cover the largest boulders to
be sluiced, but it undoubtedly diminishes the amount of gold
saved if the water is more than two inches deep. Nevertheless,
a, depth of water of from 6 to 9 inches is not unusual. Where
poor or top gravel is being "piped," it is worked off as rapidly
as possible, and with less regard to the percentage of gold saved
than when rich stuff is treated, but with more regard to the
number of cubic yards treated.
Tlie Use of Mercury. Mercury is added in great quantities
several times a day in America on the principle that " the more
quicksilver added the greater are the chances of catching the
gold," but less frequently and in less quantities in New Zealand.
In some Californian sluices from 2 to 4 tons of mercury are in use
at once. The feeding is regulated by the appearance of the
amalgam in the sluice, the additions being made near the head-
box and in the undercurrents. The loss of mercury is usually
from 10 to 15 per cent, of the amount used per run. When
cemented gravels are being treated, owing to the extra amount of
trituration required, the loss may be as high as 30 per cent.
These losses are the more serious, for the reason that amalgam
is more easily lost than pure mercury, so that a heavy loss of
mercury denotes a heavy loss of gold.

Cleaning- Up. The process does not differ from that described
under the heading of shallow placer mining. It is advisable not
to defer the clean-up too long as losses of amalgam are caused by
the wearing of the riffles. Usually from 50 to 95 per cent, of
the total yield of amalgam is caught in the first twenty or thirty
boxes, which are cleaned-up frequently. The following table*
shows the percentage yield of the various sections of the sluices,
Ac., at the North Bloomfield mine, California, for the year
1877-8:
Total yield,
*
.... $311,276.20.
Ninth Report Gal. State Min., p. 131.
Near bank, from rock cuts in mine (all in gold dust, no

4 '57 per cent.
quicksilver being added in the rock cuts), . .
Sluice in tunnel (1,800 feet), 86'26

Tunnel below sluice (6,000 feet), see p. 78, . . 4 '50
Cut below tunnel (200 feet), 0'81
Tail sluices (300 feet), 1'21
From seven undercurrents, 2*65
100-00
The first undercurrent caught five times as much as the sixth,

and nearly three times as much as the seventh, which was of
double size. The yield of the seventh ($947) induced the Com-
pany to add another undercurrent. This mine affords an example
of the difficulty of catching fine gold. The gold loss was un-
known, but was believed not to exceed 5 per cent, of the contents
of the gravel.
The bullion obtained by retorting the amalgam from the
sluices is finer than that from quartz mills, and is sometimes
990 fine in Australia, although Californian placer gold is often
as low as 850fine. The remainder is mainly silver, but copper,
lead, iron, and some of the minerals existing in the gravel also
occur. The amalgam from the head of the sluices yields finer
gold than that caught lower down and in the undercurrents.
Tailings. The disposal of the tailings is one of the most
important points to be considered in hydraulicking. Where
there is not sufficient fall to enable the tailings to be removed
from the lower end of the sluice without pumping, hydraulicking
is impossible. The tail-sluices usually terminate on the side of
a canon, in a river, or in the sea. The enormous amount of loose
sand and gravel, delivered from the hydraulic mines of Placer
County, California, and the neighbouring counties into the Yuba
and Feather rivers prior to 1880, filled up their beds to such an
extent that in rainy weather disastrous floods ensued, and much
valuable agricultural land was buried beneath sterile drift de-
posits and rendered worthless. The farmers thereupon took action
against the Mining Companies and obtained a perpetual injunc-
tion forbidding them to discharge their tailings into these rivers.
The result has been to stop hydraulicking in these districts, and
the efforts to work the deep leads more extensively
by drifting,
or on the other hand, to impound the tailings by dams made of
brushwood, or to return them to their original position, have not
resulted in unqualified success. Consequently the gold winning
industry has not been maintained on the extensive scale it had
assumed prior to the action of the courts.
The Yuba and Feather rivers, in which it was estimated that
from 750,000,000 to 800,000,000 cubic yards of gravel were con-
tained, have continued during the last ten years to carry this
material lower down and to distribute it over the level ground of
the plains. The result has been that
ground -sluicing on a
tremendous scale has been carried on, and the tailings in the
upper parts, enriched by the removal of the valueless sands, will

probably pay to work over again in the immediate future. The
impounding of tailings behind brushwood dams is now believed
to afford a solution of the difficulty in California, and hydraulicking
will probably be recommenced on an extensive scale before long.
Drift and Shaft Mining. The problem of reaching the rich
leads lying in the gutters of the old river channels, without re-
moving the superincumbent masses of poor material, was attacked
in California before hydraulicking had been invented. The
method of following rich beds which passed into the hill-side by
means of timbered tunnels was practised before the upper beds
were suspected of containing any gold. The rise and rapid
progress of hydraulicking probably checked the development of
this tunnel or drift method between the years 1855 and 1880,
but it is now the most important branch of the placer mining
industry in California, while, as already stated, the analogous
method of shaft mining has always been almost exclusively in
use in Australia. Although hydraulicking was used wherever
practicable after its introduction, many deposits occur where it
cannot be employed. Where the upper gravels are thick and
almost sterile, and in particular where the pay-gravel is covered
by a great depth of lava (as for example at Table Mountain,
California), drift mining must be resorted to. Want of sufficient
grade, or of dumping ground, or of water supply, may also render
hydraulicking impossible.
At first work was done in California only on those leads which
were intersected by one of the existing canons, so that the drift
could be carried in paying ground from the start. In later
times, when the country had been more carefully surveyed, and
the probable courses of the Pliocene rivers roughly indicated,
efforts were made to reach leads at points far from any outcrop.
In order to do this, a tunnel is driven in the bed-rock from
the nearest canon to a point just below the gravel which it
is proposed to mine (Fig. 9, p. Formerly tunnels were
64).
often driven when the proofs of the position, depth and value of
the channel rented on evidence which would now be deemed
insufficient. Consequently they sometimes completely failed in
their object. Instances are on record of tunnels being driven
hundreds of yards in directions in which there was absolutely
no chance of encountering a river channel. The tunnel was
frequently put in too high, so that it was necessary to sink a
shaft at the inner end of the drift in order to reach the pay-
gravel, and thus the extra expense of raising water and gravel
vertically for some distance was incurred. During the past few
years, the location of the ancient channels, even at distances of
several miles from the nearest outcrop, has been determined
with some precision by engineers who have made a special study
of the subject.
If the bed-rock tunnel is successfully placed, so that, while

there is a fair grade outwards throughout its length, its inner end
is a few feet below the lowest point of the channel, the succeed-
" "
ing operations are much cheapened and simplified, as upraises
to the gravel then serve to drain the workings and to deliver
the pay-dirt into cars, which convey it to the tunnel mouth. The
methods of cutting out the gravel bear a strong resemblance to
those used in coal mining, if the channel is wide and of fairly
uniform value, but the gravel is of course frequently tested by
panning. A detailed description of these methods of mining the
gravel, however, may with more propriety
be given in treatises
on mining, than in this volume.*
The pay-gravel carried to the tunnel mouth in cars, which
is
are propelled by hand, by horses, or by steam power, according
to the magnitude of the work. At Bald Mountain, California, t
the gravel is brought through a tunnel 7,000 feet long in a
train of eighteen cars, each holding 2 tons, by a locomotive
which performs the trip in five minutes. The cost of transpor-
tation at this mine is stated to be, by man-power 21 cents per
carload (of 2 tons), by mule-power 9 cents per carload, by steam
4f cents per carload. Outside the tunnel the gravel is dumped
into bins, whence it is delivered, by gravity if possible, either to
the sluices or to the batteries. In many cases the gravel is
cemented into a conglomerate which is too tough to be easily
disintegrated in the sluices. It is then passed through " cement
mills," which closely resemble the stamp battery to be described
in the next chapter, the chief differences to be noted being
in the facilities for delivery. Double discharge mortars are
used, and the screens are very coarse, the mesh being usually
about ^-inch, but varying up to j-iiich in diameter. One
battery of ten stamps, each weighing 950 Ibs., making 94 drops
of 9 inches in height per minute, will crush about 40 or 50
tons of gravel in ten hours so that it will pass through a T3F -inch
mesh screen. Mercury is put into the mortar, and most of
the gold is usually caught there on amalgamated copper plates,
but copper plates outside the mortar are also used as in quartz-
milling, and rubbers are employed to brighten the gold. If
well-arranged plates are laid down, the number of sluice boxes
which can be added with advantage is very small, a length of
from 50 to 300 feet being used, the former limit being most
common. No attempt is made to save the auriferous magnetic
sands and sulphides which these conglomerates usually contain.
In cases where cement mills are not required, the gravel is
washed in sluices which differ little from those already described.
The boxes are not more than from 18 inches to 24 inches wide
*
A detailed description is givenby Russell L. Dunn, in Eiqhth Report of
Cal Mate Min., 1888, p. 736.
t Ninth Report of Cal. State Min., 1889, p. 118.
*and deep, and the series is seldom more than 300 or 400 feet
long Iron riffles are most in favour. Where the amount of
gravel to be washed is small, or the water is scarce, the gravel is
allowed to accumulate for some time and the water stored in a
tank or reservoir. It is in some cases a great advantage to keep
compacted gravels exposed to the air during a few months before
" "
washing them, as they slack and disintegrate under the in-
fluence of the weather, and subsequently are more easily treated,
while for a similar reason, tailings are sometimes impounded,
and re- washed after some time has elapsed. The disintegration
"
of cemented material, which has been " slacked by exposure to
the weather, is usually completed in a cement-pan. This is a
cast-iron pan with perforated bottom, and with a gate in the side
for the removal of boulders, which are mostly barren and are
separated from the auriferous material by this system, instead of
being crushed and mixed with it, as is the case when stamp-mills
are used. In the pan, four revolving arms, furnished with
plough-shares, break up the gravel, which is carried through the
apertures in the bottom by a stream of water, and falls into the
sluice. A pan of 5 feet in diameter and 2 feet in depth will treat
from 40 to 120 tons per day, according to the nature of the
gravel. (See also p. 60.)
The Hydraulic In this machine a jet of water
Elevator.
under high pressure forces water, gravel, and boulders up an
inclined plane, and delivers them all at the head of the sluice,
which may be as much as 100 feet above bed-rock. The differences
in construction between the machines made in Australia, New
Zealand, and the United States are only matters of detail.
They consist essentially of an upraise pipe, usually of wrought
iron, having a diameter of from 12 to 24 inches, which terminates
below in an open conical funnel ; a hydraulic nozzle, delivering
water under the pressure given by a head of from 100 to 500
feet, projects into this funnel, and sand and gravel can also enter
round the sides or through a special orifice. The inclination of
the upraise pipe is usually from 45 to 65. The top of the
upraise pipe is turned over and terminates above a sluice, into
which the gravel falls and is washed in the ordinary way. The
subjoined figure shows the arrangements at the base of the
upraise pipe of the elevator manufactured by Mr. J. Henry of
San Francisco. In Fig. 12 Nos. 1 to 13 are castings, No. 14
consisting of wrought iron ; a ball joint is formed by
Nos. 3, 4,
and 5, enabling the pipe bringing the water to be moved. The
nozzle and the lower part of the upraise pump are sunk in a
sump excavated in the bed-rock, and the gravel is washed down
by any means (usually by a jet from an ordinary hydraulic
nozzle) into this sump. The entrance to the upraise pipe is
protected by a coarse grating which prevents large stones, pieces
of wood, (fee., from entering it. The force of water is enough to
complete the disintegration of the gravel during its passage

through the upraise pipe, so that a sôrt sluice is enough to
effect the washing proper. If the excavation is carefully arranged,
it may be kept funnel-shaped, so that the elevator, once placed in
a sump, may be worked there permanently without being moved.
When the pit is large enough, the washing may be done inside
it, only the tailings being raised
to the surface by the hydraulic
elevator. This was done at Oroville in 1880,* the flume to
deliver the tailings into the river being about 500 feet long.
The head of water required varies according to the vertical
height through which the gravel must be raised a head of
;
about 70 feet is required for every 10 feet of vertical upraise.
Since it is just as expensive to raise water as gravel, arrange-

ments must be made to deliver as much gravel into the sump as
can possibly be raised by the jet, otherwise the expense per
cubic yard will be greater, and there will be too much water
with the gravel for satisfactory treatment in the sluices.
Wherever the necessary head of water is available, the
hydraulic elevator is now recognised as the best method of
working flat placers, river-bars, &c., or any deposits which are
either below the water level of the district, or which have not
sufficient fall for the disposal of the tailings by gravity. It is
in wide use in California and New Zealand. The following
instances of work in both countries may be given: At the
Blue Spur Consolidated Gold Mining Company, Gabriel's Gully,
New Zealand,! tailings, which have accumulated close to the
sea on the foreshore, are sluiced in this manner. The vertical
upraise is 60 feet, the angle of inclination of the upraise pipe
*
Prod. Gold and Silver in the U.S.A., 1880, p. 15.
{New Zealand Mining Commissioners' Report, 1891, p. (55.
being 6 3 -5 ;
about 480 tons of gravel are raised per shift, the
head of water used being 400 feet, while the amount of water used
in each elevator is seventeen government heads. The sluice is
short, and has an inclination of only 3| inches in 12 feet; the
upper parts are fitted with transverse, patent [""- shaped, angle
iron riffles, in which the angle faces up stream (see Fig. 11). The
lower parts of the sluice have a false bottom of wrought-iron
plates, perforated with round
holes beneath these plates is the
;
true bottom of the sluice, covered with cocoanut matting in

which fine gold is caught. The tailings are discharged into the sea.
At Quartz Valley, Siskiyou County, California,* on hard
ground, where the elevator was first used, it took forty-three
days to work out a piece of ground 300 feet by 250 feet, which
was of an average depth of 18 feet. The bank was washed down
with 600 miner's inches of water, and went to the elevator
through a 30-inch bed-rock flume, which had a grade of 5 inches
in 12 feet. The water and gravel were raised through a 20-inch
elevator pipe without any contraction at the throat. It was
set at an angle of 40, and the pipe was 42 feet long, the vertical
upraise being thus 28 feet. The force used was 1,000 inches of
water with a head of 230 feet, delivered through a 6-inch nozzle,
and the gravel was emptied into a sluice 6 feet by 3 feet, with
a grade of 1J inches in 12 feet. When 3,500 inches of water
were running in this sluice, they could not carry off all the
gravel raised by the elevator. The work was done without any
delay from stoppage of the machine, and there were no repairs,
the wear of the elevator being very little.
Economic Conditions in Placer Working Sliallow Placer
Deposits. In work by individuals the results differ greatly,
both according to the strength and skill of the worker and to
the contents of the gravel. Under the best conditions of climate,
a strong, well-nourished, American digger may be able to raise
by the shovel from 10 to 12 cubic yards of gravel per day and
throw it into a receptacle 3 feet above the ground. Native
labour cannot be expected to effect so much, and in French
Guiana it is reckoned that only about half a cubic yard of earth
per man per day can be shovelled into the sluice. If the work-
man must wash the gravel, as well as raise it, much less can be
accomplished. For an active man, it is a fair day's work to
dig and wash from fifteen to twenty pans of dirt, the amount
treated thus not exceeding about 10 cubic feet. On the other
hand, with the cradle, the output may be from 1J to 2 cubic
yards per man in a day, while with the long-torn it may rise to
3 or 4 cubic yards per man. In the Siberian trough, only from
1 to 1J cubic yards can be treated by one worker per day, but
a large percentage of the gold is believed to be saved in this
*
Egleston's Silver, Gold and Mercury in the United States, 1890, vol. ii.,
p. 307.
apparatus. The minimum contents in gold, which will make

the gravel worth treating, depends, of course, on the cost of
labour in the country in which the deposit exists, since few men
will continue to work for less reward than they could obtain in
other employments. The wages of placer miners in various
countries is as follows : In California, from $2 to $4 per day ;
in Colorado, Montana, Dakota, and the neighbouring States, from
$2 to $3 1 per day ; in Australia, from 6s. to 10s. per day ; in
French Guiana, the negroes and coolies are paid about 3s. 4d.
per day, exclusive of rations. The wages in Siberia are given
below. In California, the Chinese are content to earn about
75 cents or 3s. per day. In early times, in California and
Australia, when the virgin shallow deposits were being worked,
large sums were often realised by individual diggers, cases being
on record in which 5 ozs. of gold were obtained from one pan of
bedrock- scrapings lying under heavy gravel, and earnings of
several hundred dollars per day were not uncommon. The
rich gravels, from which these results were obtained, are now all
worked out, and instances are rare in which more than $1 or $2
per day can be earned with the pan.
Concerted work, with the aid of the sluice, is much more
effective ; in California gravel containing about 1 pennyweight
of gold per cubic yard is worked at a profit, the dirt being lifted
into the sluice by hand-labour, and the tailings removed by
sluicing with water ; at Ballarat, in Australia, where the gravel
is raised to the surface from underground workings through a
vertical shaft several hundred feet deep, and subsequently
washed, 12 grains of gold per cubic yard of material pay for the
treatment, while in Siberia, as stated below, the cost is even less.
In French Guiana, the unhealthiness of the climate and the cost
of supplies render it impossible to work gravel containing less
than about 3 pennyweights of gold per cubic yard.*
In Siberia,! the distance of the workings from the nearest town,
and the traditions of the industry, require workmen to be hired
by the year, or, in cases where no work is attempted in winter,
for the season. The best workmen, employed in actual washing,
receive 1 rouble (about 2s.) per day; the diggers and carters
75 copecks (Is. 6d.), and the women and boys from 40 to 50
copecks (9d. to Is.) per day rations are also distributed in mines
:
far from towns. Sometimes piece-work is done, the price paid

being from 2J to 2| roubles per cubic sagen, or from 4d. to 5^d.
per cubic yard (!); the horses and men are then both supplied by
the workman. The total cost of treatment of the gravel varies
greatly with its geographical position. On the banks of the
Lena, where the season only lasts for five months, it is estimated
that the gravel must contain 2 zollatniks of gold per 100 ponds,
or 4J dwts. per cubic yard ; but in the neighbourhood of Ekater-
*
Fr&ny, Ency. Chim., vol. v., L'or, p. 48. t Ibid, p. 47.
ineberg, the deposits in the bed of the Pechma, a tributary of the

Obi, are worked when they only contain from 8 to 9 grains per
cubic yard. This is less than that required in other countries
in order that a profit may be made from similar deposits, viz.,
those in which the gravel must be handled at least twice.
Deep Placer Deposits. The cost of hydraulicking and drift
mining necessarily varies enormously with the conditions. In
hydraulicking, it depends
largely on the magnitude of the opera-
tions. Withlarge quantities of water available at a cheap rate,
big banks of soft gravel, and a well-constructed and convenient
tailings -tunnel, the cost has been reduced to from 2 to 3
cents per cubic yard in California, while the average cost there
in 1884 was about 10 cents, and only in exceptional cases
amounted to as much as 50 cents per cubic yard. This is
without reckoning interest on capital expended. As already
stated, hydraulicking is rarely possible in Australia, and then
only on a small scale. A
case in Australia is cited, however, by
Mr. G. Warnford Lock, in which three workmen washed 150
cubic yards of gravel per day with 1,875 cubic yards of water, the
jet being directed against a bank about 30 feet high. The cost
of working was 2 per day, being made up as follows :
Three men at 8s., 140

Water, 10
Maintenance, repairs, &c., . . . . 060
200
The cost was, therefore, a little more than 3d. per cubic yard,
and a saving of 1^ grains of gold per cubic yard would, in this
case, pay expenses. The cost of production in the year 1875
of 1 ounce troy of gold in hydraulicking is given opposite for
two Californian mines,* viz., the La Grange Company, which
worked on a thin bed of gravel, and the North Bloomfield Mine,
in which the bank was 250 feet high.
In drift mining, the cost of treating uncemented gravel is less
than that of treating conglomerate. The cost at the Hidden
Treasure Mine, on the Damascus Channel, Placer County, Cali-
fornia, was $0-9236 per carload of 1 ton in the spring of 1888, the
yield being $1-2347 per ton. Here the ground could all be
" "
picked down, requiring no blasting, and the gravel could be
sluiced without crushing, but the timbering was costly. The
working of this mine was exceptionally cheap, as, under ordinary
conditions, a cost of from $1*50 to $1-75 per carload is as low as
can be expected, and in many mines the cost rises to $3 per
carload. The cost of milling cemented gravel is from 20 to 40
cents per ton, the capacity of the mills being from 5 to 12 tons
per stamp in twenty-four hours.
* Ninth
Report Cal. State Min., 1889, p. 133.
-
QUARTZ CRUSHING IN THE STAMP BATTERY. 89
operation was conducted, have been found in Wales, in Central

America, in the Pyrenees, and in Transylvania. Diodorus
Siculus, the Greek historian, has given a detailed description of
the method of gold-quartz reduction employed in the mines of
Upper Egypt 1,900 years ago. The ore was first reduced to
coarse powder in mortars, then finely crushed in hand-mills
resembling the flour-mills of the present day, and finally washed.
In order to separate the pulp from the uncrushed lumps, the
Egyptians, in common with other races in ancient times, employed
sieves, but in extracting the gold from auriferous sands they used
raw hides, on which the flakes of gold were entangled.* These
devices closely resemble those still in use in many parts of the
world, f
The date of the first use of mercury for amalgamation is
unknown. Pliny mentions the fact that mercury will extract
gold from its ores.! and it has no doubt been used for this
purpose for the last 2,000 years. It is mentioned by Biringuccio
in his treatise (which was published in Italian in 1540), and
had apparently been a secret art for some time previously.
Mercury was introduced into Mexico as a means of extracting
the precious metals by Bartolome Medina in the year 1557, and
its use doubtless soon spread to Peru and the neighbouring
countries. When Barba wrote in 1639 there were three amal-
gamation machines in use in Peru,|| viz., the mortar (used in the
Tintin process), the trapiche or Chilian mill, and the maray.
The mortar was hollowed out in a hard stone, the cavity being
about 9 inches both in diameter and in depth. The ore was
triturated with water and mercury in this mortar with the aid of
an iron pestle, while a stream of water, flowing through, carried
away the crushed particles. The slimes were roughly concen-
trated, but most of the gold remained in the mortar with the
mercury.
The maray, although equally primitive, was probably of greater
capacity; it makes use of the principle employed in the bucking
hammer. About the year 1825 MiersIT saw it in Chili, where it
is probably still in use. It consists of a flat or slightly concave
stone, 3 feet in diameter, on which lies a spherical boulder of
granite about 2 feet in diameter. This is rolled to and fro by
two men seated on the ends of a long, wooden pole, which is
*Beckmann's Hist, of Inventions, vol. ii., p. 334.

t An interesting account of some primitive methods of treating gold quartz
now employed by the Chinese is given by Henry Louis in the Eng. and
Mng. Journ., Dec. 31, 1892, p. 629, from which it appears that these methods
bear points of resemblance to those of the Egyptians.
t Nat. Hist., xxx., cap. vi., sec. 32.
Georg Agricola's Bermannus . . . ilbersetzt von Friedrick August
Schmid, 1806, p. 49.
||
Arte, de los mrtales, lib. iii, cap. 16.
1 Travels in Chile and La Plata, by John Miers, vol. ii., p. 390.
firmly fixed to the boulder. Ore, water and mercury are ground
together in this machine, and then washed down.
The Chilian mill closely resembles the edge-runner mill of the
present day, which is used for grinding and mixing mortar, &c.
The Peruvian trapicJie had a similar circular bed of hard stone,
but only one stone- runner, which was driven by mules. The
Chilian mill is still used to prepare ores for treatment in the
arrastra, which was not mentioned by Barba, and may perhaps
be regarded as an outcome of the trapiche.
The Arrastra was also one of the earliest crushing machines
in use in America, being introduced at the same time as the
Patio process -i.e., about 1557 and is still in wide use in Mexico,
although chiefly in the treatment of silver ores by the Patio
process. It is a circular, shallow, flat-bottomed pit, 10 to 20 feet
in diameter, and paved with hard, uncut stones. Granite, basalt,
Fig. 13.
and compact quartz are all used for the rock pavement, which is
made 12 inches thick, and is either placed on a bed of well-
puddled clay from 3 to 6 inches thick, or set in hydraulic cement,
so that no chink or cranny remains, into which the mercury or
amalgam can find its way. In the centre a vertical shaft revolves,
carrying two or four horizontal arms, to each of which is attached
a heavy stone by thongs of bullock hide, or by chains. These
grinding stones weigh from 400 to 1,000 Ibs. each, their forward
ends being about 2 inches above the floor, whilst their other ends
drag on it. They are moved by mules walking round outside
the arrastra, or by water- or steam-power, the speed varying from
four to eighteen turns per minute. Fig. 13 represents an
arrastra of the simplest description ; at the front the stones
forming the edge have been removed, so as to expose a section
of the rock pavement.
Ore of about the size of pigeons' eggs is introduced, enough
water being added to make the pulp of the consistency of cream,
and mercury is sprinkled over it arter most of the grinding has
been done. When the ore cannot be ground any finer, more
water is added to dilute the pulp, the mule is driven

slowly for
half an hour to collect the mercury, and the pulp is then run
out and another charge shovelled in. An arrastra of 10 feet in
diameter takes a charge of 500 to 600 Ibs. of ore, and treats about
1 ton per day of twenty-four hours. The amount of wear sus-
tained by the grinding stones is equal to from 6 to 10 per cent,
of the ore crushed. The output is so limited that the use of the
arrastra has never been general outside Mexico, although it has
been used in almost every district in the United States for a
short time after the commencement of quartz mining in the
particular locality.
It is often stated that the results obtained, so far as the
percentage of gold extracted from refractory ores is concerned,
cannot be equalled by any other amalgamating appliance, and
that the Mexicans, using the arrastra, formerly treated at a profit
ores which hardly yield any gold to the stamp mill, or even to
the amalgamating pan. In consequence of its power of saving
gold and the cheapness with which it can be erected and worked,
the arrastra is still valuable for prospecting. In preparing the
bed for this purpose, every care must be taken that the surface is
even, and all joints properly cemented, or else the mercury and
amalgam will in great part find their way into the founda-
tions.
The reasons for the high extractive power of the arrastra, when
treating certain ores, are no doubt to be found in the extreme
fineness to which the ore is reduced, and in the prolonged contact
between the ore and mercury, which is maintained while they are
being ground together. Moreover, the grinding action of the
dragged stones keeps the particles of gold bright and in a suitable
condition for amalgamation, without exercising force enough to
flatten and harden them. The relative advantages of grinding
surfaces consisting of iron and stone are less certain, and probably
vary with the nature of the ore in course of treatment. An
instance is given by Colonel Harris,* in which stone was better
than iron. In this case, at Cerro de Pasco, Peru, the old method
of grinding ores in circular pans, by edge-runners of stone or
granite, was found to entail a rapid wearing of the edge-runners,
and, in order to remedy this, the runners were shod with iron.
The returns at once fell off, and on careful trial it was found that
the yield in the old machines was 15 per cent, more than in the
new ones. Other similar instances are on record, but it must
nevertheless be conceded that, with some ores, the presence of
iron is necessary for good work, chlorides of silver and mercury
being reduced by it, which would otherwise be lost in the
tailings. The somewhat extravagant views often expressed,
upholding the great superiority of the arrastra over its more
modern rivals as an amalgamating machine for gold ores, are
*
Min. Journ.y December 24, 1892, p. 1466.
perhaps hardly justified, since a direct comparison with stamps

or pans has been possible in only a few instances.
One of the most remarkable of these instances is that afforded
by the experience at the Pestarena Mines, Val Anzasca, Italy.
The ore of this mine contains about 20 per cent, of pyrites, and
is of somewhat low grade, rarely containing as much as 15 dwts.
of gold per ton. Efforts were at first made to treat it by means
of stamps and amalgamated plates, with the result that only
65 per cent, of the gold was extracted. Better results attended
the introduction of the Francfort mill, a modified arrastra, driven
in this instance by steam; this mill is substantially a wooden
pan, with dies and shoes of stone. The mercury was added to
the pulp after it had been finely ground, and the amalgamation
and grinding of the pulp subsequently kept up for seven hours.
From ore containing 12-3 dwts. per ton, the mill extracted 1O2
dwts., or 83 per cent., for a considerable period of time, while in
the year 1890 the average extraction was 79 -4 per cent. There
were twenty-eight mills, each of which treated 1,200 Ibs. of ore
per day. The stone bed is said to last ten months, and the shoes
from six to eight weeks, the total cost of treatment being very
low.
Successful work by the arrastra on gold ores is also being done
at the present day at the Bote Mine, Zacatecas. The material
treated is an auriferous silver ore, the gold being extracted in
the arrastra, and the silver subsequently obtained by means of
the Patio process. Here the closely-fitting cut-stones forming the
bottom of the arrastra are first coated with an amalgam of silver,
and the ore then ground for twelve hours, mercury being added
to it to the extent of one and a-half times the weight of the gold
present, so as to form a "dry" amalgam. The extraction of
gold in this case, however, is only 60 per cent., a result which is
probably not better than that which would be obtained in pans
and settlers, if economical conditions admitted of their employment.
The necessity of keeping the amount of mercury low so as to
"
obtain " dry amalgam, and thus to prevent loss by leakage into
the foundations, is one of the objections to the use of a roughly
made stone-floor for amalgamation, the percentage of extraction
being necessarily reduced by this precaution. There were about
100 arrastras at work in California in 1889, each treating from
1 to 3 tons of ore per day. They are used where only small
quantities of high-grade ore are available for treatment.
At Smartsville, Nevada County, the arrastra has lately been
applied to a new purpose, that of crushing and amalgamating
hard cement-gravel obtained from a drift mine.* The cement is
coarsely crushed by being passed through a Gates rock-breaker,
and is then charged into four arrastras, each of which is 12 feet
in diameter and 3 feet deep, and capable of containing from 5 to
* Eleventh
Report Gal. State Min. t 1892, p. 315.
9 tons of gravel. The grinding is effected by four blocks of diabase,

each weighing from 600 to 1,000 Ibs. ; the rate of revolution is
14 times per minute, the time of grinding being one hour. A
tablespoonful of mercury is fed in wifch each charge, and the
total loss is only 10 per cent, of this. The pavement costs $40
and lasts for six months. At the end of the hour, a gate is
opened in the side of the arrastra and the charge run out into a
sluice, 200 feet long, where the mercury and amalgam is caught
by means of riffles. The capacity of one arrastra is 50 tons of
hard cement per day, or 75 to 90 tons of soft surface-gravel per
day. The cost is from 6 to 8 cents per ton, and more gold is
extracted than when a stamp battery was used at a cost of from
20 to 40 cents per ton. One man attends to the four arrastras,
and another to the Gates crusher.
Iron Prospecting Arrastra. The difficulties of getting suit-
able stone for the arrastra in some localities, and of constructing
the pavement so as to prevent loss of mercury, have led to the
introduction, for prospecting purposes, of wrought-iron pans with
steel bottoms on which stone drags are used. One of these so-
called arrastras is 3 feet in diameter and 12 inches deep, with an
axis revolved by bevel gear, which is placed below the bottom.
The upper part of the axis carries a yoke to which stone drags
are attached by chains. It is doubtful if the steel bottom is as
good for amalgamation as the stone pavements, but the losses
of mercury by leakage are certainly avoided. Perhaps a thin
pavement of flat stones might be put inside such a prospecting
pan with advantage, when working on some ores.
The Stamp Battery. The stamp battery was, no doubt,
evolved from the pestle and mortar, but was not introduced until
a comparatively recent date. Beckmann* states that mortars,
mills and sieves were used exclusively in Germany throughout
the whole of the 15th century, and in France stamps were
unknown as late as the year 1579. It has been suggested that
the origin of stamp mills was probably due to the gunpowder
manufacture, and it seems certain that in 1340 stamp mill,
used in connection with this industry, existed in Augsburg, and
that Conrad Harscher, of Nuremburg, owned one in 1435. They
were first applied to the gold industry at the beginning of the
sixteenth century, a doubtful record stating that they were intro-
duced into Saxony by Count von Maltitz in 1505, whilst in 1519
the processes of wet-stamping and sifting were established in Joa-
chimsthal by Paul Grommestetter, who had some time previously
introduced them at Schneeberg. The improvements gradually
spread through Germany, and detailed descriptions and drawings
of the apparatus are given by Agricola f in 1556, from which it
appears that the earliest battery consisted of a single stamp,
* Hist,
of Inventions, vol. ii., p. 334.
t De re. Metallica, vol. viii., p. 247.
raised by means of two levers fixed to the axis of a wheel. Dry

crushing was at first employed, but the great production of dust
soon led to the use of water in the mortar. In some Hungarian
*
mines, Bennett H. Brough recently saw some primitive stamps
in use, resembling those drawn by Agricola, weighing only
100 Ibs. each, and having stamp heads made in some cases oi
hard blocks of quartzite. At that time, in cases where the
conditions of water supply were favourable, these stamps were
able to treat with profit an ore containing as little as 2| ozs. of
gold to 50 tons of ore, and, at Zell, in the Tyrol, they were
able to treat a slaty material containing 1 oz. of gold to 50 tons
of ore. Such economical work is seldom possible with the heavy
modern Californian stamp under the most favourable circum-
stances.
The German stamp has a rectangular stem made of wood or,
latterly, of iron, with an iron head, the total weight never
exceeding 300 to 400 Ibs. It was introduced almost unchanged
into France, Cornwall, and, after the discovery of gold in 1849,
into the United States, but has given place in all new districts
to the Californian stamp, and need not be fully described here.
In 1850, the first Californian stamp-mill was erected at Boston
Ravine, Grass Valley. The stamps consisted of tree-trunks shod
with iron, and the framework was constructed of logs.
The ordinary method of reduction and amalgamation of gold
quartz in a stamp battery now consists of the following opera-
tions :
1. The oreis broken down to a moderate size, varying from

that of a man's fist to a nut, by passing through the jaws of a
stone -breaker, or by hand hammers.
2. The ore is then fed into the mortar-box of a
stamp mill,
where it is pulverised to any required degree of fineness. In
wet crushing, a stream of water is introduced also, and the blows
of the stamps splash the water and pulp against screens set in
the side of the mortar, the finely-divided ore being ejected in
this way. In some cases the mortar-box is partly lined with
amalgamated copper plates, by which some of the gold is caught
and retained, mercury being in this case usually fed into the
mortar-box with the ore and water.
3. On issuing from the
battery, the pulp is allowed to run
slowly over a series of inclined, amalgamated, copper plates, by
which a further percentage of the gold is amalgamated and
retained.
4. The tailings are sometimes further treated
by running
over rough hides or blankets, by which some particles of gold
and pyrites are retained, or the pyrites are separated from the
valueless sands by concentration on some form of vanner or
jig.
These concentrates are subjected to further treatment, usually
either by smelting or by chlorination.
*
Proc. Iiist. Civil Ewj., Session 1891-1892, part ii.
5. At intervals the gold amalgam is wiped off the copper

plates, the excess
of mercury separated by squeezing through
filter bags of chamois leather, buckskin, or canvas, and the solid
amalgam thus obtained is retorted so as to distil off the mercury,
and the gold is then melted.
In the following pages much information has been derived
from (1) "The Milling of Gold Ores in California," by J. H.
Hammond, published in the Eighth Report of the Ccdifornian
State Mineralogist, 1888; (2) the series of articles on "The
Variations in the Milling of Gold Ores," by T. A. Rickard,
published in the Engineering and Mining Journal in the years
1892 to 1895 ; and (3) " The Amalgamation of Free-Milling Gold
Ores," by Louis Janin, Jun., published in the Mineral Industry
for 1894.
The stamp battery must be regarded from two different points
of view viz., (a) as a crushing machine, (6) as an amalgamating
machine, and it should be remembered that the modifications
designed to make a more efficient crusher often reduce its
it
power as an amalgamator, and vice versa. Rickard has pointed

out that two typical methods are the Californian practice, for
the treatment of "free-milling" gold ores, and the Colorado
practice, for the treatment of ores, sometimes called "refractory,"
which, however, yield most of their gold when carefully treated
in the stamp battery. The word "refractory" is better reserved
for those ores which cannot be satisfactorily treated by direct
amalgamation, whatever be the method adopted. The stamp
battery is also used purely as a crushing machine in preparing
certain silver ores (whether gold is present in addition or not)
for treatment by other processes (e.^.,theWashoe, Reese River, and
lixiviation processes), which properly appertain to the metallurgy
of silver and will not be described in this volume, although a
short account of pan-amalgamation is given on pp. 158-163.
The Californian practice consists briefly in crushing the ore and
effecting its discharge from the battery as rapidly as possible.
With this object in view heavy stamps are used, running very
fast, with a small drop ; the screen area is large and placed as
low down as possible, and the mortar is made narrow, with
nearly vertical sides. These arrangements all increase the out-
put of the battery. There are usually amalgamated copper plates
inside the mortar on the discharge side only. The method is
suitable for ores containing coarse, free gold which is easily
amalgamated, and which is caught largely on the inside plates,
in spite of the short time during which the ore remains in the
mortar. A
small amount (1 to 5 per cent.) of pyrites, especially
of clean iron pyrites, does not interfere with rapid amalgamation;
but if the percentage of this mineral is high (10 to 20 per cent.),
and especially if the easily decomposable variety of sulphide of
iron (marcasite), or some sulphides of lead and zinc, or if com-
pounds of arsenic or antimony or other base minerals are present,
the amalgamation is greatly retarded or prevented, and the

Colorado practice is resorted to. If the gold is fine, or if, for any
other reason whatever, amalgamation is difficult, the Californian
practice must be modified in the same direction.
The Colorado practice was first devised in Gilpin County,
Colorado. In the early days of this gold-field, about the year
1860, after the arrastra had been discarded as being too slow,
the fast-drop and shallow-discharge batteries, like those used in
California, were introduced, and gave good results in working on
the oxidised surface-quartz, 60 per cent, to 75 per cent, of the
gold being extracted. As the mine workings got lower, however,
the percentage of pyrites steadily increased, and the mills gave
poorer and poorer results, until a return of only 30 per cent, to
40 per cent, was obtained. Areturn to efficient working was
only made after a long series of costly experiments, which
resulted in the present slow-working, long-drop stamps with a
wide, roomy mortar and a very deep discharge, the lowest part
of the screen being more than a foot above the dies, instead of
only 6 inches as in California. Amalgamated plates were put
inside the mortar on both feed and discharge sides. The object
of these arrangements was to keep the ore in the mortar for a
long time, so as to increase the chance of catching the gold
on the inside plates. The duty of the stamps was of course
greatly diminished, the output of a typical Colorado battery
being only about 1 ton per stamp per day of twenty-four hours,
while with Californian practice it is from 2 to 5 tons. Never-
theless the ultimate object is equally well attained by both
types of battery viz., the extraction of from 75 per cent, to 95
per cent, of the gold present, including that saved in the concen-
trates.
Ores containing more than 20 per cent, of sulphides', unless
they are clean, cubical iron pyrites, usually come under the
heading of true refractory ores, and cannot be treated by any kind
of stamp-battery amalgamation. Their treatment is considered
in succeeding chapters. The following is a general description
of the machinery employed in stamp-battery practice. In the
sequel the modifications adopted in different districts will be
discussed, and the conditions for successful amalgamation of
different classes of ore referred to.
1. Bock -breakers. There are two classes of these machines
in general use, viz.: Those constructed on the reciprocating-
(a)
jaw principle, and (b) gyratory crushers. The Blake and the
Dodge crushers are representative of the former class, and the
Gates crusher of the latter.
CrusJier is shown in section in Fig. 14.
The Blake The rock
is crushed between the stationary jaw, BC 1
and the swinging
,
jaw, D, which is pivoted at E, and moved by the eccentric, F,

through the toggles, J K.. The swinging jaw, in order to be
as light as possible, should be made of steel, cast hollow and

braced by ribs, but is usually composed of cast iron. The
machine works at about 250 revolutions per minute. At each
revolution the moving jaw is advanced about ^-inch towards the
other, and the lumps of rock which have dropped down between
the jaws are broken. as the moving jaw recedes, the fragments
;
slip lower down and are further crushed at the next advance,
and this process is repeated until the ore is small enough to
pass out at the opening at the bottom. The distance between
the jaws at the bottom limits the size of the fragments, and
this distance may be regulated at will by moving the wedge,
L, or by changing the length of the toggles, J K. The capacity
of the machine is great, being about 300 tons of ordinary rock
per day of twenty -four hours in the case of the machine whose
Fig. 14.
dimensions at the mouth are 20 inches by 10 inches, when the

lower edge of the jaws are set to approach within 1^ inches of each
other. The power required for this is stated to be 14 H.P.
Many modifications have been made in this machine by
various makers since the patent expired, and of these the Blake-
Marsden machine has come into more extended use than any
other reciprocating-jaw crusher. The most important alteration
is the pivoting of the moving jaw below instead of above, which
isadopted in the Dodge crusher shown in Fig. 15. The effect of
this arrangement is to make the product more uniform in size,
and as there is little or no motion of the movable jaw at the
delivery aperture, this may be made as narrow as desired, so
that a finer product can be obtained, although it is at the ex-
pense of capacity. The Dodge crusher is more particularly
recommended for fine crushing in concentration works, or where
the product is to be subsequently passed through rolls. pre-A
liminary treatment of the ore in a Blake crusher is desirable,
where fine crushing by a Dodge is resorted to.
7
Authorities differ as to the relative advantages of the two

positions of the pivot. A. Sahlin decides * in favour of placing
the pivot below, assigning as a reason that more work is neces-
sary to crush comparatively fine material than to break down
large pieces of rock in the upper part, where the points of
contact between the crushing jaws and the rock are few. The
shortest stroke and great leverage should, therefore, he con-
siders, be at the bottom, where the work is heaviest. On the
other hand, W. P. Blake dissents from all these views.
Fig. 15.
Gates Crusher. This machine has been in use for fifteen years
in crushing macadam, ballast, and iron ore, chiefly in the United
States, but has not long been applied to crush gold ores. It is
now largely used, however, both in America and in South Africa,
being probably the most economical rock-breaker where large
quantities of ore are being treated. It is shown in Fig. 16, and
consists of a vertical shaft of forged steel, G-, rotated at the
bottom by a bevelled wheel, L, placed J inch out of centre. At
the top of the shaft is a chilled-iron breaking head, F, and the
shell surrounding this is lined with twelve chilled-iron, concave
pieces, E. These form the crushing faces. The shaft, G, has a
gyratory motion imparted to it by the eccentric box, D, attached
to L, and the rock is thus crushed, without grinding, between
the head and liners. The distance between the crushing surfaces
at the bottom may be regulated by set-screws. With dry ore
this distance may be as low as f inch, no pieces larger than this
being allowed to pass. It is stated that this machine works
with less expenditure of power than the Blake crusher, and that
its product is more uniform and can be made finer. Its first
cost, however, is higher, and, what is of more importance, the
*
Trans. Am. Inst. Mng. Eng., 1892.
repairs are troublesome and expensive, but it certainly works

AY ell in many places where it has been adopted.
Fig. 16.
The Schranz Stone-breaker. This machine (Fig. 17) is

of interest, as it forms a link between the ordinary stone -breaker,
with reciprocating jaws, and the crushing rolls. The movable jaw,
instead of having a reciprocating motion, has a rocking one, some-
what similar to the motion of a circular blotting-pad. The jaws
are of cast steel, and the space between them is regulated by
means of the vertical screw, I, which adjusts the wedge, c. The
short connecting-rod, b, made of cast iron is made with a narrow
section, so that it is broken before any other part of the stone-
breaker, if some very hard material should find its way between
the jaws. The machine has the advantage, in common with the
Dodge crusher, that the opening at the bottom does not vary, so
that a uniform-sized product is secured, the maximum diameter
of which can readily be reduced to 8 mm. (0'3 inch). The
machine can thus take the place of roughing-rolls. The large
sized machine, working at 250 revolutions per minute, and with
an expenditure of 10 to 12 horse-power, is stated to crush from
100 THK METALLURGY OF GOLD.
4 to 5 tons of rock per hour to the size given above, and thus
compares favourably with the Dodge crusher. It is in use at
Laurenburg, on the Lahn, and at other places in Germany,
and gives great satisfaction.
Material Employed for the Crushing Surfaces. The
selection of the most suitable material for the working parts,
and especially for the crushing surfaces, of reduction machinery
is a matter of the greatest possible importance, as the economy
effected by using a durable material, which seldom requires
renewal, is very great. When rock is crushed by repeated blows,
as in the case of ordinary rock-breakers, stamps, and, perhaps,
Fig. 17.
it does not follow that the hardest material is

rolls, always
the most durable. A substance is wanted that will not be
broken or caused to crystallise unduly by a blow, but which,
at the same time, will show neither signs of deformation
nor rapid attrition by the impact of fragments of quartz.
Both chilled iron and hard chromium-steel have been proved to
be very useful for the surfaces of impact machines, but the softer
and tougher open-hearth mild steel has also been found advan-
tageous for the surfaces of rolls. The quality of toughness seems
on the whole to be of even more importance than mere hardness.
No doubt different qualities are required for the crushing surfaces

in different machines, and even in the same machine when
operating on different classes of ore. Thus chromium - steel,
which has been found so useful for the manufacture of stamp
heads and dies, was not a success when used in the Huntington
mill, where the action is rather one of grinding than of impact.
Perhaps Hadfield's manganese steel is the best material for the
jaws of stone-breakers. At the Penmaemnawr granite quarries,
the plates inserted at the sides of the jaws to protect the outer
casing of a Blake crusher were worn out in two months,
when
ordinary steel was used, while similar plates made of
cast
manganese steel, after being in service for more than twelve

months, had only been worn to the extent of J inch, and were
still at work. In stamp-mills, the crushing faces, if made of chilled
iron, last for six or eight months, while steel faces last twice
as long. The whole question of the nature of the materials to be
used for different purposes is one for the metallurgist, and too
much attention can hardly be paid to it in the future.
Position and Use of Rock-Breakers. The aperture of a
rock-breaker should be placed on a level with the floor, so that
the ore can be dumped down by the side and shovelled into the
jaws. As noted below (p. 121), it is now becoming customary
to place the rock-breaker in a separate building distinct from
the battery house.
Agrizzly is employed to separate the fine material, which is
passed straight to the stamps, while in some cases the material
from the rock-breaker is also screened and part returned to be put
through again. Revolving trommels and flat shaking screens
have both been used for this purpose.
The efficiency and economy in crushing, attained by the
machines on the reciprocating-jaw principle, are so fully recog-
nised that there has been a tendency of late years to use them
to reduce gold-quartz to a very small size before feeding it into
the stamp batteries, the capacity of which is greatly increased in
this way. For fine crushing, multiple-jaw crushers, on the
same principle as the Blake, have been constructed, but have
not passed into general use the use of a pair of rolls between
;
the rock-breaker and the stamp battery has also been advocated.
This is done by the Huanchaca Mining Company at Antofagasta,
Chili, and gives an increase of capacity to the stamps of over
20 per cent., while the cost is trifling. The size of the ore
best suited for feeding into a stamp battery may be roughly
put down as about J inch for light stamps and J inch for heavy
stamps. If the size of the material is much smaller than this,
no advantage in speed is gained, while the jar given to the stamps
and framework is greatly increased. At the present day few
large mills are erected without rock-breakers, which have also
been successfully added to many old mills. Nevertheless, they
are absent or rarely seen in several of the richest gold-fields of

the world. In Gilpin County, Colorado, they are not used in ;
Victoria the extent of their use may be judged from the fact
that there were 5,901 stamp-heads in operation in 1891, and
only twelve stone-breakers, and in other parts of Australia they
have been similarly neglected.
Californian " gravitation stamps are

"
The Stamp Battery. .
in general use at the present day for crushing gold ores. Astamp
is a heavy iron, or iron and steel, pestle, raised by a cam keyed
on to a horizontal revolving shaft, and let fall by its own weight.
Stamps are ranged in line in groups of five stamps each, which
have a mortar-box in common. Fig. 18 represents the side view,
and Fig. 19 the front view, of a ten-stamp battery, with the
amalgamating tables removed to show the foundation timbers or
mortar-blocks, A.
The foundations are of the highest importance, as, if they are
badly made through carelessness or false economy, the efficiency
of the battery is greatly decreased, and it soon shakes itself to
pieces. The blow of a stamp is partly employed in crushing the
ore, and is partly expended in producing a concussion or jar
acting on the framework and foundations. The amount of energy
used up in the latter way depends largely on construction, for
details of which the student is referred to the volume on Metal-
luryical Machinery. In preparing the ground for the foundations,
the earth is removed until bed-rock is reached if possible,
and the latter is then carefully smoothed and sometimes covered
with a layer of cement. The wooden mortar-blocks of from 6 to
14 feet long are placed upright in this trench, and the space round
filled up with sand, or, as in the Transvaal, solid masonry is
builtround the blocks. The framework is now usually made of
wood, which is a far more satisfactory material for the purpose
than iron. It consists of the massive battery sills, B, on which
rest the battery posts, C, and the braces, E. The posts are held
together by the stamp-guides or tie-timbers, D. Wooden braces
have completely replaced iron rods, which allow the battery to
spring. It is better to place the braces on the discharge side
alone, thus leaving more room to work on the feed side.
The Mortar. The mortars are made of cast iron, but differ in
shape according to the nature of the ore and the corresponding
modifications made in the course of treatment. They weigh
from 1J to 3 tons, being especially thick at the bottom where
there is the greatest strain. An ordinary mortar is about 4 feet
7 inches long, 50 inches high, and 12 inches wide on the inside
at the level at which the dies are set. The bottom is from 3 to
8 inches in thickness, and has a heavy flange cast on it, by which
it is bolted to the mortar-blocks. These are tarred over, all
cracks in them having been filled with sulphur, and are then
covered with three thicknesses of blanket, carefully coated with
tar on both sides. The mortar is placed on these blankets and
securely bolted down. This arrangement lessens the chance of
the mortar working loose, the jar being diminished. A sheet of
rubber, J inch thick, may be used instead of the blankets. Figs.
20 and 21 represent sectional elevations of two forms of mortars
in use in the United States, Fig. 20 showing a single-discharge
narrow mortar for wet crushing, and Fig. 21 a wide double-
discharge mortar for dry crushing. In both figures, b is the

feed -opening through which the ore is introduced into the
mortar ; c is the bed on which the die is placed d is the screen-
;
opening. Single-discharge mortars vary in shape, according as

they are intended to contain amalgamated copper lining plates
both at the front and back, or on the screen side only. The
chief difference between them is in the feeding arrangement ;
in the former case the back plate is put in a recess, and is pro-
tected from the falling rock fed into the battery. The mortar
shown in position in Fig. 18 is one of this kind. The plates
catch the coarse gold inside the mortar when the pulp is
flung against them. Occasionally copper plates are also put
at the ends of the mortar. All the plates, which are bolted
through the mortar itself or to its lining plates, must be so
arranged that they can easily be taken out and cleaned, as, when
Fig. 20. scale, l in. =3 ft. Fig. 21.
very rich ores are being treated, precious metal accumulates on

them very fast.
The width of the mortar varies from 10 to 14 inches at the
level of thebottom of the screens. As has been already men-
tioned, narrow mortars are best fitted for rapid discharge, but, if
hard flinty ores are to be crushed, a narrow mortar causes
frequent breakage of the screens, unless the discharge is deep
i.e., unless the bottom of the screens is a considerable distance
above the surface of the dies. By this latter arrangement the out-
put is reduced, since, the nearer the screens are to the dies, the
more rapid is the discharge. The depth of the discharge is only
about 6 or 7 inches in California, where an adjustable battery-
screen has been introduced to keep this depth constant, in spite
of the wearing of the dies. In the battery thus modified, the
screen-frame is supported on a wooden block, which is easily
removable, and to which the copper plate is bolted. When the
dies wear down, this block is replaced by one of less height, to
which a suitable plate has already been fixed.
Mortars often have a lining of cast-iron plates, bolted to them
near the bottom, to protect them from the rapid wear due to the
splashing of the pulp. These plates last from six to nine months,
and can be replaced when worn out.
The splasJi-box, not shown in the figures, and now often
omitted, is bolted to the outside of the mortar just below the
screens. It is rectangular, consists of wood or iron, and is of the
same length as the mortar. It receives the pulp as it passes
through the screens, and distributes it evenly over the amalga-
mating tables by a number of spouts, usually three. Instead of
the splash-box, a splash-board is now almost universally employed,
the usual material for it being heavy canvas. The old form of
mortar had its upper part, or housing, of wood (see Fig. 44,
p. 195), but, as mercury is lost through the smallest aperture,
and it was difficult to make these wooden housings quite tight,
mortars are now cast in one piece, including the housings. The
roof of the mortar is made of 2-inch planking, through which holes
are cut to admit the stems of the stamps and the water pipes.
When the mortar is in place, the dies are put into it, a layer of
sand being often introduced first. The dies consist of two parts,
the footplate and the die proper, or boss.
A Fig. 22 shows, in plan and elevation,
one of the many forms of dies in use ;
here the footplate is almost square, so
i
as to fit the mortar; it is 1 or 2 inches

*>
thick, and 10 or 12 inches square. The
boss is cylindrical, from 3 to 6 inches
high, nnd of the same diameter as the
shoe. Shoes and dies are made either of
iron or steel, hard chilled-iron being used
for wet crushing, and soft iron for dry
crushing. Cast-steel dies and shoes have
been often tried, but owing to their
tendency to chip and to their uneven
wear, their introduction formerly met
with little success. Nevertheless, in
most mills remote from foundries, where
transportation is an important item in
t k e cQst of SU ppij eSj s teel shoes and dies
have now replaced those of iron, as the life of steel is from two
and a-half to three times that of iron, and the cost only about
twice as great.* About seven years ago, chrome-steel shoes and
dies were introduced into California, and proved their superiority
over those manufactured from most other kinds of steel forged ;
steel and, according to Janin, manganese steel have also been

used with success. Sometimes steel shoes and iron dies are used,
the wear being more even than when both consist of steel. At
the Patterson mine in 1888, a set of iron shoes and dies lasted
*
Eighth Report Cat. State Min., 1888, p. 7C8.
106 THF METALLURGY OF GOLD.
on an average six weeks, crushing 1,680 tons of ore, while one

set of chrome-steel shoes is said to have lasted fourteen months,
crushing 16,800 tons, or ten times the amount crushed by iron.*

The wear of iron shoes and dies is stated to be about 2 or 3 Ibs.
per ton of ore crushed in California. In Colorado, at the New
California Mine, the wear of shoes was 1 1 '3 ozs. of iron per ton
of ore, and that of the dies 4*5 ozs. At the Robinson Mine,
South Africa, the wear of steel shoes and dies is, according to
Mr. Harland, 0*45 Ib. per ton crushed for shoes and 0'30 Ib. per
ton for dies. When the boss is worn down to within from J inch
to 1 inch of the footplate, the die is replaced. Dies wear more
slowly than shoes since they are protected by a layer of pulp,
which is from 1J to 3 inches thick. The dies should all be
renewed together, as it is important that those in the same
battery should be of equal height, otherwise one or more will
become almost bare of ore, and a disastrous pounding result.
If a die breaks, it should not be replaced by a new one, but by
one worn to the same extent as the others in the battery. Iron
false-bottoms or chuck-blocks are placed beneath partially- worn
dies, so as to keep the depth of discharge constant.
The cam-shaft, H, Fig. 19, is of wrought iron, and about
5 inches in diameter (A, Fig. 23). It is now usual to have a
separate cam-shaft for each five or ten stamps, which have thus a
separate driving wheel. The advantage of this arrangement is
that repairs can be done to one or more stamps without necessi-
tating the stoppage of the whole mill, as used to be the case when
there was only one cam shaft. The cam-shaft is placed at a
distance of from 5 to 10 inches from the stem-centre, and is 9 to
10 feet above the mortar bed. The bearings rest on supports
attached to the battery posts on the discharge side.
The cams are made of cast iron, with chilled faces, which
are 2 to 3 inches wide (B, Fig. 23). The double cam, two views
of which are shown in Fig. 23, is now in almost universal use,
though single and treble forms have been employed. Sometimes
cams are cast in two pieces which are bolted together, so that
when one is worn out, it can be replaced without first removing
the other cams on the shaft. It is stated, however, that these
sectional cams work loose, and are not much used in conse-
quence. The hub is always strengthened by a band of wrought
iron shrunk on it. The shape of the cam face is the involute
of a circle slightly modified at the end so as to stop the upward
motion gradually. The radius of this circle is equal to the
distance between the centres of the cam-shaft and stem, which
depends on the height to which the stamp is to be lifted, so that
the curve of the cam varies with the drop. A cam should last
several years unless broken through being a faulty casting, or
through carelessness in letting the stamp fall when hung up.
The cam-face works against the iron collar or tappet, shown in
*
Eighth Report Cal. State Min., 18SS, p. Co6.
plan and section in Pig. 24, which is bored out at A

to fit the
stem of the stamp. The tappet is fitted with a wrought-
iron gib, which is pressed against the stem by two or three
keys behind it, thus binding the tappet firmly on the stem
while, at the same time, admitting of rapid adjustment
to another position. The entire end surface of the tappet
comes in contact with the cam-face, by which the stamp is
raised and, at the same time, rotated, all its parts being round.
The effect of this is, that the shoe does not strike the ore in the
mortar in exactly the same place twice in succession, and the
wear of its face is made more uniform. The greater part of the
revolution takes place during the raising of the stamp, but the
latter does not quite cease to rotate as it falls, and a slight grind-
Fig. 23. Fig. 24.
ing action on the ore has been noticed by many observers. The
amount of rotation varies with the fall, the extent to which the
cam and tappet are greased, and the state of wear of their sur-
faces. A little grease is always added to reduce wear, but, if too
much is present, the stamp does not revolve at all, while, accord-
ing to J. H. Hammond, when the tappet is in the right condi-
tion, one revolution is effected in from four to eight blows, with
a 6- or 8-inch drop. Other observers find the usual rate of
rotation more rapid, and in Gilpin Co., Colorado, where the
average drop is from 16 to 18 inches, the stamp makes from 1J
to 1 J revolutions at each blow. Tappets should last for four or
five years ; and, having both ends alike, they can be reversed
when one end is worn out, and their worn and grooved faces
can be planed down when necessary. Some millmen assert that
tappets may be broken by the cam if keyed too tightly to
the
stem.
As the cam-thrust is not applied at the centre of the stamp,

there is always a considerable side pressure, which greatly in-
creases the friction in the guides and wears the latter out,
besides causing a loss of power. Moreover, another result of
this is that the stamp tends to be inclined (not vertical) when it
is released, and so the blow on the die is given slightly to one
side i.e., the side of the dye on which the cam works. Conse-
quently, there is a tendency for this side to wear down more
quickly than the other. To obviate these disadvantages, cams
have been introduced at Johannesburg with a wide hub, and
the two blades set one at each end of the hub, so that they work
on opposite sides of the stamp and cause it to revolve in different
directions at each successive uplift. The Blanton cam, now
generally used on the Rand, is fastened to the shaft by a semi-
circular wedge pinned to the cam-shaft, no keys being necessary.
The pulley on the cam-shaft (F, Fig. 19) is made of wood on
cast-iron flanges ; if iron alone were used, the rapid succession
of jars, caused by the dropping of the stamps, would soon cause
the material to crystallise and break. A tightener pulley on the
belt driving the cam-shaft is often used, by which the stamps
can be put in motion or stopped without interfering with the
driving power.
The stamp itself consists of three parts, the stem, the head,
and the shoe. The stem (G, Fig. 19) is from 12 to 15 feet
Fig. 26.
long, and from 3 to 3| inches in diameter; it is made of wrought

iron, and has both ends tapered for a length of 6 or 8 inches to
fit the heads, so that, if one end is broken off, the stem can
be inverted and the other end used. The head (Fig. 25) and
shoe (Fig. 26) are made of equal diameter viz., about 8 to 10
inches. The head is of tough cast-iron, about 15 to 20 inches
high, and has a tapered socket at each end, the upper one (A B)
for the stem and the lower (E F) for the tapered shank of the
shoe. When these are driven into their respective sockets, into
which a few strips of wood are inserted to keep the two metal
surfaces from touching each other, a few blows by the stamp
bind them securely together, no other fastening being necessary.
Slots are provided (shown in the figure), at the base of the two
sockets, through which wedges may be driven to force out the
shoe or stem when necessary. The head is often strengthened
by bands of wrought iron, shrunk on at each end, to prevent
splitting by the wedge-like action of the tapering
stem and shoe.
The head lasts several years, being rarely ruptured. The shoe
(Fig. 26, which is on a larger scale than Fig. 25) consists of two
parts, the shank, which fits into the head,
and the shoe proper
or butt. The latter is made of very hard white iron, and the
shank of softer iron ; steel is also used, as has been already
mentioned. The diameter of the shank is about half that of the
butt. The shoe is replaced when the butt, which is from 6 to
12 inches in length when new, has been worn down to about
1 inch in length. To keep the total weight of the stamp con-
stant, several sizes of heads are sometimes used in one mill,
the heavier heads taking partly-worn shoes. "Chuck-shoes"
are inserted between heads and shoes with the same object.
relative weights of tappet, stem, head and shoe, which
The
together make up the stamp, vary considerably. There is an
advantage in increasing the weight of the stem, as one of small
diameter tends to spring and bend from the blow of the cam,
or when the stamp falls, and to wear the guides rapidly. The
stem weighs from 250 to 475 Ibs., the tappet from 80 to 130 Ibs.,
the head from 175 to 370 Ibs., and the shoe from 100 to 230 Ibs.
The total weight of the stamp is usually from 650 to 1,150 Ibs.,
but is sometimes as low as 450 Ibs., and, for prospecting purposes,
the weight is only from 100 to 300 Ibs. Old dies weigh from
20 to 50 Ibs. when they are discarded, and old shoes from 25 to
40 Ibs. A steel tappet on a 900-lb. stamp weighs 112 Ibs.
The stamp stems are guided in boxes bolted to the wooden
cross-ties, which also serve to hold the battery posts together.
There are two of these guides (D, Fig. 19), one within a foot or
so of the top of the battery posts, and the other as low as the
raising of the stamp head will allow. The depth of each guide
is about 15 inches, and the stems are fitted closely to the guides,
metal boxes being used occasionally, although wood is much more
general. The guide-boards are sometimes pierced with large
square holes in which bushes of wood, with the grain parallel to
the length of the stamp, are placed fitting the stem exactly. In
this way, the guide-boards themselves are preserved from wearing
out. Sectional guides, consisting of a series of iron keys enclosing
wooden bushings, are also used. In this case each stem has a guide
to itself, and the bushings can be renewed by hanging-up the one
stamp without stopping the other stamps in the battery.
Each stamp is provided with a lever or jack (I, Fig. 18) made
of wrought iron, or of wood protected by iron. The jack is for the

purpose of raising the stamp and hanging it up out of reach of
the cam. When this is to be done, a strip of wood, an inch or
more thick, is laid with one hand on the cam as it rises, and the
stamp is thus raised an inch higher than usual, so that the jack
can be slipped in under the tappet with the other hand. The
stamp is thus suspended above the cam and can be repaired
without stopping the others, while it can only be released in a
manner similar to that in which it was hung up. Above the
stamps there is a double rail, on which is a movable pulley ; by
this the stamps, &c., can be lifted up for repairs. When the
stamps are to be set up, the head is put on the die and the stem
dropped into it, iron being left sometimes on iron, but more
usually some canvas or other packing being put into the head
socket and the stem dropped into that. The stamp is then
raised and dropped into the shoe, the shank of which is often
surrounded by strips of wood for packing. As already stated,
the parts are soon wedged firmly together by raising and letting
fall the stamp a few times.
The height of the " drop " of the stamps varies from 4 to 1 8
inches, and the number of drops per minute varies from 30 to
over 100. These depend on one another to a great extent,
an increase in the height of the drop being necessarily accom-
panied by a diminution in the number of drops per minute.
"With a drop of 8 inches, about 95 blows can be obtained, the
tappet then just having time to fall after leaving one face of the
cam, before the other begins to raise it. As, within certain
limits and under certain conditions, an increase of speed results
in an increase of yield of pulverised ore, efforts have been made
to raise the number of blows per minute. It has been proposed
to use two cam-shafts, one above and one below the gib-tappets,
a speed of 250 blows per minute being thus attained, according
to the statements of the manufacturer, but this device has
apparently not passed into use. The subject will be returned to
when the conditions for successful amalgamation are discussed.
Screens. The screens are set in iron frames, which now usually
slide in grooves cast in the mortar, and are
keyed to it, but
were formerly fitted into recesses and bolted. They are made
either of wire-cloth, or of Russian sheet-iron or steel, in which
holes are punched. The sheet-iron is about ^ inch thick and
weighs about 1 Ib. per square foot. The rough (''burred") side of
the punched plate is put on the inside, so that the holes widen
outwards, and are thus prevented from becoming clogged. The
holes are round, oval, or consist of long slots (from
J to J inch
long) ranged parallel or inclined to each other. The shortest dia-
meter of these holes ranges from about -
to -f% inch or more,
according to the nature of the ore and the method employed
in its treatment the usual size is from about
;
QUARTZ CRUSHING IN THE STAMP BATTERY. Ill
The relative advantages of wire-cloth and sheet-iron are not yet-

beyond dispute, and vary with the nature of the ore. Slots
appear to be better suited for discharge than meshes, but, on
the other hand, there is a great loss of discharge area in the use
of punched iron. Thus a wire mesh screen, containing 18 holes
to the linear inch, has 324 holes to the square inch, while a
round-punched sheet-iron screen has only 140 holes of the same
size per square inch. At Otago, N.Z., where friable quartzose
ores are treated, both kinds are used, but the wire-cloth (No. 18)
lasts a little longer, and costs nearly 25 per cent, less than the
round-punched Russia iron. In Gilpin County, Colorado, the
"
" burr-slot screen is used, having horizontal slots placed alter-
nately. The screens are inverted after a time, as the lower part
wears out soonest.
In California, both steel and brass wire, and slot- and needle-
punched tinned- and sheet-iron are used as screen materials, steel
wire screens, however, being seldom used, owing to their rapid cor-
rosion by rusting. The sheet-iron consists of the best soft Russia
iron and has a smooth glossy surface, and the slots and needle-
holes are almost universally burred. The size is named according
to the number of the sewing-machine needle by which the holes
were punched, the sizes numbered 5, 6, 7, 8, and 9 being those
most used. The following table gives details of these screens :
No. of Needle.
Mill, Yictoria, iron punched gratings lasted less than a week,

but, on introducing copper plates containing 100 holes per square
inch, the life of the screen was increased to a month, and
275 tons of ore were passed through it. At Blackhawk, the iron
screens last while from 80 to 430 tons are passed through them,
according to the position of the mill. At Grass Valley, the
average is 200 tons, at Bendigo, 134 tons, and at Otago, N.Z.,
only 40 tons. In California the brass- wire screens last from
10 to 14 days, corresponding to a passage of 120 to 140 tons, and
the Russia-plate lasts from 15 to 40 days, the average being
30 days, corresponding to the passage of about 330 tons. The
rate of wear of the screens depends greatly on their position,
being more rapid with a shallow than a deep discharge, and
more rapid in a narrow than a wide mortar. Pieces of iron
or wood in the ore may cause the screen to break, and these
should, consequently, be removed from the ore as far as possible.
The battery is hung up now and then, so that a thorough inspec-
tion of all the screens may be made, and those that are broken
replaced.
The screens were formerly set vertical, and this system still
prevails generally in Australia. In the United States, however,
they are now placed at an angle which varies somewhat but is
never far from 10, and this has been found to facilitate discharge.
In wet stamping, screens are usually placed on one side of the
mortar only viz., that opposite the feeding side. In cases where
the discharge is required to be as rapid as possible, the screen
area is increased, and double discharge (front and back) mortars
have been made, but have not been used much, except for dry
crushing, and screens at the ends of the mortars are used at
Harriettville, Victoria (Bickard). The area of the screen is from
3 to 4 square feet per battery in California, the height from
the bottom to the top of the screen being from 8 to 10 inches ;
in Gilpin County, the screens are usually 8 inches high and

4^ feet long. Opinions differ as to the necessary amount of
screen area to be used. By some experts it is contended that the
capacity of a battery is really limited, not by its crushing, but
by its discharging power. Thus, by a number of experiments
conducted some years ago at the Metacom Mill, California,
it was shown that when crushed pulp instead of the unbroken
ore was fed into the mortar, the rate of discharge was not
increased. This was taken as a convincing proof that the
discharge area is usually far too low. Nevertheless, double dis-
charge is hardly ever used for wet-crushing mills, the various
objections that are made to it being summarised as follows :
1. Inconvenience is caused in the arrangement of the

copper
plates, both inside and outside the battery.
2. So great a quantity of battery water must be used, that the
pulp is too thin for efficient amalgamation on the plates.

3. The ore does not stay long enough in the battery to be

effectively amalgamated.
When, however, battery amalgamation is not attempted, double
discharge might be advantageous, and, in particular, it is to be
recommended with heavily sulphuretted ores, containing brittle
pyrites. The object to be attained in this case is to leave these
sulphides unbroken as far as possible, so as to facilitate concen-
tration, since great losses from sliming will inevitably result if
they are subjected to repeated blows in the battery before being
discharged.
Order of Fall of Stamps. The order in which the stamps
drop is of considerable importance. If they were let fall
in succession from one end to the other of the mortar, the pulp
would be driven before them, so that the stamp which fell last
would have its die covered by too deep a cushion of ore, while
that at the other end would be almost bare. The result to be
obtained is to keep the ore equally distributed through the
mortar, so that each stamp shall do the same amount of crushing,
although it is inevitable that the middle stamps should be more
efficient than the end ones in discharging the ore. The order
most favoured in California is 1, 4, 2, 5, 3, and that on the Rand
is 1, 3, 5, 2, 4, whilst the orders 1, 5, 2,
4, 3 and 1, 5, 3, 2, 4 are
also often used. Several other orders have their advocates,
and are probably little inferior to the above for the particular
ores on which they are employed. Since the end stamps are of
less efficiency than the others, it has been argued that a
larger
number of stamps in one mortar would be advantageous, and at
Clausthal, in the Hartz Mountains, there are usually from nine
to eleven stamps in a battery,* placed close together, space being
greatly economised in this way. Long and wide experience has,
however, proved that the best number is five.
Feeding. Ore is fed into the battery either by hand or by
automatic machines. It is often asserted that really intelligent
hand-feeding is better than the automatic method, since the stamps
are not all equally efficient. At any rate, hand-feeding is still
persisted in on some Australian gold-fields, although, perhaps,
not always performed with intelligence. The feeder on small
mills is often expected to break down the big pieces of ore with a
sledge hammer, a rock -breaker not being used, but this method of
working may be safely set down as irrational and uneconomical,
and the result usually is that large and small pieces go into the
mortar together. In the United States and in the Transvaal,
self-feeders are almost universally employed in modern mills.
The art of feeding consists in keeping the depth of pulp on the
dies constant throughout the battery, as long as the work is
carried on. This is much better done by automatic machinery
than by hand, and it was found that by the introduction of the
*
Meinecke, Proc. Inst. C. Eng., Session 1891-1892, part ii.
8
former in California the capacity of the stamps was increased by

15 to 20 per cent., while the wear of shoes and dies was decreased
by 25 per cent., and that of the screens by 50 per cent. It is
not difficult to discern the cause of the advantages, for, if the
dies are insufficiently covered with ore, less crushing is done, while
a greater concussion must be taken up by the stamp and by the
die, mortar, &c. If the die is quite bare this concussion is so
great that the stem may be bent or broken, and the shoe and die
battered. On the other hand, if the ore is too deep in the
mortar, there is so thick a cushion that much of the force is
taken up in compressing without crushing it; whilst, besides
the reduction of output, the head, under these circumstances,
sometimes becomes detached from the stem, which is broken or
Fig. 27. Scale, 1 in. = 2 ft. Fig. 27a.
battered by the next blow. The maximum capacity is obtained

with " low feeding," the depth of pulp on the dies being about
2 inches or less.
One advantage of even feeding is that a larger proportion of
gold is caught, owing to the more regular and even flow of the
pulp over the plates, the danger of scouring being diminished.
There are two classes of automatic feeders, each consisting of
a pyramidal hopper with inclined floor, connected with the feed-
opening of the mortar-box by a spout or shoot. Stanford's
automatic feeder, which was the first invented (in California,
about 1870), and is still used, is a typical example of the one
class, and Hendy's Challenge feeder, the best machine devised,
and the most largely used, is a good example of the other class.
Stanford's feeder, shown in Fig. 27, consists of a hopper, A,
with an adjustable spout, B, which is hinged at C. Fig. 27 a

is a front view of B, showing how it is suspended. The spout
is attached to the vertical rod, E, which is
hung to the lever, G,
and this has its fulcrum at D. Near the top of a stamp-stem
(usually the middle one in the battery), the feeding-tappet, F,
is keyed. When the" battery is full of ore, the tappet does not
come down far enough to strike the lever, G, but, when the ore
gets low, the lever is struck,and the result is that the spout, B, is
jerked up and clown again, the ore being thus thrown forward.
Fig. 28.
This machine works well with dry ores which are moderately
fine, but, ifthe ore is wet, and especially if it is argillaceous, it
sticks in the hopper until at last a powerful jerk brings it down
with a run, the mortar-box is filled up, and all the evils of
over-feeding result. The consequence is that the battery is fed
as irregularly as by the worst hand-feeding.
The Challenge feeder (Fig. 28) is constructed so that the
tray, A, below the sheet-iron hopper, B, is revolved in
a hori-
zontal plane by means of a gear-wheel below it, shown in the
figure, and this gears with teeth set in the bottom of the tray,
A.
The gear-wheel is set in motion by a friction grip, D, placed on

the outside of the frame, and actuated through the lever, E, by
the bumper-rod, G, against which the tappet of the stamp strikes.
At each partial rotation a given quantity of ore is scraped off
by the stationary wings or side plates, II, resting on the tray, A.
This amount of ore is regulated by the condition of the mortar.
The machine is especially adapted for very wet or sticky ores.
It weighs, with the frame, about 900 Ibs., and is the most
expensive automatic feeder in the market. Other automatic
feeders, such as the Tulloch and the E oiler machines, constructed
on a similar principle, the ore being scraped and not shaken into-
the mortar, are cheaper than the Challenge, and are doing excel-
lent work on certain kinds of ore. In all cases one feeder is
sufficient for a battery of five stamps, more ore being fed to the
middle stamps, where the most work is done, than to the end ones.
The feeders are now sometimes suspended from above instead of
being supported from below.
Water Supply. The water is supplied to the stamps by
horizontal pipes passing just above the top of the housing of the
mortar-box, with an opening supplied with a stop-cock opposite
each stamp. The water is commonly made to impinge against the
stem of the stamp, either just above or below the wooden casing,
and thence run clown into the mortar, thus keeping the stem clean.
In front of the battery there is also another pipe of about half
the size, to supply water to the tables to help carry off the pulp.
This pipo is often pierced with pin-holes, so that the water is
supplied as a number of fine jets. The water is warmed by
steam in winter in many mills, from a belief that amalgamation
is promoted by warmth. This belief is founded on general
experience rather than on exact experiments, and some results
appear to point to a different conclusion.
Thus, Professor Le Neve Foster obtained the following results*
at the Pestarena Mill in Italy,
during the years 1869-70 :
The average temperature of the water supplied to the mill during

the six summer months was 52 F., and the average temperature of
the water supplied in the winter months was 394 F., and
yet,
in spite of that, and of the fact that the
average temperature, for
instance, of the month of January, 1870, was as low as 33 -6 F.,
he extracted 3-1 per cent, more gold with the cold water than
with the warm. These figures do not necessarily prove that cold
water is better for amalgamation, as there were in this instance
other matters to be taken into consideration, but
they show that
amalgamation is possible even when the temperature of the
water is on an average only 39. The difference in the results in
this case might have been due to the
turbidity of the water
(which was derived from the glaciers at Monte Rosa) in the
summer, and its clearness in winter, or to the fact that the
pyrites were more liable to decompose in warm weather, and so
additional sickening of the mercury was caused in summer.
*
Loc. cit.
The amount of water used varies from 1J to 6J gallons

per stamp per minute, the average in California being about
2J gallons, on the Rand about 5J gallons, and in Colorado
only about 1 j gallons. In California, with fast-running rapid-
discharge batteries, the amount of water per ton of rock crushed
varies from 1,000 to 2,400 gallons, the mean being about 1,700
gallons, while in Colorado the average amount is as high as
2,500 gallons. Besides varying with the method of crushing
adopted, the amount of water varies with the nature of the
gangue, clayey ores requiring more, while the large quantity
required by sulphide ores is due to the deep discharge necessitated
by the difficulty of catching the gold, as well as to the high
density of the pulverised material, which renders it more difficult
to convey in suspension over the plates. As a rule, the more
rapid the output, the less water per ton of ore is required in the
battery. Coarse crushing requires less water in the battery, but,
on the other hand, more has to be added on the plates. The
amount of battery water per ton is increased by over 20 per cent,
by the employment of double-discharge mortars. The amount of
water to be added on the plates varies with their grade, as well
as with the density and size of the particles of crushed ore. It
should be only just enough to prevent the pulp from accumu-
lating on the plates, as any excess over this tends to check
amalgamation and to scour the plates. The average duty of a.
miner's inch in a gold stamp-mill is given by P. M. Randall* as
12 tons of quartz, if the head under which the water is supplied
is 4 inches, and 15-88 tons, if the head is 7 inches. This gives
the proportion of the volume of water to that of ore as ll'l to 1.
This may be compared with the proportion of between 7 and 1Q
to in Siberian placer working, and as much as 30, or even 50
1
to in hydraulicking.
1 It may be mentioned that a ton of 2,000
Ibs. of quartz occupies about 13 cubic feet when unbroken, and
about 20 cubic feet after having been broken up, so that in a
lode a cubic yard contains about 2 tons, and in a tailings heap
only about l tons.
Amalgamated Plates. These plates are usually made of
copper,and are as much as f inch thick for the inside of the
battery and ff to inch thick for the outside. The average
weight used in California is 3 Ibs. per square foot. It was
formerly laid down as a general rule that the heaviest plates,
were the best, as they last longer and are not so easily dented,
but comparatively light plates are now used. The copper should
be of the best quality, and, if it is hard, it must be annealed
before applying the mercury, so as to make it absorbent. This
was formerly done by heating it from below as uniformly as pos-
sible until sawdust laid on the upper surface was ignited. The
plate is then straightened by blows of a wooden mallet, striking
*
Article on Practical Hydraulics in Sixth Report Gal. State Min. t 1886.
a block of wood laid on it, and the surface is carefully cleaned by

scouring with sand or fine emery paper until quite bright, and
washed with strong soda to remove all traces of grease. Cleaning
may also be effected by nitric acid diluted with 9 parts of water,
or by a 2 J per cent, solution of potassium cyanide, rubbed on with
a woollen rag and carefully washed off with water. As soon as
the plate is clean, it is rubbed with a mixture of fine sand, sal-
ammoniac and mercury by means of a brush, the sal-ammoniac
preventing the recommencement of oxidation. More mercury
is sprinkled on and wiped over with a piece of rubber until no
more can be absorbed, the whole surface being now thoroughly
coated and bright. After having been left for an hour, the plate
is finally washed with clean water.
If the plates were used in this condition they would not
catch the gold very well at first, but would continually improve
until they had become coated with gold amalgam. In order to
make them efficient from the start, they are usually coated with
gold amalgam or silver amalgam before being laid down in the
mill. Gold amalgam is most effective but is seldom used, as it is
so much more expensive than silver. The amalgam is rubbed
on with a piece of india-rubber, the plate being wetted with a
solution of sal-ammoniac to keep it bright, and such plates last for
years without further treatment. A more usual method of
preparing the plates in California is to coat them with electro-
deposited silver. This plating is done by certain establishments
in California for most of the mills, but it can be done on the
spot without much difficulty, the plant required being inexpen-
sive. After being silvered, the plates have the mercury applied
to them. They absorb a large amount of mercury, catch gold
well, and are little trouble to keep clean. The plates need not
be re-silvered, except after scraping and sweating (see p. 133),
as they become coated with gold amalgam in the course of time.
About 1 oz. of electro-deposited silver is required per square foot
of copper plate. Silvered plates are not used inside the mortar.
The position of the plates is as follows : The lower edges of
the inside plates are level with the upper surface of the pulp,
when the battery is working properly i.e., they are usually at
1J or 2 inches above the surface of the dies. The plate on the
feed side is generally about 9 to 12 inches wide, and is of the
same length as the battery ; it is bolted to the mortar itself, and
its angle of inclination varies with the shape of the latter, so
that the angle of inclination is sometimes 40, and it is sometimes
nearly vertical. The plate on the discharge side is inclined at
an angle of 10 or 20 to the vertical, and is as wide as the space
below the screen permits, being usually from 3 to 6 inches wide.
It is fixed to a wooden chuck-block, which has its top bevelled
off so as not to obstruct the screen opening. The block is bolted
to the mortar with some thicknesses of blanketing between, in
order to make a tight joint. Several sizes of these chuck-blocks,

with their copper plates attached, are kept in one and the same
mill, a wider block being substituted for a narrower one when,
the wear of the dies has proceeded to a certain extent.
At the Alaska Treadwell Mill, and at one or two mills on the
Rand, the cast-steel linings of the mortars are furnished with
several horizontal slots or recesses for the collection of amalgam,
and these take the place of the back copper plates. The front
or chuck-plate is, however, retained in these mortars. This is
shown in Fig. 45a, Chap. X.
The outside plates are fastened to a wooden table with copper
nails, or wooden clamps, or by wedges driven into the raised
edges of the table. The table is as wide as the battery (4 feet
7 inches) and usually from 6 feet to 8 feet long. In California a
length of 2 or 3 feet of plates of the same width, the apron plates,
are interposed between the battery and the tables proper, on to
which there is a drop of 2 or 3 inches. Below the tables there is
a succession of four or five sluice plates, each about 30 inches long,
with drops of 1 to 3 inches between them. They are usually
made narrower than the others, and are frequently only 12 or
18 inches wide, but this practice is not to be commended, as the
stream of ore and water, forced into a narrower channel, becomes
deeper and flows more rapidly and tumultuously, with the result
that the contact between the ore and amalgamated plates is much
reduced, and very little gold is caught. The use of the drops is
to assist in catching the float gold and to separate the amalgam
which has become floured and mixed with the pulp. In place of
the composite arrangement described above, a single length of
12 or 14 feet of plates is often used, especially in the Transvaal.
To aid in catching the float gold, swinging amalgamated plates
have been introduced, and are in use in the sluices below the
batteries of many Californian mills. They are also used in
hydraulicking. The swinging plate consists of a curved strip
of
silver-plated amalgamated plate about 3
inches deep, and of the
same width as the sluice in which it is hung ; it is suspended
on eyes through which wires pass. The plate thus hangs, half
submerged, with its concave side up-stream, and is kept swinging
by the current, so that all floating particles of gold must come
in
contact with it. It is found in practice that, immediately under
each plate, across the sluice, a line of amalgam which has dropped
from the plateaccumulates. The plates are placed a few feet
apart. They and are very effective.
cost little
Mill Site. This should be easily accessible by road, rail, and
water, if possible ; moreover, it should be near
both wood and
water, and there should be a good fall of the ground.
The least
fall that is considered sufficient in California is 33 feet from the
mouth of the rock-breaker to the floor on which the concentrators
are placed, when rock-breakers are used, followed by stamps,
copper-plated tables, sluice plates,
and two successive concentra-
120 METALLURGY OF GOLD.
tion tables. If a second concentrator is dispensed with, however,

and space otherwise economised as far as possible, 29 J feet may
be enough.
Arrangement of the Mill. The general disposition of the

machinery is shown in section in Fig. 29. This represents a
mill in which the ore is delivered from the ore-cars through a
grizzly on to the rock-breaker floor, and thence by a shoot to the

automatic feeders of the stamp battery ; the pulp, after passing
over the plates, is conveyed by sluices to the double row of
" frue-vanners " on p. 176), which are shown standing
(described
back to back on the lowest floor. The ore-bins should be
of sufficient capacity to hold several weeks' supply for the
stamps, in case of breakdowns or other delays in the mine.
Their floors are made of planking, which is laid with the
lengths in the direction of the slope, for, if placed trans-
versely, the boards wear fast, and the ore packs at the edge
of each one, with the result that its movement is impeded
and must be assisted by shovelling. The slope should be
at least 45 in order to enable the ore to move downwards by
gravity, when the lowest portion is drawn from the shoot. Ore-
bins with flat bottoms have greater capacity, but necessitate an
additional handling of the ore. The sills of the bins should be
placed horizontally on terraced ground, not on the slope of
the hill. Ashoot from the ore-bin door leads to a grizzly,
through which the fine ore drops into the main battery ore-bin.
The larger pieces of rock are discharged either into a coarse ore-
bin or else upon the platform by the side of the rock-breaker and
on a level with its mouth, into which it is shovelled by hand.
The former course is preferable, as in that case the rock-breaker
can be fed continuously by a gate in the coarse ore-bin, which is
opened and shut by a rack and pinion. By this arrangement
there is a saving of labour, but the chief advantage is that the
rock-breaker is thereby kept constantly at work. At the North
Star Mill, California, it was found that when, by arranging for
a continuous feed from the coarse ore-bin down a shoot leading
direct to the rock-breaker, the latter was in constant work, it
absorbed 12 horse-power, as against 8 horse-power when in inter-
mittent work, but its output was over 50 per cent. more. When
the stamp battery is used only to crush the ore, which is subse-
quently treated in pans or other amalgamators, or by concentra-
tion, it is of great advantage to separate the fine product
of
the rock -breaker by sieving, instead of passing the whole through
the stamps. This arrangement increases the output and pre-
vents unnecessary sliming of the ore, thus greatly reducing the-
loss of sulphides when an attempt is made to save these by
concentration. As observed on p. 101, it is becoming cus-
tomary in large mills to place the rock-breakers in separate
buildings.
The product of the rock-breaker is mixed as thoroughly as
possible with the ore, which originally passed through
the grizzly,
and led by means of a shoot direct to the automatic feeders,
which should be mounted on wheels, so as to be readily movable.
in
Plenty of space must be left behind the feeders for convenience
repairing and in the exercise of supervision.
The amalgamating
tables should also be easily accessible, space being left to pass be-
tween them, and the same remark applies to the sluices and the
tables or other appliances for concentration. All shafts, bearings,
<kc.,should also be easily accessible, so that oiling, re-lining, and
repairs may be readily done.
The into a large sluice by which they
tailings are discharged
are carried into a river, or into the sea, or run into settling pits,
or impounded behind dams. One of the two latter courses is
so that it is necessary to use it
adopted, either if water is scarce,
over again, or if the discharge of tailings is forbidden by law, or
if the tailings are rich enough to be subjected to further treat-
ment, at once or at some future time.

The whole of the machinery is contained in a strong building
of timber or corrugated iron, to protect it and the workmen from
the weather, and to prevent theft of amalgam. The interior
should be lighted by as many windows as possible, in order to
facilitate superintendence and repairs. In cold climates, the
mill buildings are sometimes warmed by passing the flues from
the boiler fires through them from end to end, before leading the
products of combustion to the stack, or by steam-pipes.
CHAPTER VII.
AMALGAMATION IN THE STAMP BATTERY.

Treatment of the Plates. In order to keep the plates in proper
condition so that successful amalgamation may be maintained,
they must be prepared carefully, and the closest watch kept over
them. The silver-plated copper table is preferred in California
from the ease with which it is kept clean, but is not used in the
Transvaal. It is not considered desirable to put on it as much
mercury as it will hold, since, if the amalgam is too fluid, losses
are sustained by scouring, but, on the other hand, if the amalgam
becomes too hard and dry from absorption of gold and silver,
further amalgamation is checked and fresh mercury must be
added.
The condition of the inside plates is regulated by the amount
of mercury supplied to the mortar. In Colorado there is an
opening at the front of the battery and above the screen frame,
ordinarily covered by canvas, which can be lifted up by the mill-
man, who introduces his arm, and determines by passing his hand
over the front plate whether the right amount of mercury is being
added by the feeder. The regulation of the addition of mercury
AMALGAMATION IN THE STAMP BATTERY. 123
is thus effected without removing the screen frame. The amount

added varies with the conditions of crushing and the richness
of the ore, but in general from 1 to 2 ounces of mercury are fed
in for every ounce of gold contained in the ore. The finer the
state of division of the gold and the more sulphides there are
contained in the ore, the more mercury is required. It is fed
into the battery at stated intervals of from half an hour to an
hour. In some mills, particularly in Australia, amalgamation in
the battery is not attempted, no inside copper plates being
provided, and under these circumstances it is not usual to feed
mercury into the battery.
The practice of feeding mercury into the battery, although
almost universally pursued in the best mills in the United
States and South Africa, still meets with opposition from certain
experienced millmen. The objections urged against it are mainly
that the mercury so introduced, and the amalgam formed through
its agency, tend to become so excessively subdivided that a
high
percentage is lostthrough flouring; moreover the mercury is liable
to sicken when the ores contain sulphurets. These evils, no doubt,
exist, and tend to increase with the percentage of sulphides present,
while arsenic and antimony in particular cause heavy losses of
both mercury and amalgam if battery amalgamation is attempted,
but with ordinary free-milling ores such losses are not serious.
The amalgamated plates are dressed as frequently as is neces-
sary, the length of time allowed to elapse between two operations
depending partly on the richness of the ore. To dress the
plates, the battery is stopped, and the amalgam wiped oif the
plates with a brush or piece of indiarubber held between two
pieces of wood, so that only about half an inch of it projects ;
and, if necessary, fresh mercury is added. The amount of
mercury put on the plates should be enough to keep their
surfaces in a pasty condition, but not enough to gather into
liquid drops or to run off. The plates are wiped three or four
times a day, or as often as every two hours when rich ores are
being treated, whilst, with exceptionally valuable materials, the
operation may be necessary at intervals of a few minutes only.
Discoloration of the Copper Plates. The plates often
become stained by the formation on them of oxides, carbonates,
or other compounds of copper through the corrosive action of the
water and pulp. Ores containing decomposing sulphides acidify
the water and thus cause the corrosion of the plates, a yellow
film being formed on the surface of the metal. The presence of
carbonic acid in the water is equally harmful, but Aaron points
out that the addition of slaked lime to the water neutralises
the acid substances and diminishes the tarnishing. The yellow,
"
brownish, or greenish discoloration, the so-called verdigris,"
appears in spots and spreads quickly, especially on new
plates, those which have been silver-plated being less liable
to become dirty than the others ; whilst when a plate has become
covered with a thick layer of amalgam it is not readily dis-
coloured. When these stains appear the plate must be at once
cleaned, as the stained part catches little or no gold. The
chemicals used for the purpose are generally sal-ammoniac and
potassic cyanide, the operation being conducted
as follows
:
The battery is stopped, the plates rinsed with clean water, and
a solution of sal-ammoniac applied to the stained parts with a
scrubbing brush, and left covering them for a few minutes in
order to dissolve the oxides. It is then washed off, a solution
of potassic cyanide rubbed on to brighten the plate, and almost
instantly washed off, fresh mercury being then added if neces-
sary. Janin states (Mineral Industry, 1894, p. 332) that long
brushing with potassium cyanide is necessary, as otherwise the
spots reappear when the water is turned on.
Whisk brooms are perhaps better than indiarubber for brush-
ing the plates ; these brooms are cut down to a short length so
as to be stiff enough. The plate is brushed all over and the
amalgam thus thoroughly loosened from it, after which, com-
mencing at the top where the amalgam is thickest, it is sub-
jected to a systematic stiff brushing, each stroke being directed
longitudinally down the table, and not towards the centre. The
surplus amalgam is thus brushed to the lower end of the plate,
whence it is removed, and a thin coating of amalgam is left over
the whole surface of the plate, excluding the air and preventing-
the formation of " verdigris."
Chemicals Used to Promote Amalgamation. The use of
chemicals to aid amalgamation was formerly much more general
than at present, although battery-men were less afflicted by the
rage for nostrums than pan-amalgamators. Almost the only
chemical now used on a battery in California or Colorado,
both to promote amalgamation and to clean the plates, is cyanide
of potassium, and the use even of this reagent is becoming less
general every year. A dilute solution is believed to promote amal-
gamation, but probably its action consists merely in thoroughly
cleaning the plates and the mercury from all trace of oil, grease,
and base metallic oxides. 1 or 2 Ibs. of potassic cyanide should
be enough to supply a 40-stamp battery for twelve months
when treating free-milling ores, by which the mercury is not
much affected. The difference between such a mill and one
running on base ores may be judged from the fact that at the
Hidden Treasure Mill, Colorado, where there are seventy-five
stamps, 260 Ibs. of cyanide were used in a year. Here the plates
are dressed every twelve hours with a weak solution containing
2 oz. of cyanide in 3 gallons of water, the operation being neces-
sitated by the acidity of the water which comes from the mine r
and is further contaminated by sulphates formed in the ore.
Sodium is now used mainly to clean the mercury, after it has.
been retorted, before again using it in the battery. A weak

solution of caustic soda is used to remove grease from the plates.
The merest trace of any kind of grease or oil is very prejudi-
cial to successful amalgamation, forming a film over the plates
and over the little globules of mercury, and thus preventing
contact between them and the particles of gold. Ammonia or
potassic cyanide removes the film, but a saturated solution of
wood-ashes, or of soda carbonate or caustic soda is more gener-
ally used than these reagents. Grease may consist of the tallow
dropped from the miner's candles, or the oil from the loose steam
which is sometimes used to warm the battery feed- water, or from
the bearings of the machinery.
Mercury. The mercury used for the battery and the plates
should be quite free from base metals, such as copper, zinc, lead,
or tin. If these are dissolved in the mercury they become
rapidly oxidised and soon cause it to "sicken," so that it
breaks up into a number of minute globules, each coated with
a layer of base metallic oxide. These globules are not only
useless for amalgamation, but as they are very minute and refuse
to coalesce, they are carried away in the tailings, together with
such gold as may have been taken up before the oxidation had
proceeded far. Such impure mercury, when used to coat the
plates, also causes their discoloration by oxidation of the dis-
solved base metals. Mercury acts best when it already contains
gold and silver, and it is customary to dissolve some silver in
the new mercury applied at the starting of a mill, if no old
mercury can be had to mix with it. When work has been some
time in progress there is no difficulty on this account, as the
quicksilver, which has been squeezed through a filter to separate
the amalgam, carries some dissolved silver and gold with it.
Grade of Plates. The grade or inclination of the plates
varies with the nature of the ore to be treated, heavy pyritic ores
requiring a higher grade than light quartz, while the coarser the
crushing the steeper must be the grade. In California the
copper tables have an inclination of from 1 to 2 inches per
foot, the apron plates from to 1} inches per foot, with an
average of 1 J inches, and the sluice plates from 1J to 1J inches.
With heavily sulphuretted ores a grade of from 2 to 2 J inches per
foot is used. The steepness of the grade is of great importance, as
on it, and on the amount of water supplied, the attainment of the
necessary contact between the ore and the plate depends. When
the pulp is flowing properly, it travels down in a series of little
waves and ripples, and, in consequence of the friction between
the plate and the film of water in contact with it, the upper
portions of these little waves travel faster than the lower parts,
so that the motion becomes one of tumbling over and over. As
a result of this, if the plate is long enough, every particle of
pulp comes in contact with the amalgamated surface, and the
perfect extraction of amalgamable gold, mercury

and amalgam is
obtained.
Muntz Metal Plates. The use of Muntz metal (which con-
sists of copper 60 per cent., zinc 40 per cent.) for amalgamated
plates is of great interest. It differs from copper in catching

to be
gold well as soon as the plate is amalgamated, not requiring
covered with gold- or silver-amalgam before it begins to do good
work. Moreover, the amalgamated surface is very superficial,
since the mercury does not sink in so far as it does into a plate
composed of pure copper, so that only a small quantity of
mercury is required to cover it. The result is that cleaning
up is easy and rapid, no iron instrument being necessary, but
rubber being always sufficient. These properties make it par-
ticularly valuable for custom mills, where it
is desirable to catch
as much as possible without mixing the amalgam obtained from

two parcels of ore crushed in succession. On the other hand, as
it holds little mercury, it cannot absorb much gold, and must be
cleaned-up at frequent intervals.

The mercury on Muntz metal plates does not suffer so easily
from "sickening" as that on copper plates it has been suggested
;
that this is due to the electrolytic action of the copper-zinc

couple, which sets free nascent hydrogen, and so reduces the
compounds of mercury and other metals which have been formed.
It follows that Muntz metal plates are preferable for ores con-
taining large amounts of heavy sulphides or arsenides. The
" "
greenish-yellow stains (called verdigris by millmen) which are
formed on copper plates when grease and other impurities are
present in the battery water, do not appear when Muntz metal is
used, and such discolorations as occur on these plates can be
better removed by dilute sulphuric acid than by potassium
cyanide. At the Saxon Mill, New Zealand, the copper plates
formerly required 7 Ibs. of cyanide, costing 23s., per month to
keep them in order, while the Muntz metal plates, by which
they were replaced, could be kept clean by 5 Ibs. of sulphuric
acid per month, the cost being 3s. 4d. It is stated, however,
that, in the treatment of highly acid ores, which have been
weathered for some time so that they contain large quantities
of soluble sulphates, or in cases where the battery water contains
acids, copper plates are less affected than Muntz metal, over
which a scum is rapidly formed. This is not the experience in
the Thames Yalley, N.Z., where Muntz metal is preferred in
spite of the extremely acid nature of the water and ore.
Generally, it may be stated that Muntz metal plates are
cheaper, wear better, and require less attention than copper. In
dressing new Muntz metal plates the following method is
adopted in New Zealand : The surface of the plate is scoured
with fine, clean sand then it is rinsed with water, and washed
;
with a dilute (1 to 6) solution of sulphuric acid. Mercury is

then applied and rubbed in with a flannel mop until it wets the
surface of the plate (i.e., amalgamates with it) in one or more
places, after which the mop is given a circular movement, passing
through these spots, until the amalgamation of the surface
spreads from them over the whole plate.
The discoloration of the Muntz metal plates is prevented by
the weak electric current produced, as has been already stated.
The same effect can, according to Aaron, be obtained when
ordinary copper plates are in use, by placing them in contact
with iron or some other metal which is positive to copper. Strips
of iron bolted to the top and sides of the plate are said to be suffi-
cient tor the purpose, the copper being in that case unaffected by
the acidity of the water, which causes oxidation and dissolution of
the iron only. Jauin's experience does not support these views.
Shaking Copper Plates. A shaking copper plate is recom-
mended by some of the best authorities to be used either below
or in place of the ordinary amalgamating tables, especially in
cases where these do not appear to give good results. An
ordinary fixed copper plate requires an inclination of from 1^ to
2 inches per foot, in order to keep it clear of sand, when the plate
is of the same width as the If, however, the plate is
battery.
subjected to a short rapid shake, the sand is kept from packing,
and amalgamation is well performed with a grade of only |- to J
inch per foot, or the amount of water needed with the pulp may
be greatly reduced and better contact thus obtained. For these
plates, silver-plated copper is the material employed. They are
affixed to a light wooden frame which is moved by a crank-shaft,
revolving 180 to 200 times per minute, placed on one side, with
a throw of 1 inch at right angles to the direction of the flow of
the pulp. In some mills, a longitudinal shake is given to the
plate instead of this side shake. The frame may be hung on
rods from above, but is more conveniently supported on four
short iron springs, forming rocking legs. The width of the tables
should be made as great as possible, while the length is of less
importance, as, the thinner the current of pulp flowing over them,
the better the chance of the gold particles coining in contact
with the plates and being retained. These shaking plates were
first used in Montana in 1878, and have since been
employed
at several Californian mills, giving satisfactory results. It is
advantageous to add to them an amalgam- and mercury-saver.
A simple device for this purpose is to nail a strip of wood,
half an inch thick, across the copper plate near the top, thus
forming a shallow riffle, the angle of which is soon filled with
sulphides and coarse sand, which are kept in agitation by the

movement of the table. This is stated by W. M'Dermott and
P. W. Duffield* to be the most effective contrivance yet devised
for catching quicksilver and hard amalgam. If the inside copper
*
Gold Amalgamation, London and New York, 1890, p. 16.
plate should become hard by accident or neglect, chips of amal-

gam escaping through the screens are retained in this riffle, and,
becoming spherical by rolling up and down under the effect of
the shaking motion, increase in size just as a snowball does when
rolled in snow.
Corrugated. Plates. Another device to increase the catching
power of copper plates is to make them in a corrugated form so
that they will hold little pools of mercury. Where much float
gold is believed to be passing off, this is supplemented by the
addition of a second plate placed above and facing the other with
its amalgamated surface downwards. The pulp is allowed just
sufficient room to pass between them, in contact with both; and
there is little doubt that in this way some fine gold is caught
that would otherwise escape.
Mercury Wells. Another method of saving mercury and
amalgam, which would otherwise be lost in the tailings, consists
in the application of mercury wells or riffles. A mercury well
or riffle consists of a shallow gutter filled with mercury, over the
surface of which the pulp flows or through which it is forced to
pass by suitable machinery. Attwood's amalgamator, formerly
much used in California, was a machine of the latter class.
Such wells are even now often used in modern mills. It is
believed by some of the best authorities that such devices catch
no more mercury than could be secured by either well-arranged
stationary or shaking plates, while the practice of placing a well
between the screens and the amalgamating tables is especially to
be condemned, as it prevents proper supervision being kept over
the feeding of mercury into the battery, over-feeding being
difficult to detect under these circumstances.
Galvanic Action in Amalgamation. In amalgamation in
the mortar, on plates, or in pans, not only are free metals
absorbed, but the dissolved salts, and, to a less extent, the
insoluble compounds of the heavy metals, are reduced and
amalgamated, chiefly by galvanic action. The copper of the
plates, or the iron of the mortar or pan, constitutes the positive
element, and all metals less oxidisable than this reacting metal
are reduced by it, and are then amalgamated by the mercury.
In this way iron reduces both lead and copper, although, if these
are present in the form of undecomposed sulphides, this action
will be very slight. Now, if lead is introduced into the amalgam,
the latter becomes pasty, and is subjected to considerable losses,
and copper has an equally harmful effect. It is for this reason that
the arrastra is found to be better than the stamp battery or even
the pan for certain plumbiferous ores. This action of iron is of
course enormously increased if the ores are subjected to a
chloridising roast before being amalgamated, as in the Reese
River process for the treatment of auriferous silver ores.
In a number of mills, this galvanic action is increased by the
passage of a weak electric current through the charge by means

of a dynamo. The amalgamated plates or the walls of the pan
are connected with the negative pole, while the positive pole is
formed of a plate of lead or iron dipping into the pulp. Under
these conditions the mercury is still further protected from
attack, and remains bright and lively, but the deposition of base
metals in it is favoured, and the stronger the current the more
this action is induced. Consequently, such methods are attended
with the best results when dealing with ores containing little or
no copper, lead, &c., since in these cases the strength of current
can be increased, and the mercury kept clean, without any ill
effects. The principle is made use of in Bazin's centrifugal
amalgamator, Molloy's hydrogen amalgamator, and other similar
machines.
Designolle Process of Amalgamation.* In this process a
solution of mercuric chloride is used. It was tried at the Haile
Mine, South Carolina, the method being as follows Charges of
:
600 Ibs. of roasted ore were placed in cast-iron barrels with

1,000 Ibs. of cast-iron balls, or in pans. The barrel was partly
filled with water, and 1 gallon was added of a solution containing
1-7 per cent, of mercuric chloride, the same amount of hydro-
chloric acid, and twice as much salt, if the ore contained less
than 15 dwts. of gold per ton. This would be equivalent to
about 10 parts of mercury to 1 part of gold. The barrel was
rotated for twenty minutes and then discharged into a settler,
and the suspended amalgam caught on copper plates. The
mercury was supposed to be reduced by the iron thus
HgCl 2 + Fe = FeCl 2 + Hg,
and the metallic mercury thus freed amalgamated with the gold.
If no common salt was present, some mercurous chloride was
formed according to the equation
2HgCl 2 + Fe = Hg 2 Cl 2 + FeCl 2 ,
and the subsequent reduction of the insoluble calomel by iron

was not complete. Hydrochloric acid was supposed to hasten
the amalgamation by setting up some electrolytic action.
The total cost of the process at the Haile Mine was said to be
only 35 cents, or Is. 7Jd., per ton, but it was abandoned when
the percentage of iron in the material under treatment increased,
owing to improved methods of concentration. Large quantities
of oxide of iron were then amalgamated, and rendered the
resulting mass harder to treat than the ore itself. By repeated
washing, settling, and regrinding with fresh mercury, it could
be partially purified, but not without a loss of gold. It is stated
that 87 per cent, of the gold in the ore was extracted.
*
This account is abridged from that given by Louis Janin, Junr., in
Mineral Industry in 1694, p. 346.
130 THE METALLURGY CF GOLD.
The Clean-up. The amalgam, both on the inside and outside

plates, does not accumulate evenly, but in ridges and knots
which serve as nuclei for the collection of more. It is not
advisable to allow the coating of amalgam to become very thick,
since, although the plates catch better as the amalgam accumu-
lates, losses experienced by scouring. The wiping of the
may be
outside plates usually takes from 10 to 15 minutes for each
battery. The amalgam so obtained is ground with more mercury
in a clean-up pan in order to soften it, the skimmings from the
mercury wells, &c., being added to the charge. The inside plates
are not wiped until the amalgam stands up in ridges on it ; the
operation may be necessary as often as twice a week, but it
usually takes place twice a month, when a general clean-up
is made. All amalgam, however obtained, unless it is already
hard and dry, is usually at once separated from the excess of
mercury contained in it by being squeezed in filter-bags, and the
pasty residue alone passed to the clean-up pan or the retort.
In cleaning-up, the stamps are hung up, two batteries at a time,
the screens, inside plates, and dies are all taken out, and the
"
headings," or contents of the mortar, consisting of the pulp,
mercury, sulphides, and pieces of iron and steel, amounting in all
to a quantity sufficient to fill two or three buckets, are carefully
scraped out and fed into the mortar of one of the other batteries,
which has not yet been cleaned up. In California, the headings
from the last batteries are panned, the iron removed by a magnet,
and the remainder ground with mercury in the clean-up pan
In Gilpin County, however, it is not panned, but merely
returned to the battery on restarting. Amalgam is found
adhering to the inside of the mortar and to the dies, and is
carefully detached and added to the clean-up pan, and all the
plates are well scraped with a sharp-edged piece of hard rubber,
care being taken not to scratch them. The plates are then
redressed and put back, the batteries restarted, and the next ones
stopped and cleaned up. Three men can clean up forty stamps
in from five to seven hours, ten stamps being thus idle for the
whole of this time.
The amalgam obtained from the batteries, outside plates,
mercury wells, or sluices, is rarely clean enough for immediate
retorting ; it is usually found to contain mixed with it grains
of sand, pyrites, magnetite and other minerals, together with
fragments of iron and other foreign substances. The skim-
mings from mercury wells are still more impure. These materials
must be purified by grinding with fresh mercury and washing
before they can be passed to the retort. The scraps of iron
consist of fragments of shoes, dies, shovels, picks, hammers, and
drills: they are knocked about in the mortar until a
quantity
of gold and amalgam has been driven into their interstices. At
the Jefferson Mill, Yuba County, California, about ^ ton of such
scrap, picked out by hand or by a magnet, had accumulated in

1885. It was attacked by warm dilute sulphuric acid until the
surface had all been dissolved off, and the residue was then well
washed, and gold to the value of $3,000 thus recovered. The
shoes, dies, &c., which were too large for this treatment, were
boiled in water for half an hour, and then struck by a hammer,
when the gold dropped out.* In Colorado, and in small mills
generally, the dirty amalgam is ground in a mortar by hand with
fresh mercury and hot water, until it is reduced to an even thin
consistency, when the dirty water is poured off, and the mercury
poured backwards and forwards from one clean porcelain basin to
another until the pyrites, dirt, fcc., have risen to the surface, when
they are skimmed off. The clean mercury is then squeezed
through canvas or wash-leather, when the greater part of the gold
and silver contained in it, together with about one and a-half
times its weight in mercury, remains in the bag, the rest of the
mercury, with a small quantity of the precious metals dissolved
in it, passing through. The skimmings obtained are put back
into the mortar, and re-ground by themselves with fresh mercury.
In large mills a clean-up revolving barrel is often employed to
mix the amalgam. This is made of iron, and is similar in con-
struction to chlorinating barrels, but without the lead lining.
At the Plumas Eureka Mill, the barrel is 3 feet in diameter and
4 feet long, and revolves twenty times a minute ; the charge is
700 Ibs. of amalgam and 20 Ibs. of mercury, or more if the
amalgam is very rich. A
dozen or more pieces of iron, such as
worn out battery shoe shanks, are put into the barrel, which is
filled nearly up to the top with water, and then revolved for
from six to twelve hours. The use of the iron is to help to mix
the amalgam and mercury, but it causes some loss by flouring,
and is omitted with advantage in Australian mills. The barrel
is then opened and washed out with water, the tailings being
run over amalgamated plates and through a mercury well or
some other form of amalgam-saver, after which the amalgam is
scooped out of the barrel and squeezed again in wash leather.
The Clean-up Pan is even more extensively used than the
barrel. One of the oldest in use in the United States, the
Knox Pan, is still a gre it favourite, especially for treating
battery sands, skimmings, &c. It consists (Fig. 30) of an iron
pan, 5 feet in diameter and 14 inches deep. Wooden or iron
shoes are attached to the arms, ,7, which make from twelve to
fourteen revolutions per minute. Iron shoes are considered
better for brightening or polishing the particles of gold con-
tained in the pyrites, and so rendering them fit for amalgama-
tion. The charge for this pan is about 300 Ibs. of concentrates,
skimmings, &c. Concentrates are treated in this way only when
the quantities of them are small, while at the same time they
*
Sixth Report Gal. State Min., 1886.
contain mercury and amalgam. The charge is made into a pulp

with water and ground for three or four hours, after which 50 Ibs.
of mercury are added, and mixing is carried on for a^few hours
longer, before the pulp is diluted, settled, and discharged. The
tailings suspended in the water are often caught in settling pits,
and either sold or subjected to further treatment on the mill, as
they are frequently of high value. When the Knox pan is
required for mixing dirty or impure amalgam with additional
mercury, the wooden shoes are used, as shown in the figure.
The waste matter is then washed off with a stream of water as
before, and the amalgam retorted.
Fig. 30.
The position from which the greater part of the gold is

obtained in a clean-up varies according to the ore and the
method of treatment. In California from 50 to 80 per cent, of
the gold saved is caught on the single
plate inside the battery,
the remainder being caught on the outside plates and the sluice
plates, or being contained in the concentrates. In the Grass
Valley, at the Original Empire and the North Star Mills, from
70 to 85 per cent, is caught inside the battery. The
amalgam
from the battery plates is usually richer than that from the
outside plates, especially if the gold is coarse. The reason for
this is that coarse gold, being
easily amalgamable, is almost all
caught on the inside plates, while fine gold, even if amalgamated
in the battery, forms a more fluid
amalgam which passes through
the screens and is caught outside. Coarse gold forms a richer

and amalgam than fine gold, because it is more easily-
stiffer
taken up by amalgam which is already almost saturated with
gold. At the two mills last named the value of the plate
amalgam is $4.50 per oz., and that of the battery amalgam
$8.50 per oz.
About every three or six months the plates are scraped with
a sharp iron chisel or palette knife, or scaled by chipping with a
hammer and chisel. At the North Star Mill, California, the
plates are immersed in boiling water so as to soften the amalgam
before they are scraped. They are sometimes "sweated" in
California by heating them over a wood fire before scraping
them. At the Empire Mill, Grass Valley, California, the sweat-
ing of the outside and apron plates of four batteries produced
bullion to the value of $19,000.* These plates had been down
for eighteen months, and the ore which had been run over them
averaged 18 dwts. of free gold per ton. After scaling and
sweating, the plates may require replating. In course of time
they are worn out, the copper becoming brittle and worn into
holes, but they usually contain gold enough when discarded to
pay for a new set.
Several methods are in use for recovering the gold from old
plates. For example, they may be dissolved in nitric acid,
when the gold is left nearly pure. A
more economical method
of detaching the gold, much used in Australia, is described by
Mr. M'Cutcheon as follows The plates are placed on the-
:
hearth of a reverberatory furnace, or on a fire made with logs

in the open air, and the mercury expelled at a gentle heat. If
the temperature is too high, the gold sinks into the copper at
once, and the copper must then be dissolved. After the mer-
cury has been driven off, the plate appears to be more or less
coated with gold on one side. This surface is treated with
hydrochloric acid for eight or ten hours, and the plate is then
replaced on the hearth and exposed to a dull red heat until well
blackened. On plunging it into cold water, the gold now scales
and is collected and freed from copper by boiling in nitric acid.
off,
Retorting. The solid amalgam, which is retained in the

canvas or wash-leather filters, usually contains from 30 to 45 per
cent, of gold and silver, according to the state of division of
the gold present in the ore, and also to the degree of care exer-
cised in squeezing out the excess of mercury. For separating the
gold from the mercury there are two kinds of retorts in general
use the pot-sliaped retort, which is sometimes cast with trun-
nions to swing on supports, in small mills and the cylindrical
;
retort, shown in Fig. 31, in larger mills. The pasty amalgam

is rolled up into balls or kneaded into cakes, and squeezed
into the pot -shaped retort, and often rammed down with a
*
Eiyhth Report Col. Stale Min., 1888, p. 714.
bolt-head, although this course is deprecated by some managers,

who prefer to leave the amalgam as spongy and open in
texture as possible, believing that a uniform product is thus
obtained more rapidly at a lower temperature and without so
much loss. In the horizontal retort the amalgam is placed in
iron trays divided into compartments by partitions. In either
case, the retorted metal is prevented from adhering to the iron,
either by laying it on three or four thicknesses of paper, the
ashes of which remain beneath the amalgam, or by covering the
iron trays with a coating of whitewash. The former plan is
preferred. The mercury is condensed in cooling tubes passed
through water ; the loss through volatilisation is usually very
Fig. 31.
small, and may be taken as being about one grain of gold per
pound of mercury.
The charge is heated slowly until the boiling point of mercury
is reached, when the fire is checked, and the retort kept at an
even temperature for one or two hours, or until the bulk of the
mercury has been driven off. The retort is then raised gradually
to a bright red heat to expel the remainder, and after cooling
it is opened, the trays are withdrawn, and the retorted metal is
loosened by a chisel, if necessary, and turned out on a table.
In retorting amalgam containing considerable quantities of
base materials, there is a danger of the vent being choked up by
condensation of solid material. The retort should be so arranged
that a rod can be passed through the condensing pipe so as to

clear it of obstructions, if
necessary. The front of the retort is
luted on carefully with chalk or wood-ashes and salt, and firmly
clamped so as to be quite tight, otherwise a loss of mercury
is incurred. In all retorts the lid is turned and ground so. as to
fib on perfectly. The condensing pipe should not have an open
end dipping freely into water, as in that case a sudden cooling
of the retort would cause the water to be sucked in, and an
explosion would occur. The open end of the pipe may be
enclosed in a rubber bag immersed in water, in which case the
amount of distention of the bag will indicate the progress of the
operation.
The pot-shaped retort requires no brick fittings, and can be
heated over an assay furnace or forge fire, or in a fire built on
the ground, when it is placed on a tripod stand. In the latter
case the fire is lit at the top and burns slowly downwards. The
pot-shaped retort is not filled to more than two- thirds its capacity,
and must be heated very gradually at first.
The retorted metal is porous and from 500 to 950 fine in gold,
the remainder being in general chiefly silver, with base metals
and sulphides in smaller quantities. The gold is melted in
crucibles with carbonate of soda and borax, and suffers a further
loss in weight, due to the volatilisation of a small quantity of
mercury, which is obstinately retained until the melting takes
place. This amount is usually not more than from 0*5 to
1 part
per 1,000 of the retorted metal, if the retorting has been
carefully performed.
Loss of Mercury. The loss of mercury is due to two causes,
" minute mechanical sub-division, due to excessive
flouring," or
"
stamping or grinding, and sickening," or extreme sub-division
caused by chemical means. In the latter case, a coating of some
impurity is formed over the minute globules of mercury, which
are thereby prevented from coalescing, from taking up gold
and silver, or from being caught by the plates and wells, as the
coating prevents all contact between the mercury and other
bodies. The impurity may be an oxide, sulphate, sulphide, or
arsenide of some base metal, either originally present in the
mercury, or taken up from the ore by it ; occasionally the
mercury itself may be partly converted into a sulphate or other
salt, although this latter condition is not common.
The employ-
ment of mercury, which contains no base metals dissolved in it,
will reduce the loss due to sickening, but such pure mercury
is not always obtainable (except by very careful distillation, in
which the first and last portions condensed are rejected), and it
soon takes up fresh impurities when used with sulphuretted
ores. The base metals usually present in mercury are rapidly
oxidised in the air, especially in contact with water ; the oxida-
tion is made much more rapid by the presence of any acid in the
water, and this acidity (due to the presence of acid sulphates

from decomposing pyrites) is rarely quite absent from battery and
mine waters. The metallic oxides thus formed are not soluble in.
mercury, and they float on its surface in the form of little black
scales, which soon form a coating. A surface coating may also
be formed of grease. The use of sodium amalgam in preventing
the sickening of mercury is noticed below (p. 137). One of the
impurities in mercury most to be feared is lead, as the amalgam of
this metal tends to separate out of the bath of mercury in which
it is dissolved. According to Prof. J. Cosmo Newbery, it rises
to the surface by degrees, taking with it any gold amalgam that
may have been formed, and floats as a frothy scum, coating the
mercury and preventing any further action by it, whilst it is
readily powdered and carried away in suspension by a current of
water flowing over it, so that the gold contained in it is lost.
Sickening of the mercury is also promoted by base minerals
present in the ore. Most gangues, except heavy spar, hydrous
silicates, &c., have no action on the quicksilver ; even clean
cubical iron pyrites, and other iron and copper pyrites, if.
they are undecomposed, are harmless, although the materials
last named cause sickening when partly decomposed. The
other sulphides are all more or less harmful, their action
being, however, much less energetic than the compounds of
arsenic and antimony. J. Cosmo Newbery conducted a number
of experiments in Australia many years ago, to determine the
action of some of the base metals on mercury, and found that
compounds of arsenic and antimony are particularly harmful, and
that if gold containing metallic arsenic is amalgamated, the result-
ing amalgam is black and powdery, and floats on the mercury,
being coated by black metallic arsenic, which separates out and
refuses to unite with mercury. Arsenical pyrites seems to act
in the same way as metallic arsenic, a large amount of black
sickened mercury being produced by it, the action being especially
energetic if the pyrites is partly decomposed. The black coating
is, in this case, a mixture of pyrites, arsenic, and mercury, in a
very finely divided state. Sodium amalgam acts beneficially

when arsenic is causing loss of mercury.
Sulphide of antimony breaks the mercury into black powder
even more quickly than arsenic, some sulphide of mercury being
formed if there is any trituration, whilst the antimony forms an
amalgam. The action of sodium amalgam on this mixture is of
no avail, as sodic sulphide is formed, more antimony amalgam
produced, and sulphuretted hydrogen set free, the results on
the amalgamation of the gold being very disastrous. Bismuth
sulphide acts similarly, but with less rapidity.
Floured mercury is perfectly white in appearance, like flour,
sickened mercury, as already stated, being blackish. If this
floured mercury is examined with a lens, it is seen to consist of
a number of minute particles many of them microscopic each

of which is perfectly bright and pure, shining like a mirror.
They are prevented from coming into contact and coalescing
by being surrounded by films of air, concerning which Prof.
" If
Huntington has remarked you scratch the film of air, the
particles run together and form one globule." Floured mercury
is readily carried away and lost in the tailings, but if
passed
through and agitated with a large body of clean mercury much
of it is at once absorbed in the mass. The loss through flouring
is experienced in the milling of refractory and
free-milling ores
alike.
In California the total loss of mercury varies from i to 1 oz. of
mercury per ton of ore crushed, the mean being about ^ oz. per
ton. Most of the mercury is lost as such and not in the form
of amalgam, as is proved by the fact that where the largest
proportion of mercury is fed into the battery the greatest loss,
takes place but the highest percentage of gold is recovered.
Thus, at the North Star and Empire Mills the greatest loss in the
State occurs, 1 oz. of mercury being lost per ton, but over 90 per
cent, of the gold is extracted. In the Blackhawk Mills, Colorado,
where base ores are crushed, containing from 12 to 20 percent, of
pyrites, the loss of mercury is from i to ^ oz. of mercury per ton.
The mills in which the greatest loss of mercury occurs have the
deepest discharge, the ore and mercury being in these cases
pounded together for a greater length of time before being
ejected from the mortar, so that more flouring takes place.
Many suggestions have been made at various times for the
reduction of the loss of mercury. The use of certain chemicals
in keeping it clean and lively and in neutralising the bad effect of
base minerals has already been noticed. Sodium amalgam was
suggested by Prof. Wm .
Crookes, F.R.S., many years ago as
likely to effect this purpose. It is prepared by heating a basin
or iron flask of mercury to about 300 F., and dropping in little
pieces of sodium not larger than a pea, one by one. Each
addition causes a slight explosion and a bright flash of flame.
The sodium may be added with less loss and less danger to the
operator if the mercury is kept at a somewhat lower temperature
and the sodium stirred into the mercury with an iron pestle or
pressed below its surface with a spatula. When about 3 per
cent, of sodium has been added to the mercury, the reaction
becomes less active and the amalgam is then poured out upon a,
slab or shallow dish, allowed to cool and solidify, and then
broken up and kept in stoppered bottles under naphtha. When
it is necessary to revivify a quantity of mercury, a few small
pieces of the amalgam are added to it and stirred in, or are

previously dissolved in clean mercury before being added to the
impure stuff. The strong affinity of sodium for oxygen enables
it to reduce the oxides of the base metals which are coating the
i38 THE METALLURGY OF GOLD.
mercury globules, and as sodic oxide is very soluble in water it

is at once removed in solution, neutralising part of the
acidity of
the water at the same time. The base metals are redissolved by
the mercury which is then in good condition to take up the
precious metals or to be caught on amalgamated surfaces or in
riffles. Sodium amalgam is not much used except in amal-
gamating-pans or in mercury wells or riffles i.e., wherever large
bodies of mercury can be directly acted on by it. It is of
comparatively small value when added to the mortar of a
stamp battery, although this use of it is not unknown.
Loss of Gold. The losses of gold in amalgamation may be
ranged under the following heads :
1. Loss of free
gold or amalgam due to a want of proper care
in amalgamation.
2. Loss of gold or sulphurets imbedded in particles of rock.
3. Loss of gold which "floats" in water and is carried away
with the slimes.
4. Loss of gold which is not in a condition to be
directly
amalgamated.
The latter heading may be subdivided into two, viz. :
(a) Loss of gold contained in sulphurets.

(6) Loss of free gold, which is prevented from being amal-
gamated by being coated with a film of some mineral ("rusty"
gold), or with grease, or by being in an unsuitable physical
condition (hammered gold).
In the preparation of gold ores for amalgamation every care
must be taken that the course best suited to each particular case
is being pursued. In some instances, which are not common,
the whole of the gold may be present in a form in which it can
be directly amalgamated. In general, however, the gold is pre-
sent in two or more forms, one capable and the others not
capable of amalgamation. In such cases there is no reason to
be dissatisfied with the action of an amalgamating machine if it
extracts a high percentage of the free gold, even though the
total extraction obtained by it is comparatively low. It has
frequently been declared that the greater part of the loss of
free gold, which really takes place, is not due so much to the
imperfection of the various amalgamating machines which are
now at the disposal of the metallurgist, as to the lack of care
and facility of resource on the part of the men in -charge of them.
It is probable that no two ores can be treated to the best advan-
tage under exactly the same conditions. A millman experienced
in the treatment of the ores of one district may be quite at fault
when attempting to amalgamate an ore unlike those to which he
has been accustomed. A silver mill, in particular, has been pro-
nounced to be the worst possible school for a gold amalgamator,
whose work must be closer in proportion as his amalgam is richer
than that obtained from silver ores. At the risk of recapitulating
much that has been already said, it may be worth while to direct
the student's attention to a few general rules which can be
applied in many cases. In the first place the stamps and
screens must be such as are calculated to produce the largest
possible output, without rendering the pulp unsuitable for the
processes of amalgamation or of concentration, or both, which
are to follow. The ideal crushing would be to " crack the nut
and leave the kernel entire," or in other words, to liberate the
particles of gold without breaking them. It was formerly be-
lieved that a light stamp working rapidly was best adapted for
this, and even quite recently light fast stamps, with narrow
heads, only 4 inches across, have been re-introduced experimen-
tally. On the whole, however, the tendency is in favour of
heavy stamps working fast with a low drop, as producing the
maximum output with the minimum amount of slimes. In con-
nection with this, the small production of slimes by the steam
stamp (described on p. 149) should be noted.
The subject of delivery is closely connected with that of
crushing and must be considered at the same time. The screens
are not usually placed quite close to the level of the pulp in the
mortar on account of the rapid wear caused by the violent pro-
jection of pulp against them when in that position. Their
height above the dies is varied according to the ore, the
delivery being slower in proportion to this depth of discharge.
The best size for the mesh of the screens must be determined by
direct experiment. It has often been contended that, as the
crushing must be fine enough to liberate the particles of free gold
from their matrix, therefore the size of the screens depends on
the state of division of the precious metal in the ore. Never-
theless it does not follow that an ore must be reduced so that
the gangue is as fine as the gold particles contained in it,
although this is sometimes erroneously assumed. On the con-
trary, even when the gold is of extreme fineness, coarse crushing
through 20 or 30-mesh screens, may be the best practical method
to adopt, since otherwise the sliming of the ore may cause the
loss of valuable mineral which should be obtainable by concen-
tration, besides reducing the yield on the plates. In the course
of the crushing, much of the ore will have been reduced to a
comparatively fine state of division, and probably this portion
will be found after amalgamation to contain but little gold; from
this, the coarser material, in which the gold is still locked up,
" "
may be separated by sizing in pointed boxes (see p. 170) and
reground in an amalgamating mill or pan. If, on the other hand,
the slimes are found to be as rich as the coarse sand, it is obvious
that no finer crushing is required, as the output would be
thereby diminished without any corresponding increase in the
yield per ton. If the slimes, after separation of the sulphides, are
found to contain more free gold than the coarse sand does, this
fact points to the conclusion that a coarser screen might be used

without detriment, and experiments in this direction should be
made, and the limit of economy thus found by trial.
The evils of overstamping, due to slowness of discharge, have
been often dwelt on, and probably are frequently exaggerated. It
is true that the excessive production of slime thus caused is an
undoubted evil, but, setting this aside, it has been frequently
asserted that particles of gold are reduced in size by cverstamping,
so that they will float off in suspension in water, whilst, even
if not reduced in size, they are hammered and flattened so as
to be rendered incapable of amalgamation. Other authorities,
however, consider it doubtful whether much loss is due to
either of these causes. Small particles of gold, it is contended,
are not likely to suffer further comminution when mixed with
larger particles of gangue, which would usually prevent them
from coming into contact with the shoes and dies these small;
particles, too, would probably only receive a single blow, which

would throw them off the die, and a large piece of gold, which might
remain on the die while several blows were struck, would not be
easily lost even if partially subdivided. As regards hammered
gold, it does not seem to be beyond doubt that flattening and
hardening alone will prevent gold from being amalgamated.
Prof. T. Egleston has described a number of experiments* which
tend to show that amalgamation is retarded by this treatment of
gold, but, on attempting to repeat these experiments at the Royal
Mint, the author could not obtain similar results to those of Pro-
fessor Egleston. Pieces of pure gold when subjected to repeated
blows with a clean 7-lb. hammer on a clean anvil occasionally
showed a disinclination to amalgamate, but, if these pieces were
washed with dilute ammonia, so as to remove any grease that
might be adhering to them, they were instantly wetted by mercury
and were dissolved by it at about the same rate as clean annealed
gold. It thus appeared that, in these cases at least, grease from
the fingers of the operator formed a more potent preventive
of amalgamation than the hardness of the gold. Moreover,
gold-leaf which has been subjected to an extended course of
hammering is readily amalgamable. The matter may perhaps
be left as an open question for the present.
There is still much difference of opinion as to the desirability
of attempting to amalgamate the gold inside the battery. It is
considered by some authorities that the addition of mercury to
the mortar is a mistake, and that no copper plates should be put
inside. In spite of numerous practical illustrations of the
soundness of the contrary view, they hold that no machine can
be successful at once as a crusher and an amalgamator. The
practice of adding mercury to the mortar when no inside plates
*
Mcta'luryy of Gold, Silver, and Mercury in the United State*, 1887,
vol. ii., p. 586.
AMALGAMATION IN THE STAMP BATTERY. 14 J
are used is certainly not now much favoured,

although it is still
adhered to in Ballarat and some other districts in Australia.
In treating rich ores, however, when the gold is coarse and of hi<*h
standard, there does not seem to be any valid objection to be
raised against catching the gold on inside plates in a concen-
trated form, instead of letting it all go to the outside. In this
case mercury must be added to the mortar, and this is probably
not so important a cause of loss of mercury by flouring and
sickening as is often assumed. If a decomposing ore is mixed
with lime beforehand, and the acidity of the water used is
corrected, the conditions do not appear to be favourable to the
production of salts injurious to the mercury, and the latter when
charged in is probably almost instantly washed through the
screens or else dashed against and retained by the plates, the
condition of which is thereby improved by the softening of the
amalgam, so that gold is more readily caught on it. The mercury
does not remain on the die, subjected to repeated blows, which
would no doubt cause much flouring. Finely divided gold, how-
ever, particularly if it is of low standard, containing much silver,
requires more mercury for its amalgamation than coarse gold, and
in this case it is difficult to keep the plates in good order, so that
it is usually advantageous to save the extra trouble and labour,
caused by looking after and cleaning-up inside plates, by putting
all the plates outside. This has been the experience in a number
of mills, including that of the Montana Company, where the
inside plates have been entirely discarded.
Amalgamation outside the battery has also been the subject
of much discussion and some careful investigation. Numerous
amalgamating machines have been patented, the inventors in
every case praising their own contrivances and decrying the
copper plate, but the latter has not as yet been superseded, and
its principle applied to almost all its more promising rivals.
is
To securesuccessful amalgamation it is necessary that the
particles of gold should be brought into absolute contact with the
mercury. This contact is obtained in one of three ways, viz. :
1. The mercury and ore are ground together in pans, arrastras

and similar machines, contact being secured by pressing the gold
and mercury together.
2. The ore is allowed to flow over or even through a bath of
liquid mercury, or the endeavour is made to ensure contact by

letting the pulp fall from a height upon the mercury.
3. The ore is allowed to flow over either stationary or shak-
ing amalgamated copper plates, drops being introduced between

the plates to break up the pulp and to assist in catching the
amalgam.
The first method is undoubtedly the best for ensuring contact,
but the operation is tedious and, in most cases, unnecessary.
Opinions are divided as to the relative merits of the two latter
methods, but the great majority of metallurgists are advocates

of the superiority of the plates. They point out that in a mer-
cury bath, in spite of the first impression to the contrary felt
by every one on approaching the subject, contact with the ore is
very difficult to obtain. If wet pulp is introduced at the bottom
of a bath of mercury, it rises to the surface in lumps, surrounded
by films of water, and dry pulp is still more effectually pro-
tected from contact with the mercury by films of air. If a thin
stream of pulp is run over the surface of a bath of mercury of
sufficient size, the chances of the particles of gold coming in
contact with the quicksilver (by settling through the stream)
are greatly improved, but in this case the more convenient
copper plate could be substituted for the bath. Moreover, it is
well known that rich gold or silver amalgam catches gold more
readily than pure mercury, and, whilst the surface of the plates
can readily be covered with such pasty amalgam, it would
involve a large and unnecessary sinking of capital to keep any
considerable percentage of precious metals in the baths. For
these reasons, and for the practical reason that plates are found
to work better than baths, the use of the latter has been
gradually more and more restricted, until they are now only to
be found in narrow wells and riffles for the purpose of catching
hard amalgam and floured mercury, and as purely supplementary
aids to the plates, whilst in the most modern mills they are
often dispensed with altogether. When a proper disposition of
the plates is made, it is rare to find amalgamable gold escaping
into the tailings. Even if the latter contain several penny-
weights of gold to the ton, this does not show that the
amalgamation effected on the plates is unsatisfactory, and, where
tailings have to be reground or roasted, or treated in any way
that may alter the condition of the gold, before a further extrac-
tion by mercury is obtained, no proof is afforded that the plates
are not doing their work.
The loss of finely divided or " float " gold, particularly when
it cannot be checked by the use of swinging plates, is often
another name for the loss of slimed sulphides. Many examples
have been adduced by the vendors of patent amalgamating or
chlorinating machinery of the large percentage of the gold in
the ores crushed in particular mills, which has been carried
away suspended in water in a form not easily recoverable by
settling. In the majority of these cases, however, no attempt
seems to have been made to distinguish between the values
contained in slimed sulphides and those existing as particles
of free gold. Where this is not done there are no grounds for
the assumption that any free gold is escaping at all. Thus
G. M'Dougal, of Grass Valley, California, found* that a gallon
of water, in the stream, ^ mile below two mills, contained on
*
.
Mines West of the Pocky Mountains, by R. W. Raymond.
AMALGAMATION IN THE STAMP BATTEItY. 143
an average 1*18 cents worth of gold. He called this " float" gold,
but did not try to find out its physical condition, and it was
very likely contained in sulphides. Again at the Spring Gully
Mine, in Queensland, the tailings from the battery, if settled in
the ordinary way by running off the water, were found to con-
tain 7 dwts. of gold per ton, but, if carefully filtered, assayed 15
dwts. All such examples prove only that the slimes are rich,
not that "float" gold is being lost, and although it is of course
likely that some finely divided gold is carried away in suspen-
sion in water during the treatment of many ores, nevertheless,,
if sufficientcare were taken in ascertaining this loss, it would
probably prove to be less than is generally believed.
The following scheme of examining tailings with a view
to determine the causes and amount of loss is given by
M'Dermott & Duffield * :
Small samples are taken at intervals from the waste outflow

of the mill, until a bucketful is collected ; this is allowed to-
settle for several hours, the clear water is decanted, preferably
through a filter, and the remainder evaporated to dryness. Care
must be taken to avoid spilling anything out of the vessels
containing the samples. The sample having been well mixed,
portions are treated as follows
A. One part is panned and examined for free gold, amalgam,
and quicksilver. If these are present, it is probably the fault of
the millman, and nothing further need be done until this state of
things is remedied.
B. The tailings are sized on brass screens, and the coarse,
medium, and fine materials (the latter consisting, say, of that
portion which passes a 100-mesh screen) are weighed and assayed
separately, the coarser portions being reground and panned to>
find whether their values are in free gold or in sulphides.
C. The sulphides are separated from the tailings on a vanning
shovel or batea, and are weighed and assayed. It may thus be
determined whether they are worth saving, and the size of the
mesh used for the screens will depend largely on this, and on
the nature of the sulphides, which will in many cases be badly
slimed and difficult to catch if the ore is finely crushed.
"
D. The loss due to fine or " float gold may be determined
by assaying the slimes after the sulphides have been carefully
removed by concentration. This requires much skill and patience,
but can in almost all cases be successfully accomplished by the
vanning shovel. The concentrates may be examined under the
microscope for fine specks of gold, but these and the fine sulphides
can be recovered by concentration on suitable machinery. The
assay value of the tailings from the vanning shovel
will give
some idea of the amount of float gold which is being lost. It
will usually be found to be smaller than may be expected. If it
*
Gold Amalgamation, London and New York, 1890, p. 7.
is large, the use of some system of amalgamation more perfect

than that by copper plates (such as pan-amalgamation) or of a
method of smelting, or of a wet method, may be considered, if
the advantages appear sufficient to pay for the presumably
increased cost.
Non-Amalgamable Gold. The appearance in the tailings
of free gold, which is not especially finely divided, but, neverthe-
less, is not in a condition to be amalgamated, may be regarded
as a rare occurrence, but deserves some consideration. 'Amal-
gamation is in these cases prevented by the existence of a
thin film of some neutral substance over the surface of the
gold. The film may be so thin as to be transparent, but it is
enough to prevent contact between the gold and the mercury.
The disastrous effect of a film of grease covering gold particles
has already been remarked upon. It is said to have been a
fruitful source of loss in the treatment of certain ores in the
Transvaal that they were impregnated with mineral oil. The
effects of grease may be combated by the use of chemicals (caustic
alkalies, potassic cyanide, kc.), but it is, of course, better to use
every precaution to avoid the introduction into the pulp of oil
from bearings, guides, &c., or contained in steam from the boiler.
Losses in amalgamation are also caused by the greasy sub-
stances contained in some ores, such as the powdered hydrated
silicates of magnesia and of alumina, which cause frothing, and
coat the gold with a slime which prevents the action of the
mercury.
Other films are formed of oxide of iron, compounds of sulphur,
arsenic, &c., or of silica. Some years ago J. Hankey, of San
Francisco, had a collection of particles of native gold which
appeared as bright and lustrous as usual, but were coated by
thin translucent films of red oxide of iron. These particles of
" "
rusty gold could not be wetted by mercury, but, if a piece
were snipped off" one end, the mercury seized on the fractured
surface at once. Such gold seems to be rare in nature.
In 1867, William Skey, of the Geological Survey of New
Zealand, after a series of experiments on the ores and tailings of
the Thames Valley, came to the conclusion that the bright gold
particles which refused to amalgamate were always coated by
some compound of sulphur. He found that gold takes up
sulphur from sulphides of ammonium or sodium, or from sul-
phuretted hydrogen, when brought in contact with their solu-
tions, and that after this the gold refuses to amalgamate. He
supposed that these compounds of sulphur were often formed by
the action of acidulated water on the minerals in ores, and that
"
consequently a large area of the natural surfaces of native gold
is covered with a thin film of auriferous sulphide, and that the
greater part of the gold which escapes amalgamation at the

battery consists of this sulphurised gold."
Gold in Pyrites. In the preceding pages no account has

been taken of the loss of gold which is contained in pyrites, as
it has been assumed that these are saved
by concentration if
they are valuable, and this subject is dealt with in Chapter ix.
Nevertheless, as this gold comes under the head of non-amal-
gamable gold, its physical state and the causes of its disinclination
to unite with mercury may conveniently be considered here. In
generaj, pyrites yields an extremely low percentage of its gold
contents if it is run over the amalgamated plates, and if it is
ground very fine in a pan with mercury the percentage extraction
is better. However, in general, pyrites, even if it is reduced to
the very finest slimes, and its prolonged contact with mercury
ensured by continuous grinding, cannot be made to yield more
than about 40 per cent, of its gold, and this is at the cost of much
mercury, which is floured or sickened during the process. Iso-
lated instances of better results are on record, but must be
regarded as exceptional cases. Among the old processes used
for the amalgamation of the gold in pyrites may be mentioned
the treatment in revolving wooden barrels with mercury, as
practised at the St. John del Rey Mine, and the practice of
leaving the pyrites to be decomposed by weathering before
grinding it with mercury. This method of oxidation seems to
be decidedly inferior to the alternative plan of roasting the
sulphides, by which the oxidation is rendered more complete
and the particles of gold agglomerated to some extent. However,
the amalgamation of pyrites, even when roasted, is far from
perfect, part of the gold still remaining in a condition unfit for
extraction in this way. The ores, which have been met with in
various parts of the world, consisting mainly of lirnonite or
hydrated oxide of iron, and in most cases believed to be the
result of decomposition of pyritic ores by atmospheric agencies,
are also extremely refractory, causing the mercury to sicken
rapidly, and yielding only about the same percentage of gold as
can be obtained from unoxidised pyrites.
The most celebrated case of this kind is that of the Mount
Morgan ore, in Queensland, which is an ironstone gossan con-
sisting of silicious brown iron ore, derived according to one view
from the decomposition of pyrites. Although the gold appears
to be free, it cannot be amalgamated, yielding only about 30 per
cent, when crushed in batteries and subjected to prolonged
grinding in pans with mercury. When the ore is dehydrated by
of
roasting in reverberatory furnaces the extremely fine particles
gold are agglomerated, and between 80 and 90 per cent, can then
be extracted by amalgamation, the remainder being presumably
coated by oxides of iron. The richness of the ore, however,
makes even this result unsatisfactory, and a process of chlorina-
tion has been adopted in practice. A
similar case was noticed
Mr. Mactear * in South where a limonite ore which
by America,
*
Mining Journal, Jan. 24, 1893, p. 70. 10
only yielded from 35 to 40 per cent, of its gold when treated in

Huntington pans, was made to yield between 85 and 90 per
cent, by merely subjecting it to a dehydrating calcination before
amalgamating it. Louis Janin, Jr., mentions another case*
in the ores of the Southern Cross Mine, Deer Lodge County,
Montana, which consist of limonite derived from the alteration
of pyrites. In panning large samples, only one or two specks of
gold could be seen, although the ore contained from 1 to 2 ounces
per ton. This ore yielded only about 40 per cent, on being
amalgamated, but over 90 per cent, was dissolved out by leach-
ing the raw ore with cyanide of potassium, and similar results
were obtained by chlorination. Here the ore was thoroughly
decomposed, but yet the gold would not amalgamate to a much
greater extent than if it were still contained in the original
pyrites, whilst the chemicals at once dissolved it.
In seeking to explain the behaviour of gold in pyrites, various
theories have been propounded. According to one of these, the
gold is supposed to exist in the pyrites in the form of sulphide
combined with sulphides of iron, silver, copper, <fcc., and the
refusal of the gold to amalgamate is explained in this way, auric
sulphide not being acted on by mercury. Some observers have
endeavoured to dissolve gold out of pyrites by the action of
alkaline sulphides, and when, after many attempts, this was at
length successfully accomplished, it was put forward as additional
evidence that the gold must have been in the state of sulphide,
although metallic gold is known to be soluble in these menstrua.
The balance of evidence, however, seems to be in favour of the
theory that gold, at any rate in great part, exists in pyrites in
the metallic state. Although the metal is generally invisible in
undecomposed crystals of pyrites, it becomes visible when such
crystals are oxidised either by air and water in nature, or by
means of nitric acid, or by being roasted or subjected to
deflagration with nitre. As a result of such decomposition,
particles of bright lustrous gold, angular and ragged in shape,
but of considerable size, often become apparent. These particles
may be separated from the oxides of iron by washing, and the use
of nitric acid, followed by panning, is frequently resorted to in
order to detect gold in pyrites. Moreover, although usually
invisible, gold can sometimes be seen in unroasted pyrites. As
long ago as the year 1874, Richard Daintree and Latta found
specimens of cubical pyrites, f in which gold could be seen
under a microscope gilding the cleavage planes of the crystals.
Again, G. Melville-Attwood, on examining crystals of auriferous
pyrites from California in 1881,1 found that the faces of the
crystals were gilded in some places, and that here and there
*
Mineral Industry for 1892,
p. 249.
t Proc. Royal So<-. of New South Wales, 1874 paper on " Iron
; Pyrites.'*
$ Precious Metals in the United States, 1881, p. 604.
little specks or drops of gold occurred, partially imbedded in the

pyrites. These films were too thin to be detected by an ordinary
lens, so that it did not seem surprising that such impalpable
material could not be taken up by mercury. Louis Janin,
Jr., more recently* found crystals of pyrites in a porphyritic
gangue from the Republic of Colombia, which had gold in small
globules on their surfaces. Lastly, it has long been known that
crystals of pyrites are often found adhering to an amalgamated
plate, the particles of gold on their surfaces having been amalga-
mated. It seems likely, in view of all these facts, that some of
the gold at any rate is in the metallic state, and its refusal to
amalgamate is not very surprising, when it is remembered how
completely a thin coating of certain sulphurised compounds

prevents amalgamation, and how readily sulphuretted hydrogen
would be evolved from decomposing pyrites. Some authorities
have contended that the metallic gold is disseminated mechani-
cally through the mass of pyrites, but the action of potassic
cyanide, in dissolving the whole of the gold out of comparatively
coarsely crushed pyrites, seems to point to the correctness of the
view that the interior of the crystal is not auriferous, the
deposition of the gold being superficial, so that the enrichment
of the pyrites is confined to its crystalline faces, and possibly, but
not probably, to its cleavage planes.
The following details f of a microscopical examination by Prof.
Morton, of the condition in which pyrites is left after being
leached with cyanide, confirms this view to some extent :
"
Upon the ordinary auriferous sulphide of iron, or arsenical
pyrites, the solution of potassium cyanide acts readily, not by
dissolving the sulphuret, but by attacking the gold upon its-
exposed edges, and eating its way into the cubes by a slow
advance, dissolving out the gold as it goes. An examination
with the microscope of the pyrites after the gold has been
removed, suggests the method of the operation. A. sample of
very rich pyrites, from a mine north of Redding, was treated
with a weak solution, containing less than two-tenths of 1 per
cent, of cyanide, for 168 hours; the assay showed a complete
extraction of the gold ; as the sulphurets showed no change in
their appearance to the naked eye, some of them were placed
under the microscope.
" There is no
change visible in the form of the crystals as a,
whole ; along the fractured faces the mispickel looks clean and
unaltered, showing the silvery-white colour and intense refraction
of the arseno-pyrite. Upon the faces of the crystals appear dark
lines, short, and parallel to each other. In places they are
crowded close together ; in other parts they are at considerable
distances, but always in parallel lines. The lines vary in length,
the
being from four or five to over a hundred times their width;
* t Loc. cit.
Mins-i al Industry, 1892, p. 249.
lines are very irregular and often broken. These lines are fissures
in the pyrites, and extend so deep into it that the microscope does
not reveal their depth. By using the higher powers the walls of
one of the fissures were seen to be completely honeycombed,
looking somewhat like two empty honeycombs set opposite each
other evidently the mineral removed was crystallised along its
;
contact walls at least. As the raw or untreated pyrites does not

show any such fissuring, but, upon the contrary, shows a surface
marked only by striation lines common to pyrites, I assume that
the fissuring in the treated sample is caused by the solution
acting upon some soluble mineral, probably gold, arranged in
plates, occurring in groups, but which, by its colour and iso-
morphism and the extreme tenuity of its lines, is undistinguish-
able from the mass of pyrites enclosing it."
CHAPTER VIII.
OTHER FORMS OF CRUSHING AND AMALGAMATING

MACHINERY.
Special Forms of Stamps. Since within certain limits and
under certain conditions the capacity of a stamp battery depends
on the number of blows given per minute and on the momentum
of the fall, various contrivances have been suggested with a view
to increase both of these. Among these special stamps Husband's
pneumatic stamp, the Ball Steam stamp, and the Elephant stamp
may be noticed.
Husband's Pneumatic Stamp consists usually of two stamps, the
stems of which are attached to pistons of small diameter working
in pneumatic cylinders, D (Fig. 32). These cylinders have a reci-
procating up and down motion given to them by the revolution of
a crank shaft, C, above them, with which they are connected by
iron rods, E, affixed to trunnions, d on the cylinders.
1
,
When a
cylinder is raised by the crank shaft, the air in it, below the
piston, is compressed and the stamp stem thus lifted. Similarly,
the downward motion of the cylinder causes a compression of the
air above the piston, which urges the stamp downwards with
greater velocity than it would have by virtue of its weight alone.
There is also a contrivance for rotating the stamps so as to give
even wearing of shoes and dies. These stamps have not been
much used except on tin-ores in Cornwall. Their output is from
20 to 30 tons per head per day through a 36-mesh screen, the
power required being about 20 to 25 H.P. per head. One of
CRUSHING AND AMALGAMATING MACHINERY. 149
the chief difficulties encountered in attempting fine crushing

with these stamps of enormous capacity is the effect on the
screens. Wire screens do not stand the excessive impact in a
satisfactory manner, while Russia-iron screens, if punched for
HARVEYS PATENT HUSBAND'S STAMPS

Seal* :Viineh,-2 foot.
Fig. 32.
fine crushing, are more effective when thin, and are then of little
durability.
Steam Stamps. The ordinary form of steam stamp consists of
a direct-acting vertical engine, having a steam cylinder and
slide valve at the top, the piston-rod being rigidly connected

with the stamp. Each stamp head works in a separate rectan-
gular cast-iron mortar, with screens both at the front and the
back, and sometimes all round. The screens are of Russia sheet-
iron with punched holes of about to y ^ inch in diameter, the
-J-
1
steam stamp being best adapted for coarse crushing. The speed
Fig. 33.
of working is from 90 to 100 blows per minute, and the out-turn

isfrom 100 tons to as much as 225 tons of ore per head in twenty-
four hours. It is obvious that gold could not be economically
saved on plates inside the mortar of one of these stamps, and, as
a matter of fact, until recently they were only employed in
coarsely crushing the copper ores of the Lake Superior region.
Nevertheless, as the curious fact seems to be well established

that these stamps, with their heavy blow, do not make so much
slimes as the ordinary gravitation stamp, it is to be expected
that they will be largely used in the future for crushing before
concentration. They are already in use to crush silver ores in
Montana, and gold ores at the Homestake mine, in the Black
Hills, through a 30-mesh screen. As their capacity is so great
their use is limited to cases in which large quantities of ore are
available, one Ball steam stamp, such as is used in the Lake
Superior district, being equal to at least fifty head of gravitation
stamps.
The chief advantages of the steam stamp are economy of space
and labour. With gravitation stamps more work is thrown on
the rock-breaker, but it is doubtful whether there is any waste
of power, as the whole of the work can be done by means of a
single steam engine. The advantage of subdividing the work
among a number of batteries is that stoppages for repairs and
breakages affect only a small part of the crushing capacity at one
time. Also, it has been pointed out that water-power can be
directly applied to gravitation stamps with little loss of efficiency,
whilst it is more difficult and wasteful to apply water-power to
compress air for use in place of steam in the Ball stamp.
The ElepJiant Stamp consists of a bent or compound lever of
hammered-iron or steel (a, a, Fig. 33), one end of which is
pivoted on a fulcrum-pin, 6, and the other end is fitted with
stamp shoes. The figure represents a battery of two stamps.
The levers are connected with the crank-shaft, c, by strong semi-
circular springs, d, secured to short connecting rods, e. The
stamp heads work in a mortar, f, fitted with dies in the usual
way, and having discharge screens on three sides, and coarsely
crushed quartz is fed in through an inclined shoot at the back.
The action of the bow-springs, d, introduced between the crank
shaft and the lever, is to receive and store up the force of the
recoil from the blows of the stamps, and by giving it out at the
next descent of the crank to effect a certain saving of power ; in
addition to this the springs act as cushions, taking up the jarring
effect of the blows, and so diminishing the wear and tear of the
machine. A mill, consisting of two heads of Elephant stamps, is
said to be capable of stamping from 12 to 15 tons of hard gold-
quartz in twenty-four hours through a 30-mesh screen. The
advantages over gravitation stamps consist in the lightness,
cheapness, ease of transportation and erection of Elephant
stamps, and the driving power required, and they are conse-
quently especially suitable for prospecting purposes, and in pre-
liminary work during the development of new mines. It was
found, however, in India that the lateral wear and tear at the
top of the mortar-box by the levers which carried the crushing
heads was very great, and frequent renewals were necessary.
The constant position of the heads, moreover, makes the wear

very uneven, and the machines require much oil, whilst their
capacity rapidly falls off as they are being used.
The Huntington Mill. The Huntington Roller Mill, having
now been in steady use for several years in the United States
and elsewhere, has conclusively proved its claim to rank with
stamps as a practical machine for fine crushing. It consists of
an iron pan, at the top of which a ring, B (Fig. 34), is set, and
attached to this are three stems, D, each of which has a steel
shoe, E, fastened to it. The stems are suspended from the ring
and are free to swing in a radial direction, as well as to rotate
round their own axes, whilst the whole ring, B, with the stems
Fig. 34.
and shoes, revolves round the central shaft, G. The shoes or

rollers, asthey are called, are thus driven outwards by centri-
fugal force and press against the replaceable ring die, C.
The
most modern yoke for the suspension of the rollers is shown in
Fig. 35.* In front of each roller is a scraper, F, which keeps
the ore from packing. The rollers are suspended with their
bases at the distance of 1 inch from the bottom of the pan,
which can also be replaced when worn. The lowest part of the
screen is situated a little above the top of the rollers, and out-
side it there is a deep gutter into which the ore is discharged,
*This figure is taken from Mr. A. Harper Curtis's paper on "Gold
Ouartz Reduction." Proc. Civil Engineers, vol. cviii., part ii., 1892.
and from which a passage leads to the copper-plate tables. The

ore and water being fed into the mill through the hopper, A,
generally by an automatic feeder, the i?otating rollers and the
scrapers throw the ore against the sides where it is crushed to
any required degree of fineness by the centrifugal force of the
rollers acting against the ring die. From 17 to 25 Ibs. of
mercury are placed in the bottom of the pan, the clearance below
the rollers permitting them to pass freely over the mercury
without coming in contact with it, so that it is not stirred up and
Fig. 35.
Scale, | inch = 1 foot.
"
floured," but the motion is such as to bring the pulp in contact
with the quicksilver. The speed of the mill is from 45 to 75
revolutions per minute. The ore should be broken in rock-
breakers to a maximum size equal to that of a walnut, or, better
still, of a cobnut, before being fed in. The action of the rollers is
one of impact rather than of grinding, the ore being granulated
without the production of much slimes. The free gold, as soon
as it liberated from its matrix, is in great part amalgamated
is
and retained by the mercury at the bottom of the pan, the
remainder being caught on the plates outside the mill. Coarse
gold is caught inside and fine gold outside the mill, but the yield
inside iscomparatively small when ores with high percentages of
sulphides are in course of treatment.
The mill is particularly adapted for the treatment of ores con-
taining brittle sulphides, which, if pulverised by stamps, are
liable to become "slimed," and so to be in an unsuitable condition
for concentration. It is also suitable for argillaceous quartzes,
"
which yield their gold more readily under the " puddling action
of the rollers than when pounded by stamps. Moreover, the
Huntington mill does much more satisfactory work than stamps
on soffc ores or in regrinding coarse tailings. The reason for this
lies, of course, in the relatively large amount of screen area in
the mill and its consequent high efficiency of discharge, a point
in which stamps are decidedly inferior to it. In the experi-
ments at the Metacom Mill, California, already quoted, it was
proved that stamps will take as long to pass a ton of tailings
through a screen as a ton of the original rock, but with the
Huntington mill the finer and softer the material, the more
rapidly it is passed through. As the splash is heavy against
the sides the wear of the screens is somewhat rapid, but they
-can be very quickly replaced.
The mill is made in three sizes viz., 3^, 5, and 6 feet in
diameter respectively and the capacity of a 5 feet mill, the one
which is most commonly in use, is from 10 to 20 tons of rock
per day through a 30-mesh screen, the power required being
from 10 to 12 H.P. The weight of the ring die in the mill is
611 Ibs., and of each of the roller shells, which are replaceable,
about 140 Ibs. The wear and tear on these replaceable parts
is very great, amounting to about 14 oz. of steel per ton of rock
crushed when soft ores, previously broken small, are being treated.
If large pieces of hard quartz are fed into the mill or if the mill
is mercury is splashed against the screens and
overfed, the
passes through with the pulp, and when by accident pieces of
iron or steel are introduced, the ring die is occasionally broken.
Another source of disaster in Huntington mills lies in the use
of acidulated water, such as that derived from mines or encoun-
tered when decomposing pyritic ores are treated ; the mill is
rapidly corroded and rendered unfit for work by such water.
The chief advantages supposed to be gained by the use of
Huntington mills instead of stamps may be thus epitomised:
1. Reduced First Cost. The cost for the same capacity is not
more than two-thirds that of stamps, even at the manufacturers'
shops, while the difference in favour of the mill is even more
in outlying districts from its light weight, and corresponding
low freight, and from the cheapness of its erection.
2. Saving of Power. The mill is said to run with about one-
half the power per ton of ore crushed.
3. The wear and cost of renewals is less for the mill than for
CRUSHING AND AMALGAMATING MACHINERY. ^55
stamps, the cost being from twopence to threepence per ton of

ore for the former, against about fivepence or sixpence for
the latter.
4. There is less loss by flouring of amalgam and quicksilver,
while the good discharge and absence of grinding leaves the pulp
in a better condition for concentration.
The first three advantages appear to refer only to such soft
and brittle ores as are especially suited to the Huntington mill.
The mill requires to be set to work in an intelligent manner by
experienced and skilful hands, and watched carefully. The
dangers of over-feeding have been already alluded to. One
difficulty in automatic feeding is that self-feeding,
such as is
carried on by stamps, is impossible. The automatic feeder must
work separately and be set to feed a certain weight of ore per
hour, this weight having been determined by trial. If, after
this, there is any change in the hardness of the rock, no automatic

change in the rate of feeding takes place, and the machine may
be choked up or run at below its maximum capacity unless
watched and the feeder regulated. Another difficulty is in the
quantity of water to be added. An excess of water, making
thin pulp, does not favour internal amalgamation, and it may
be stated that, in general, the pulp should be kept as thick as
possible, consistent with its prompt discharge through the screens
when sufficiently fine. If the pulp is too thick to run easily over
the copper plates outside the mill, water may be added there by
means of a perforated pipe. The rate of running should be as high
as possible, since, if the other conditions are the same, the crush-
ing power varies as the cube of the number of revolutions per
minute.
A few examples are appended of the results obtained in
actual practice by this excellent machine. At the Spanish
Mine, Nevada County, California, there are four Huntington
mills, three of 5 feet diameter and one of 4 feet diameter. The
ore is free-milling and is passed through a Blake stone-breaker
and thence to the mills. The four mills run at 58 revolutions
per minute, and pulverise 35 tons of ore each in twenty-four
hours, to pass through a slot screen equal to 20-mesh. The
pulp is passed over the usual amalgamated plates after leaving
the mills, and J oz. of mercury is added with each ton of ore.
Forty-five per cent, of the gold recovered comes from the inside
of the mill, where the amalgam obtained is much richer than
that from the plates. The loss of quicksilver is from -f^ to -^ oz.
per ton of ore, and the total cost of milling is about one shilling
per ton, while for the month of November, 1887, it was only
tenpence per ton. The ore is a soft talcose slate, containing
streaks and veins of ferruginous quartz, carrying gold. The
chief trouble in working lies in the frequent re-adjustment of
feed which is found necessary. In a special test run of one
month 42 -4 per cent, of the gold contents was extracted, and

the remainder lost in the tailings. This poor result was pro-
bably due to over-feeding, but profits were made, nevertheless,
although the ore yielded only a little over 1 dwt. of gold per ton
in 1887 and 1888. Twenty-two horse-power were used by the
Huntington mills in crushing from 120 to 140 tons per day.
At the Shaw Mine, El Dorado County, a 5-foot mill, making
50 revolutions per minute, pulverised 10 to 12 tons per day, so
as to pass through a 25- to 30-mesh screen. At the Mathines-
Creek Mine, in the same county, a 5-foot mill pulverised 9 to
10 tons per twenty-four hours, so as to pass a screen equal to a
40-mesh. At the Monte Christo Mine, Mono County, two 5-foot
mills, running at from 65 to 75 revolutions per minute, pulverised
2 tons of ore per hour to pass a screen equal to a 40-mesh ; 25
Ibs. of quicksilver were charged into each mill at the commence-
ment of the run, and about J oz. more added each half hour.
Although cases have been adduced in which 90 per cent, of
the gold contents of an ore crushed in the Huntington mill was
retained inside the machine, this is decidedly exceptional, and
the mill is probably inferior to the stamp mill as an amalgamator
on many ores, although its product is better adapted for treat-
ment on copper plates and by concentration.
Crawford Ball Mill. This mill (see Fig. 36 *) consists of an
annular trough, divided concentrically by a vertical slit, which
passes through its deepest part all round ; the outer part of the
trough is fixed to the framework, whilst the inner portion,
having a circular spindle passing up through it, is capable of
being revolved by bevel gearing as in the Huntington mill. In
the annular trough, which is 4 feet in diameter, are placed eight
chilled cast-iron or hard-steel balls, each about 8 inches in dia-
meter two of these are shown in position in the figure. Below
:
the vertical slit in the trough there is another vessel, in which the
mercury for amalgamation is placed; this vessel communicates
with the trough by the narrow continuous slit passing all round
the machine. The ore, previously partly crushed, is fed into
the mill through the hopper, and the central spindle, carrying
with it the inner half of the annular trough, is revolved about
180 times per minute. This causes the 8-inch balls to revolve,
each on its own axis, and all of them around the annular trough,
and the ore is ground by them. A
current of water is intro-
duced into the lower cavity of the machine, and passing over
the surface of the mercury moves up through the vertical slit
in the bottom of the trough, and through the ore which is in
process of being crushed, and overflows at the top. The direction
of movement of the pulp is shown by the arrows.
The idea is that, as the ore is crushed sufficiently fine, it will
rise and pass away with the current of water, and that the
particles of gold will fall down through the slit at the bottom
*
From Proc. Inst. Civil Eng., vol cviii., part ii., 1892.
of the trough, and coming in contact with the mercury, will be

amalgamated there. It is claimed that the current of water
may be regulated so that gold only can fall through it, while
not only quartz but sulphides also are carried up by it, and not
permitted to come in contact with the mercury,
which is thereby
kept quite clean and preserved from agitation, so that the loss
due to flouring and sickening is entirely obviated.
Fig. 36.
It is obvious that rather too much is expected from this

machine. Since the effect of the current of water is exactly
the same as that in the ascending - current classifiers (see
p. 170), it is clear that "equal-falling" particles will not be
separated from each other by its action. Consequently, only
the very coarsest particles of gold, which could not escape
amalgamation in any machine yet devised, can be separated
irom sulphides in this way. Moreover, it seems likely that
the agitation produced by the movement of the balls would

occasionally permit the escape of large particles of ore before
they were finally comminuted, and, what is of much greater
importance, the rapid swirl of the water and pulp, travelling
in places at little less than 24 feet per second, which is the
speed of the periphery of the moving part of the trough, would
appear to effectually prevent that perfect separation by gravity
of heavy and light particles, which is essential to the successful
working of the machine. The machine is of little practical value,
and has only been described as a type of ball mills.
Other Ball Mills. Among other mills in which the crushing
is done by iron balls may be mentioned the Cyclops mill, in
which a single large ball, 12 inches in diameter, is carried
round in a vertical plane by frictional contact with a pair of
flexible discs fitted on a horizontal revolving shaft. The grind-
ing is done between the ball and a grooved circular path in a
vertical plane, against which it is pressed by centrifugal force.
The ball makes from 250 to 600 revolutions per minute, and the
largest sized mill is said to crush over 40 tons per day. It is
fitted for both wet and dry crushing, but records do not seem
to exist of successful work done on a large scale during any
considerable length of time.
Speaking generally, most of the ball machines which have
been devised have already been proved to be very uneconomical
owing chiefly to the great wear and tear of the grinding
surfaces, and the enormous driving power required. It is
difficult to obtain a ball which can be relied on to wear evenly,
and in many cases the effectiveness of the machine is materially
reduced after it has been in operation for a very short space of
time.
Amalgamation Pans. An amalgamation pan consists of a
circular cast-iron pan, provided on the inside with a renewable
false bottom of cast iron constituting the lower grinding sur-
face and a "muller," or upper grinding surface (d, Fig. 37),
attached to a vertical revolving spindle, g, which is set in motion
by bevel wheels, t, placed below the pan. The muller grinds to
impalpable pulp ore which has been already reduced to a coarse
powder by stamps, and also mixes the ore with mercury, intro-
duced into the bottom of the pan, and so amalgamates the gold
and silver. The origin of the pan is clearly to be traced to the
Mexican arrastra, and some of the varieties of the pan are merely
slightly modified arrastras. One variety consists of a sectional
wrought-iron pan, fitted with a granite grinding bottom, and
with granite mullers, which are attached to a vertical spindle
rotated by hand or by animal power. The Berdan pan also
forms a transition between the arrastra and the modern pan,
see p. 205.
In work on gold ores the use of amalgamating pans is mainly
limited to regrinding skimmings, blanket sands, and concentrates
obtained in working a stamp mill. Silver ores in many cases do

not readily yield their precious metal if merely treated in a stamp
battery and run over copper plates. These ores are then crushed
in a battery, roasted with salt if necessary, and then amalgamated
in pans. Silver ores containing considerable quantities of gold
are often similarly treated, but with purely gold ores it is seldom
necessary to resort to this process, and a detailed description
will, therefore, be more in place in the volume devoted to the
metallurgy of silver, only a brief account being given below.
Gold ores which do not yield a fair percentage of their values
by copper-plate amalgamation are occasionally treated in pans.
In such cases the ore may be roasted or treated raw. As already
stated, p. 145, it is seldom advantageous to roast a gold ore
before amalgamation, since, although in a roasted pyritic ore
specks of free gold may often be detected where none were
visible in the raw ore, a part of the precious metal usually
appears after roasting to be difficult to bring in contact with
mercury. The cause of this is not always easy to discover, but
it may sometimes be due to the coating of gold by thin films of
iron oxide or other mineral. Moreover, the addition of salt to
a gold ore in the roasting furnace, as is pointed out in the
chapter on chlorination, is often attended by appreciable losses
by volatilisation. These two causes are sufficient to account for
the low percentage of gold usually extracted when an auriferous
silver ore is treated by roasting with salt and pan-amalgamation.
When no base metals are present a gold ore may sometimes
prove to be satisfactorily handled by roasting and amalgamation,
but such cases are exceptional.
Pan-amalgamation, whether the ores treated are raw or roasted,
may be conducted in one of two ways. The older system is to>
crush wet in the stamp mill, and collect the ore in large shallow
settling pits or pointed boxes (see p. 168). A sufficiently dry
pulp having been obtained by draining, it is dug out by hand
and charged into the pans. The modern system, which is already
extensively in use, is called the Boss Continuous System, after
the name of its inventor, M. P. Boss. It is briefly described
on p. 162.
Old System. The amalgamating pans in use are very-
numerous, and vary greatly in form. The shape of the bottom
was formerly much in dispute, flat, cone-shaped, and hemi-
spherical bottoms each having its advocates, but it is now
generally believed that flat-bottomed pans are the best, wearing
more evenly and doing more work. The pans are often heated,
so as to increase the rate of amalgamation by means of steam
led through a chamber below a false bottom in the pan, but the
more economical device of introducing steam into the pulp
itself has also at all times been in use. The objections to the
latter course are that the pulp may be so much diluted that
amalgamation is checked, and that oil is liable to be introduced

with the steam with equally disastrous results. When the ore
is roasted before
being treated in the pan, it is in some mills
charged in hot, hot water being added also, and as the pan is
covered up and is still warm from the previous charge, it
remains at a sufficiently high temperature throughout the
operation without further treatment. The grinding of the ore
by the muller is an additional source of heat.
One of the common forms of amalgamating pans is shown in
Fig. 37. This pan is 5 feet in diameter, with cast-iron bottom, a,
Fig. 37.
and wooden The mullers are shown resting on the cast-

sides, h.
iron dies, c, which protect the bottom from wear, whilst re-
placeable shoes attached to the
lower surface of the mullers
are also shown. The shoes and dies can be kept in contact
while the spindle, g, is rotated, so that the ore can be ground,
or the muller can be raised by rotating the hand -wheel and
centre screw, j, on the top of the spindle, so that only circulation
and mixing of the charge take place. In some pans copper plates,
/,
are introduced, being attached to the side walls and projecting
into the interior. These are intended both to mix the
plates
pulp and to catch the amalgam, much of which is retained on

them. The more usual system is to employ separate vessels
called settlers for the collection of the quicksilver and amalgam,
after the pans are discharged. The speed of the muller is
usually from 65 to 75 revolutions per minute below the muller
;
the pulp is continually worked from the centre of the pan to the
circumference, being returned towards the centre above the
muller and passing down through the latter by inclined slots
which terminate near the centre. In Fig. 37, which represents
the form known as the Patton pan, n is the main through
which steam is passed into the chamber, b, to heat the pulp, and
m is the outlet pipe.
The method of operation is as follows : The charge of ore is
introduced with the mullers raised slightly and kept revolving,
water being added at the same time in quantities sufficient to
make the pulp of a pasty consistency, so that globules of mercury
remain suspended in it without subsiding. The mullers are then
lowered and the ore ground for from two to four hours, after
which the mullers are raised and the mercury added gradually,
and thoroughly mixed with the pulp for 6 to 8 hours longer. The
object in raising the mullers is to prevent the sulphides from
being ground up with mercury, which would cause considerable
losses by flouring and sickening. Nevertheless this raising of the
mullers is not an invariable practice. When the amalgamation
is thought to be complete, water is introduced to dilute the pulp,
and the whole is discharged into a settler ; or else the diluted
pulp is stirred by the raised muller at a reduced rate of speed
until the globules of mercury have re-united and sunk to the
bottom, when the pulp is gradually run off, beginning at the top,
usually by pulling out in succession plugs set in the side of the
pan at different levels. The discharge takes place into a bucket
or tub, where some of the mercury accidentally carried over is
caught. The bulk of the mercury in some mills is drawn off
from the bottom of the pan before the pulp is discharged.
In order to facilitate amalgamation various chemicals have
been recommended as desirable additions to the charge. At the
present day it is recognised that most of these are either useless
or absolutely harmful, and only salt, sulphate of copper, nitre,
cyanide of potassium, lime, and sodium amalgam are now used.
In treating gold ores, cyanide of potassium and sodium amalgam
are added to keep the mercury clean and lively, but the latter
chemical is now comparatively rarely resorted to. Salt and
sulphate of copper are chiefly added to silver ores, their use
having been suggested by the Patio process. They are believed to
decompose certain base minerals, and so to prevent the sickening
of mercury, which would otherwise be caused by their presence,
and also to liberate silver from some of its compounds and thus
render it capable of amalgamation. The use of lime is of course
11
to neutralise any acid sulphates of iron, tkc., which may be

formed by the partial decomposition of the ore, and so to prevent
the sickening of the mercury. If added when the pulp is
diluted, lime is said to be efficacious in assisting the mercury to
collect together and settle.*
The Boss Continuous System. In this system of pan-amalgama-
tion the pulp is continuously run direct from the stamp battery
through a series of pans arranged so that each overflows into the
next one, which is placed at a slightly lower level. The first
Fig. 38.
Scale, 1 inch = 3 feet.
two or three pans are arranged as grinders, the battery pulp not
being fine enough for complete amalgamation, and the pulp is
then passed through a series of amalgamating-pans supplied with
mercury, after which the mercury and amalgam are separated
from the ore in settlers, which are larger pans in which the pulp
is diluted and stirred less vigorously. The tailings overflow
*For a full account of the chemical reactions involved in pan-amalgamation,
the student is referred to the Report of the United States Survey of the
Fortieth Parallel, vol. in., chap. v.
from the settlers, and are run to waste or led over concentrators.
The number of pans arranged in series through which the pulp
must pass, in order to yield a fair percentage of its precious
metals, is determined by experiment for each particular ore. It
is obvious that the consistency of the pulp must be thinner than
has usually been considered desirable for successful amalgama-
tion, and, as a matter of fact, its volume is usually doubled by
the introduction of the continuous process, but in spite of this
the percentage of extraction is not lower than by the old method.
By the Boss system there is a large saving in labour, fuel, and in
wear and tear ; the settling pits or pointed boxes are dispensed
with, and no movement of the pulp by hand is needed. The
mercury is collected in wells and pumped up into tanks, whence
it isfed automatically into the amalgamating pans. One of the
pans in use in this process is the Boss Standard Pan, shown in
Fig. 38. It will be noticed that there is a steam chamber below
the false bottom of the pan, extending up into the conical space
in the centre, for warming the pulp.
Concentration before Pan- Amalgamation. Some refer-
ence must be made to this subject and also to that of concentra-
tion after pan-amalgamation, as much attention will probably be
directed to them in the near future. In most cases it is desirable
first to remove the sulphides by concentration and then to treat
the tailings in pans. In this way the objectionable, but often
valuable, minerals, which complicate the reactions in the pan and
cause losses in mercury and amalgam by sickening, are prevented
from doing mischief, whilst they are saved by concentration
after the preliminary crushing in the battery, more effectually
than if they are also subjected to the grinding and sliming action
of the mullers. After concentration, the tailings of course con-
tain too much water for immediate pan-amalgamation, and the
excess of water must be removed by settling in pointed boxes.
The chief disadvantage in this course is that, where the ore con-
tains chlorides, sulphides and other compounds of silver, the
slimes which remain in suspension in the water are the richest
to collect them as
part of the pulp. It is, therefore, necessary
perfectly as possible by settling, but, in these cases, a certain
amount of loss is unavoidable. The settled tailings are now in
good condition for amalgamation, and can be treated expedi-
If, on the other hand, the
ore is very
tiously and effectively.
rich in silver, so that the loss caused by immediate concentration
is heavy, it is first treated in pans, and the product is then passed
over Frue vaniiers or other slime tables.
Treatment of Concentrates in the Pan. Some details of
the treatment of concentrates by pan-amalgamation are given in
x. in the accounts of work in special localities. It is, how-
Chap.
ever, a survival of old methods, and does not represent the most
modern practice, in which concentrates are either smelted or
164: THE METALLURGY OF GOLD.
treated by wet methods. Whether they have been previously

roasted or not, the treatment of concentrates in pans is seldom
attended by the successful extraction of a high percentage of
the gold. A recourse to this method is justifiable only when the
concentrates are not in sufficient quantity to warrant ths erection
of a chlorination plant, and when there is no smelting works near
where they can be sold. Under these circumstances a stone
arrastra usually gives better results in treating roasted concen-
trates than an iron pan. In Australia the method is often
adopted of employing a large excess of mercury and little water,
and of keeping the roasted material from contact with iron, and
in some experiments conducted in Mexico, C. A. Stetefeldt found
that by the use of gold amalgam instead of mercury, and by
grinding in stone vessels, a high percentage of gold was extracted
from low grade ores. The chief advantage in roasting lies in
the reduction of the loss of mercury which is effected by the
elimination of the compounds of sulphur, the percentage of
extraction of gold not being, as a rule, much affected. This
subject is discussed under the heading Gold in Pyrites, on p. 145.
Molloy's Hydrogen Amalgamator. A description may be
given of this machine as ithas gained the approval of some
well-known metallurgists, although it has never been highly
esteemed by practical men, and, in spite of the fact that it has
been before the public for several years, has not yet had any
success in prolonged operations on a large scale. It consists
of a circular pan of cast iron, about 40 inches in diameter and
4 inches in depth, which is nearly filled with a bath of mercury.
In the centre is a bottomless ebonite box, the sides of which dip
into the mercury for about J inch. This is called the anode box,
and is filled with an aqueous solution, which in the earlier
form contained sulphate, but in the later form carbonate of soda.
A leaden plate dips into this solution, and is connected with the
positive pole of a battery, the negative pole of which
is con-
nected with the mercury bath. A wooden disc, nearly as large

as the pan, touching the mercury near the periphery of the basin,
but riding free above it towards the centre, is rotated, when the
machine is at work, at a speed of not more than sixteen revolu-
tions per minute, and crushed pulp and water is fed in near the
centre of the machine, but outside the anode box. This pulp
travels outwards in widening circles under the influence of the
rotating disc, which causes it to come in contact with
the
mercury, and overflows at the edge, leaving its precious metals
amalgamated with the bath of quicksilver. An electric current
passed through the machine is supposed to decompose the
car-
bonate of soda, liberating free sodium and hydrogen over the
whole surface of the mercury, which is thereby kept clean and
lively. It is doubtful, however, if any free sodium is ever
present in the bath. If formed it would be at once oxidised
by the water, liberating hydrogen ;

but this element, when
nascent, is doubtless of great efficacy in reducing base oxides
which may be contaminating the mercury, and so in keeping
the latter from sickening. The absence of all grinding action
prevents any considerable loss from occurring by flouring,
consequently, the loss of mercury in operating the machine
would probably be small.
The value of this machine depends on the success of the efforts
to induce perfect contact between the mercury and the gold
particles contained in the ore. It does not seem likely that this
contact is obtained. The machine is said to be able to treat
from 10 to 20 tons per day, and, therefore, to deal with the
product of from five to ten head of stamps. The area of the
amalgamating surface, however, over which the pulp passes is
only about 8 square feet, whilst the area of the copper plates
used in the stamp battery in the ordinary way would be from, say,
32 to over 100 square feet. The layer of pulp between the disc
and the mercury, if this quantity is treated, must therefore be
much thicker than tne corresponding layer flowing over copper
plates, so that there is less chance of every part of the pulp
being brought in contact with the amalgamating surface in the
former case. As already intimated, the machine has not yet
been proved a success in prolonged operations on a large scale.
Jordan's Amalgamator. This machine is intended for the
treatment of auriferous tailings. It consists of a number of
shallow amalgamated copper pans fixed one below the other on
a vertical axis, which is revolved slowly by bevelled gearing
placed at the top. Between each two pans is a circular shelf,
projecting from the wall of the surrounding casing. The tailings,
in the form of fine pulp, are fed in to the first pan, and, by the
centrifugal force exerted by the revolution of this pan, the pulp
is caused to pursue a spiral course, until, after several revolu-
tions, it falls over the edge on to the shelf below. The pulp
then flows back on the shelf to the centre of the machine and into
the second pan. In this way it traverses the surfaces of all the
pans, one after the other, and finally falls into a conical separator,
which, it is claimed, saves any heavy material and also any mer-
cury that has escaped. The machine has not yet passed into
use on any extended scale, although it has been tried at a few
mills for short periods of time. It seems well adapted to catch
fine gold if the amount treated is not too large to admit of
perfect and prolonged contact with the plates.
CHAPTER IX.
CONCENTRATION IN STAMP MILLS.

Concentration. The object of concentration is the separation
of the heavy valuable mineral from the light worthless gangue.
In Europe, where careful attention has been paid to the
mechanical treatment of ores for some hundreds of years, compli-
cations are introduced by the fact that various base minerals
must be separated from one another, an ore being subdivided
into several products. Almost all gold ores, however, only re-
"
quire separation into two parts the concentrates," in which
the precious metal is contained, and the " tailings," which are
thrown away. Consequently, the German machinery has not in
general proved applicable to gold ores without modification, and
most of the best appliances now in use have had their origin in
the United States. The German system of coarse crushing, sizing
by means of screens, and concentrating on jigs, is not, as a rule,
applicable to gold ores proper, although much used on auriferous
lead, zinc, and copper ores. This system will not be described
here, where only the methods in use for the treatment of battery
sands, after passing over the amalgamated plates, will be con-
sidered, and some reference made to the machines in use for
the mechanical treatment of pulp, either before or after pan-
amalgamation.
All the concentrating machines depend for their action on the
effect of a difference of densities on the fall of bodies in a fluid.
The fluidemployed in almost every instance is water, although
several machines have been devised in which air is used as the
concentrating medium. The use of any other fluid than these
may be safely condemned as chimerical. It has often been pro-
posed to use some fluid which shall have a lower density than
the valuable mineral to be saved, but a higher density than the
worthless gangue the mineral would then sink, while the gangue
;
would remain floating. The high cost of any such fluid is suffi-
cient to put this method out of the question, without discussing
any further disadvantages.
The fall in water of solid materials takes place according to
two laws, one applicable to very shallow strata, through which
the particles fall with increasing velocity, while the other is true
when the depth of the water is considerable, so that the particles
for the greater part of their course proceed at- their maximum
velocity. In shallow water the fall is almost entirely according
CONCENTRATION IN STAMP MILLS. 167
to density,* so that those machines which utilise only the first in-
stants of the fall will have great efficacy in concentrating. It is
this fact which has necessitated the use of shallow currents in
concentrating tables, sluices, tfcc. In almost all these machines
the fine sand and slime is brought into suspension in water, and
the liquid is then run over an inclined surface. The deposit of
sand, which is thus formed on the table, tends to become enriched
in heavy minerals, because the stream moves faster at the surface
of the water, where the lighter particles remain, than it does
next the bed, where the heavy particles have settled. The
deposit is continually worked up and brought again into sus-
pension by a rake or broom, or by a series of shakes or blows

imparted to the apparatus, so that the effect mentioned above
is repeated frequently. If the stirring up is violently performed
all the slime and very fine particles are kept in suspension in
the water and carried away and lost, a slow stream of water
and very slight agitation being favourable to their retention in
the deposit which is formed. When the stream of water is rapid
and voluminous, fine material, whether heavy or light, is swept
away and lost in the tailings, whilst if too small a stream of
water is used, much worthless sand is deposited with the con-
centrates. It follows that the amount of water used must be
regulated according to the work which it is proposed to do.
Frequently, it happens that clear water must be added to the
pulp to dilute it sufficiently. On the other hand, it occasionally
happens that the pulp is too thin, especially after hydraulic
sizing, and water is then removed by means of settling boxes,
described below.
Another operation, which it is often of the utmost importance
to perform before concentration, is that of classification by size.
The necessity of this is obvious, when it is remembered that a
shallow stream of water swift enough to carry down fine sul-
phides, might be powerless to move a pebble of quartz. The
usual method of classification is based on the varying rates of
fall of particles through a deep column of water. In this way
equal-falling particles are obtained together, and since a sphere
of galena is equal-falling with a sphere of quartz of four times the
diameter, it follows that the sizing cannot be perfectly performed
by this method. Nevertheless, such sizing as is possible in this
*At the very beginning of the fall, before the velocity has had time
enough become great, the motion varies nearly as g \^~ fjf wnere ^ i s
to
the density of the medium (in this case, water), and D the density of the
solid particle. When the depth is sufficient for the velocity to attain its
maximum, this maximum velocity will be the same for all particles for
which a (D - 3) has the same value, where a- is the area of the section of
the particle at right angles to the line of fall. Such particles are called
equal-Jailing. For full investigations of these formulae the student is
referred to Gallon's Lectures on Mining, English edition, vol. iii., p. 47.
way, when efficiently performed, is of great assistance as a pre-

paration for treatment by shallow-stream concentrators. Some
sizing machines are described below, p. 169.
Settling Boxes. These are for the purpose of allowing the
sand and mineral in suspension in a flowing current to settle, so
that the part of the water not needed in the subsequent treat-
ment of the ore may be run off; if it is desirable, this excess of
water is available for use over again. These boxes will deliver a
small stream of thick concentrated pulp at the bottom and give
an overflow of almost clear water at the top. The boxes manu-
factured by Messrs. Fraser & Chalmers are shown in plan and
Fig. 39.
elevation in Fig. 39. The main points observed in the con-

struction of these boxes are as follows :
The sides are at an angle of at least 50 to the horizontal, to

insuie the uninterrupted descent of the slimes. The pulp is
delivered evenly across the end of the box by means of some
such arrangement as the movable tongues (a, Fig. 39). Surface
currents are prevented and the incoming pulp thoroughly mixed
with the mass of the water in the box by the partition, b, which
extends across the box near the inflow end. The discharge is
made by cutting the other end of the box from 2 to 3 inches lower
than the sides this overflow is made perfectly level so that the
;
water flows out in an even sheet. The discharge pipe for the
pulp is at the bottom of the box. The siphon discharge shown
CONCENTRATION IN STAMP MILLS. 1C9
is the most convenient, as the aperture need not be so much con-

tracted as if direct discharge is used; and, besides this, less fall
in the mill site is required. A
contracted orifice is liable to be
choked up, and a smaller diameter than 1 J to 2 inches is to be
deprecated.
These boxes have the advantage over the ordinary settling-pits-
used to retain tailings and to catch pulp for pan-amalgamation,,
that they do not require to be dug out, while the settling, owing
to the elimination of surface currents, is more perfect. One-
large box of 12 feet by 6 feet at the top and 7 feet 6 inches deep
is sufficient to settle the product from five or ten stamps, accord-
ing to the degree of separation of water and ore required.

If the pulp is required for treatment in amalgamating pans,,
such settling boxes do not give material of sufficiently thick
consistency for the purpose. To adapt them to this use
M'Dermott suggests that the pulp should be allowed to flow
through two small square tanks placed above the pointed boxes,
so as to catch the heavy sand. As soon as one tank is full, the
pulp is turned to pass through the other, while the first is dug
out after a short draining. The slimes are caught in the pointed
boxes, and after mixture with the drier heavy sand the pulp is
not too thin for amalgamation in pans. Concentration before
amalgamation is, however, as yet seldom resorted to.
Classification according to Size. Classification according
to size, although seldom attempted by
mill-managers in the
treatment of battery pulp, is desirable as a preliminary to con-
centration by almost every machine ; these work more efficiently
if engaged in the separation of
equal sized particles of different
specific gravities than if pieces of all sizes are mixed up together.
The sizing machines most commonly employed are of two-
kinds, viz. :
1. Revolving or flat shaking screens, which are suitable

for coarsely ground ores.
2. Hydraulic classifiers i.e., boxes containing ascending
currents of water ;
these are suitable for finely pul-
verised ores.
1.
Sizing by screens is almost always conducted on wet
material, spraying jets of water being used to carry the small ore
through. Flat shaking screens offer advantages in cheapness
and simplicity, but revolving cylindrical screens, inclined at a
slight angle to facilitate discharge, are in more general use. It
is customary to employ screens of the minimum fineness of 8 to-
16 mesh. Where the sizing of finer particles than could be
treated by this mesh is required, hydraulic classifiers are used,,
since their cost in repairs and renewals is far less than that of
screens. The latter are seldom used in the sizing of battery
sands, but are recommended for the purpose by Resales.*
*
Loss of Gold in the Reduction of Auriferous Veinstone in Victoria. By
H. Resales, Melbourne, 1895. This exhaustive aud closely reasoned
treatise will well repay a careful study.
2. Hydraulic sizers were introduced by Prof. P. von Rittinger

more than forty years ago, for use in the Hartz, and from their
shape were known as Spitzkasten or pointed boxes. These boxes
have the shape of inverted pyramids, the stream of unsized pulp
entering at one side and flowing out at the other, whilst there is
also a small discharge at the apex. The current slackening on
entering the box, the heavier and larger particles in suspension
at once begin to settle, and, escaping the influence of the current,
fall quietly to the bottom of the box where they are discharged.
It is evident that in this classification the particles which reach
the bottom will not all be of the same size. A
piece of galena
of specific gravity 7*5 will fall through water at a faster rate
than a piece of quartz of the same size, but only one-third of
the weight. The rate of fall is dependent both on the size
of the particles and on their densities, according to a well-known
law.* By applying this law it is easily shown that in water a
sphere of iron pyrites (specific gravity 4*9) of 1 mm. in diameter
will fall at the same rate as a sphere of quartz (specific gravity
2-6) of 4 mm. in diameter, and a sphere of galena of 1 mm. in
diameter. In each successive box, then, a number of approxi-
mately equal-falling particles are removed, the closeness with
which the subdivision is made varying with the size of the box,
and the corresponding extent to which the current carrying the
pulp is checked on entering it. The larger the box the more
material is collected by it, and, therefore, the more heterogeneous
the particles caught. In a small box the material collected con-
sists more nearly of truly equal-falling particles.
These equal-falling particles however are, in any box, whatever
its size may be, carried away through the aperture in its apex
by muddy water containing material of all sizes down to the

iinest slime, and it was to eliminate this material that the
ascending current was introduced. This is a current of clear
water, which enters at the apex of the box in greater quantity
than can be discharged by the outflow near the same spot, so that
there is an upward current of water into the box. The result is
that no muddy water is discharged below, but only the particles
of ore which have weight enough to drop through the ascending
current, so that by regulating the strength of this, any desired
class of ore can be obtained. In Fig. 40, a form much used in
America is shown. The sliding partition assists the settling, by
causing the pulp to pass downwards, rapid surface currents
across the box to the overflow being thus prevented. The
discharge of the heavy particles is effected through A, the
clear water pipe itself, by the arrangement shown. The launder
above the box supplies the clear water current, and shows the
head of water used, which must be kept constant to ensure
uniformity of results.
*
See footnote on p. 167.
CONCENTRATION IX STAMP MILLS. 171
The chief defect of the early forms of pyramidal boxes was

that, as the area of the vatbecame larger and larger towards the
top, the velocity of the ascending water naturally became less
and less, so that many particles were able to settle down below
the level of the overflow, but were stopped by the increasing
force of the current, so that an accumulation of ore took place
half way up the box, and ultimately became so great as to inter-
fere with the Several remedies have been devised
classification.
which one of the simplest and most effectual is
for this defect, of
to make the box pyramidal below, but with vertical sides in the
Fig. 40.
tipper part. This construction is partly carried out in the box

described above, which, however, though given as a type, is not
theoretically the most perfect of its kind. A slime-pit is added
to catch the stuff which is too light to settle in the boxes, in all
cases in which these slimes are of sufficient value to pay for
treatment. The number of boxes used depends on the number
of classes of material which it is deemed advisable to make.
Usually two or three classes are sufficient, and, as already stated,
in the majority of cases no sizing is attempted.
Early Concentrating Machinery. One of the oldest and
most primitive machines employed in concentration of fine sand
by means of a shallow stream of water was the German huddle,
which has a distinct but imperfect resemblance to the Long-Tom

described on p. 46.* Canvas tables were used below these
buddies in Germany, and probably suggested the use of blanket-
strakes or tables, which were adopted in the early days of the
gold fields of the United States and Australia, and are still
retained in some places. The rough surface of the blanketing
seems to be particularly efficacious in catching and holding thin
plates and spangles of free gold or of sulphides, which are readily
washed off smooth surfaces by a current of water, and these
rough appliances, although almost useless for catching slimes, still
find favour where considerations of economy prevent the purchase
of modern high-priced concentrators. The blanketing is usually
in strips of 16 or 18 inches wide, and several feet long, and is
nailed or stretched on wooden frames, which have an inclination
of about J inch per foot.
At intervals of about half an hour, when a quantity of mineral
has already been collected on the rough surface, the blankets are
taken off and washed in tubs of water, where a deposit collects
which is afterwards dug out. At the St. John del Rey Mine,
the framework supporting the blankets was hung on pivots aboAê
a shallow tank. When it was necessary to clean them, the
framework was turned so that the upper surface of the blanket
was inclined downwards, and the mineral washed off its surface
by a hose, much time and labour being thus saved.
At this mine the trays supporting the blankets were 18 inches
wide and 30 feet long, with a fall of 1 inch per foot. The upper
16 feet were covered with bullock's skins, tanned with the hair
on them, and in lengths of 26 inches below these were a series
;
of blankets or baize cloths of the same length, made of coarse

wool with a long nap. The fall from the battery box upon the
tray was 4 inches, a screen being placed across the end to break
the fall of the water, and cause it to strike the tray nearly at
right angles. About 90 per cent, of the gold contained in the
ore was caught on these blankets. The blanket sands contained
95 per cent, of sulphides, and were so fine that 90 per cent, of
them passed through a 100-mesh sieve. They were amalgamated
in revolving wooden barrels, yielding 96 per cent, of their assay
value, but this was due to the fact that very little gold was
contained in the pyrites, most of it being present in the form of
free particles.
Blanket sluices have been declared unsuitable for catching fine
sulphides, and their concentrates are usually contaminated by
admixture with much sand. It' set at a proper inclination they
will save fine amalgam and free gold, but even in this respect
they are less satisfactory than shaking copper plates and riffles.
*
For a description of the similar buddle formerly used in Colorado for
the treatment of battery sands, see Eaymond's Mines, Mills, and Furnaces
of the Pacific States. New York, 1871, p. 357.
at the same time as blankets for

Riffled sluices were employed
effecting rough concentration.
The riffles were formed of half-
inch strips of wood nailed across the sluice box, the grade of
which was about three-quarters of an inch to the foot. As
soon as the concentrates had accumulated until they reached
the top of the riffle, another strip was nailed on, and the process
was repeated until the bed of concentrates was several inches
thick, when they were scraped out and a fresh start made.
Similar to this device was the raising-gate concentrator, which
was practically a riffled sluice in which the riffle was raised
continuously by machinery, instead of being adjusted at intervals
by hand.
The Hound Buddie was invented in Cornwall, where it is still
used in dressing the tin ores to the exclusion of almost every
other concentrator. There are two varieties.
1. The convex round buddle, in which the ore and water are
added at the centre of the machine, and flow down over the
surface to the periphery.
2. The concave buddle, in which the pulp is added at the
periphery of the machine and flows down to the centre.

In both cases revolving arms, carrying brushes, pass over the
surface and stir the deposit as it is being formed, and the spouts
distributing the ore also rotate so as to deliver the pulp evenly.
The buddies are from 12 to 18 feet in diameter.
These machines are not continuous in action, and after the
deposit has accumulated to a depth of a few inches the operation
is suspended and the deposit dug out. The "headings" or
material on the upper 12 or 18 inches of the inclined surface are
kept separate, and the stuff near the bottom of the slope is called
the "tailings." Eound buddies are not adapted to obtain a
finished product in one operation. The headings and tailings
must, as a rule, be subjected to farther treatment, and between
them is a large quantity of material which differs little from the
original ore. Thus handling and re-handling of the stuff is

necessitated, and it is on this account that these machines are
not now much used on gold ores. The principle on which they
depend is favourable to the collection of slimes, and the modern
improved buddies are perhaps better adapted for fine ores than
any other machines, except those employing a travelling belt.
Modifications have been proposed to adapt these buddies to the
treatment of gold ores by adding riffles containing mercury, and
by other devices, but have not found much favour. One of the
chief changes, which was proposed in the United States, is to
keep the brushes and ore-spouts stationary, and to rotate the
inclined bed. During the earlier part of the revolution an
attempt is made to separate the pulp into headings, middlings,
and tailings, and each of these is washed into a separate launder
by strong iets of water before the completion of one turn, the
table being left clean for further work. A

saving of labour in
digging out the material is thus effected. The inclination of a
buddle is usually from 1 to 1J inches per foot. device A
formerly much in vogue in Western America consisted in the
use of a cylinder of iron in the centre of a concave buddle,
through which the discharge took place. This cylinder was
made to rise slowly as the concentrates accumulated on the
buddle, the effect being similar to that obtained in the sluices
with self-raising gates.
For working large quantities daily of low grade ore containing
a high percentage of pyrites, the revolving buddle is still used.
When the percentage of pyrites is low, or if the ore is rich, so
that closer work is desirable, or if the slimes are worth catching,
revolving-belt tables are substituted for the older machines. The
advantages of the revolving buddies consist in low first cost,
large capacity, and ability to stand large quantities of water,
which are always present in the pulp after the operations of
screening and hydraulic sizing.
Centrifugal Concentrators. Among concentrators with
which were apparently suggested by the
circular revolving beds,
revolving buddle, may be mentioned Hendy's and Duncan's
machines. Hendy's concentrator was one of the earliest forms
employed in California, having been patented in 1868, and is
stillretained in some mills. It consists of a shallow cast-iron
centre by a
pan, 5 feet or 6 feet in diameter, supported in the
vertical shaft. A
rapid horizontal oscillating motion is given to
the pan by the revolution of a crank shaft, which is joined to

the periphery of the pan by a connecting-rod. The bottom of
the pan slopes gently downwards from the centre to the peri-
phery, and there is a basin-shaped depression in the centre,
and
an annular gutter running round the outside. The pulp is fed
into the pan near the periphery, and a depth of about 3 inches
of material is maintained. In consequence of the rapid oscillating
motion the heavy particles move outwards by centrifugal force
and accumulate in the peripheral gutter, while the lighter
particles are discharged into the central circular
basin and are
removed by opening discharge gates at intervals.
The Duncan concentrator resembles the Hendy pan in con-
sisting of a circular iron pan, in which the heavy sulphides
are
moved to the periphery by centrifugal force. The pan, however,
revolves, making 8| turns per minute, while the pulp is supposed
to make about 3 revolutions in the pan, thus allowing time for
the pyrites and gangue to separate before the former settles at
the bottom towards the sides, where it remains until discharged.
The lighter particles remain more or less in suspension in the
water and are carried to the centre, where their removal from
the pantakes This machine is used to some extent in
place.
California, and, like the Hendy pan, is effective in the separation
of coarse sulphides. Neither of these concentrators do good

work in saving slimes, for the reason that the movement of the
concentrates towards the periphery, depending on centrifugal
force, takes place in consequence of the superior mass of the
heavy minerals. With slimes, the mass of the particles of pyrites
is so small that it is not enough to enable them to overcome the
effects of the stream of water which carries them towards the
centre, and they are, consequently, swept away with the tailings.
Percussion Tables. In these machines the work of keeping
the pulp in a state of agitation, done by the rakes or brushes-
in the German and Cornish buddies described above, is effected
by longitudinal shakes imparted to the table. The table is
made of wood or sheet iron, the surface being as smooth as-
possible, and the sides being flanged. It is hung by chains or
in some similar manner, so as to be capable of limited movement,
and receives a number of blows delivered on its upper end.
These blows are given by cams acting through rods, or else the
table is pushed forward against the action of strong springs by
cams on a revolving shaft, and then being suddenly released is-
thrown back violently by the springs against a fixed horizontal
beam. The movement of the pulp depends on the inertia of the
particles, which are thrown backward by the blow given to the
table, the amount of movement varying with their mass, and
depending, therefore, both on their size and density. The vibra-
tions produced by the percussion also perform the work of the
rakes in destroying the cohesion between the particles, and a
stream of water washes them down. The result is that the
larger and heavier particles may be made to travel up the table
in the direction in which they are thrown by the blow, by
regulating the quantity of water, while the smaller and lighter
particles are carried down. It is obvious, however, that the
inertia of the fine particles of heavy minerals will not be sufficient-
to move them against the current, and, in consequence, all the
slimes, both rich and poor, will be carried down and lost in the
tailings.
" Gilt
Gilpin County Edge Concentrator" This variety of
shaking table was devised in Colorado, and has displaced the
blanket sluices at almost all the mills at Blackhawk. It consists
(Fig. 41) essentially of a cast-iron or copper table, 7 feet long and
3 feet wide, divided into two equal sections by a 4-inch square
bumping-beam. The table has raised edges, and its inclination is
about 4 inches in 5J feet at its lower end, the remaining 1 J feet
at the head having a somewhat steeper grade. The table is
hung by iron rods to an iron frame, the length of the rods being*
altered by screw threads, so as to regulate the inclination to the
required amount. A shaft with double earns, A, making 65
revolutions per minute, enables 130 blows per minute to be
given to the table in the following manner; on being released
by the cam, the table is forced forward by the strong spring,

B, so that its head strikes against the solid beam, G, which
is firmly united to the rest of the frame. The pulp corning
from the copper plates is fed on to the table near its upper
end by a distributing box, D, and is spread out and kept in
agitation by the rapid blows. The sulphides settle to the bottom
of the pulp, and are thrown forward by the shock, and eventually
discharged over the head of the table at the left hand of the
figure, while the gangue is carried down by
the water and dis-
charged at the other end. One machine is enough to concentrate
the pulp from five stamps. If the table consists of amalgamated
-copper plates, it is of some use for catching free gold also,
treating about 8 cwts. per hour. This machine, like other
Fig. 41.
percussion tables, is very effective for separating coarse pyrites,

but almost all the slimes are lost in the tailings. This is of
less importance, as the sulphides with which it has to deal
are not of high value, containing usually only from 10 to 12
dwts. of gold per ton. The high price of the Frue vanner and
other slime tables prevents their introduction on this field, as
they cost more than three times as much as the Gilt Edge con-
centrator for equal capacity.
The Frue Vanner. This machine is described in detail as
being typical of the shaking travelling-belt concentrators. Brief
accounts of some other forms are given subsequently. Machines
of this class are especially adapted for treating finely-crushed
battery sands which do not contain a large percentage of
" mineral"
(that is, sulphides and other heavy materials). They
.are frequently set to concentrate unsized pulp
coming straight
from the amalgamating tables, and although they often do good
work under these circumstances, when the screens used are
equal to 40-mesh sieves or finer, nevertheless if the pulp con-
tains coarse stuff the results are not so satisfactory. Rosaies
recommends * that the tailings from the plates should in general

be passed over percussion tables, and should then be separated
into four classes, viz. :Coarse sand, medium sand (which is
retained on a 40-mesh sieve), fine sand (which will all pass
through a 40-mesh sieve), and slimes. The coarse sand may
often be thrown away as worthless, although, if necessary, it is
reground, and the other classes may be advantageously treated
separately on some form of travelling-belt table or round buddle.
The Frue vanner (Fig. 42) consists essentially of an endless
rubber belt, mounted on a frame, with its upper surface slightly
inclined to the horizon, and subjected to two movements, a slow
constant longitudinal movement and a slight and rapid side shake.
The belt forms the bed or plane on which the dressing of the ore
is effected, being an inclined plane, 12 feet long, and bounded
down the two sides by projecting rubber flanges, which prevent
the water and sand from dropping over the sides. An arrange-
ment of rollers permits of the belt being slowly revolved in the
direction of its length and up the incline; thus, though the
dimensions of the working plane remain always the same, its
surface is constantly travelling. The crushed rock in a stream
of water is delivered near the upper end of the belt by means of
the sand distributor, No. 1, Fig. 42, and flows down the belt
towards its lower end. Now, as the inclination at which the
belt is set is only from 3 to 6 inches on the 12 feet, and as
the stream of water is not large and spreads over the whole
width of 4 feet, it is obvious that, if it were not for the
movements of the belt, much of the crushed rock contained in
the water would settle on the belt, while the water and the finer
and lighter particles of sand would alone reach the foot of the
table and drop over into a waste launder.
In order to separate the heavy metallic minerals from the
accompanying gangue or rock, a second stream of water is
applied, whilst a gentle side shake is given to the belt, to keep
the sand in a state of agitation, prevent it from " packing," and
facilitate the sorting process. The water distributor, 2, is placed
about 1 foot above the pulp distributor, and delivers small jets
of water, 3 inches apart, over the entire width of the belt. The
side shake thoroughly mixes this water with the pulp, spreads
the whole uniformly on the belt, and enables the heavy particles
of mineral to settle through the sand and cling to the belt, when
they are carried up by it past the small jets of water and
deposited in the collecting tank, while the lighter gangue is
carried down by the stream and delivered into the tailings
launder.
The Frue vanner is shown in plan in Fig. 42a, in side eleva-
tion in Fig. 426, and in end elevation in Fig. 42c. The following
description is abridged from that given by the manufacturers :
* Loc. cit. ,
AA are the main rollers that carry the belt and form the ends
of the table. Each roller is 50 inches long and 13 inches in
diameter, and is made of galvanised sheet iron. B and C are of
the same diameter, and are made in the same way as A
A. The
roller part of is shorter than that of AA and B, and also has
rounded edges, the upper surface of the belt with its flanges
passing over it. The belt E passes through water underneath B,
depositing its concentrations in the box, No. 4 and then, pass-
;
ing out of the water, the belt, E, passes over C, the tightening
roller. By means of the hand screws, B and C can be adjusted
on either side, thus tightening and also controlling the belt.
The boxes holding AA in place have slots and adjusting
screws, so that, by moving them out or in, AA can be made to
exercise an influence on the travel of the belt, E ; when, as
sometimes happens, it travels too much towards one side, this
tendency can be stopped most quickly by altering the screws
on one end or the other of AA ]
the change of position of B or
C also controls the belt.
D D, &c., are small galvanised iron rollers, which support
the belt, E, and cause it to form the surface of an evenly
inclined plane table. This moving and shaking table has a
frame, F, of ash, bolted together, with Aand A
as its ex-
tremities. The frame isbraced by five cross-pieces.
The belt, E, is 4 feet wide, and 27 \ feet in entire length ; it is
an endless belt of rubber with high flanges at the sides.
G G is the stationary frame. This is bound together by three
cross-timbers, which are extended on one side to support the
crank shaft, H.
F is supported on G G by uprights, N, &c., four on each
side. The bearings of A, the upper or head roller, are higher
than those of A, the foot roller, so that the former is a trifle
higher than the regular plane of the table, and the first small
roller, D, should be raised by a corresponding amount.
The shape of the lower or bottom bearings of the uprights, N,
&c., can be understood by examining b, as shown in the end
elevation and also partly in Fig. 426. This lower bearing, b,
extends across G, underneath, and is supported by a bolt passing
through G. A lug on the upper side and on the outside end of
b rests on G ; and b hangs on the head of the bolt, and is kept
stationary by the weight of N and its load. By striking with a
hammer the face of b shown in the elevation, b is moved, chang-
ing the position of the lower bearing, and thus making N
more
or less vertical. By thus moving the lower supports of N, &c.,
the sand corners on the belt hereafter explained are regulated.
The cranks attached to the crank shaft, H, are J inch out of
centre, thus giving a throw of 1 inch, which is the amount of
the lateral throw. I is the driving pulley that forms with its
belt the entire connection with the power. J is a cone pulley
on the crank shaft, H. By shifting the small leather belt

connecting J and W, the uphill travel of the main belt, E, is
increased or diminished at will; the pulley, W, is moved by
the hand-screw, m. R, R, R are three flat steel spring connec-
tions bolted underneath the cross-pieces of the frame, F, and
attached to the cranks of the shaft, H. These springs give the
quick lateral motion about 200 per minute.
No. 2 is the clear water distributor, and is a wooden trough
which is supplied with water by a pipe, and the water discharges
on the belt in drops through grooves 3 inches apart.
No. 1 is the ore-spreader, which moves with F, and delivers
the ore and water evenly on the belt.
There is also a copper well that fits in (and shakes with) the ore
spreader as shown in the drawing. This is used in concentrating
gold ores, for saving amalgam and quicksilver escaping from
the silvered plates above, and can be taken out and emptied at
any time. Into this well falls all the pulp from the battery.
Its ends are lower than the wooden blocks of the spreader, so
that the pulp passes over the ends of the well and is evenly
distributed.
For some gold ores it is desirable to use on the ore-spreader a
silvered copper plate the size of the spreader, and, when this is
used, the wooden blocks of the spreader are fastened to a movable
frame on top, so that they can be removed when the plate is
cleaned-up, once or twice a month. No. 4 is the concentration
box, in which the water is kept at the right height to wash the
surface of the belt as it passes through. No. 8 is a section of
the launder to carry off the tailings.
Method of Working. The ore is fed with water on the belt, E,
by means of the spreader, No. 1, which distributes it uniformly.
A small amount of clear water is added by No. 2. The depth of
the sand and water is kept constant at from f to J inch.
The main shaft, H, is given the proper speed for each kind of
ore this usually varies from 180 to 200 revolutions of the crank
;
shaft per minute, the former speed being for light, fine " slimes,"
the latter for somewhat coarser materials. The effect of increas-
ing the speed of the side shake is to increase the percentage of the
material which is being discharged as tailings, and so to tend to
loss of pyrites. The diminution of the speed, on the other hand,
tends to the production of concentrates containing much sand.
The speed of the uphill travel of the belt varies from 2 to 12
feet per minute, and the grade or inclination from 3 to 6 inches
in 12 feet, according to the ore. If the ore treated be poor in
pyrites, the upward motion of the belt should not exceed 20 inches
per minute ; if richer, the speed is increased accordingly, and in
agreement with the inclination of the belt, being greater as this
inclination increases, but usually not exceeding 3| feet per
minute. The inclination can be changed at will by wedges at
CONCENTRATION IN STAMP MILLS 181
the foot of the machine, these wedges being under the lower end
of G, G, and resting on shoulders of uprights from the main
timber of the mill. The motion, the water used, the grade, and
the uphill travel is regulated for every ore individually, and
must unfortunately be adjusted with every change in the pulp,
if good work is to be maintained.
In treating ore coming direct from the stamps, too much
water may possibly be present in the sand for proper treatment
by the machine. In such a case there should be a box between
the stamps and the concentrator, from which the sand with the
proper amount of water can be drawn from the bottom, whilst
the superfluous water will pass away from the top of the box ;
but as sulphides will also pass away with this water, settling
tanks should be provided, and the settlings can be worked from
time to time as they accumulate.
The main body of the belt suffers hardly any wear at all,
since it merely moves its own weight slowly around the freely
revolving rollers the life of the belt is lengthened by taking
;
the precaution of keeping it clean from sand at every point

except the working surface, so that sand cannot come between the
belt and the various rollers.
The concentration box, No. 4, which is kept full of water, and
through which E passes, may be of any size or depth desired.
Though not indispensable, it is best to have a few jets of water
playing above and underneath on the belt as it emerges from the
water in No. 4, so as to wash back any fine material adhering to
the belt, and as such a method will cause an overflow in No. 4,
the waste water, being full of finely divided mineral, should be
settled carefully in the boxes, Nos. 7, 7, 7. Every few hours the
concentrations may be scraped out with a hoe into the box, No.
9, and if this box be on wheels, it can be readily run on a track
to the place where the concentrations are stored.
The amount of water used on the machine is from 1 to 1J
gallons per minute of clear water at the head, and from 1 J to 3
gallons per minute with the pulp.
The capacity per day of twenty-four hours is usually put
down as about 6 tons of material, fine enough to pass a 40-mesh
screen, but the machine does better work
when set to treat a
smaller amount. In California, in general, two Frue vanners
treat the product of each battery of five stamps. Where the
is one Frue vanner is sometimes used for five
gangue light, only
the pulp
stamps. Sizing of the material is frequently omitted,
direct from the stamps to the copper plates, and thence
passing
to the vanners. however,
Resales, out * Chat this is a
points
mistake the coarse sand should be removed by sizing, as other-
;
wise it interferes with the successful conduct of the work, caus-
*
Reduction of Auriferous Veinstone in Victoria, 1895.
ing serious losses of sulphides. If there are large quantities of

slimes in the pulp, the capacity of the machines is proportion-
ately reduced. The machine requires careful watching by a
skilled workman, as any change in the condition of the pulp
necessitates readjustment of the belt.
It is very important to use the proper quantity of water
with the pulp from the stamps, and this should be carefully
regulated. There should be formed on each side of the belt a
slight corner of sand i.e., there should be, on each side, sand
with less water in it than there is in the remainder of the pulp
on the belt. If there is not a slight sand corner, the corner will
be sloppy, and there will be a loss. Sloppy corners are caused
by using too much water with the pulp from the stamps passing
on to No. 1.
Frequently, on the other hand, there may not be enough water

with the pulp from the stamps, and as a result the sand corners
formed will be too wide. The remedy for this is to use more
water in the pulp coming on to No. 1.
As regards the proper amount of water to be used in the
water spreader, No. 2, use just enough (no more) to keep
the field between No. 1 and No. 2 covered, so that no points
(or fingers) of sand shall show on the surface. The whole width
of the belt between the water spreader and ore spreader should
be kept quite wet. If dry streaks or points occur, and water,
as a consequence, runs in streaks at the junction of the wet and
" "
dry channels, sulphides will be picked up and floated away on
the surface of the water; this "floating" of pyrites is caused
by its dryness, not by its lightness ; it has been coated with a
film of air. When the proper amount of water has been fixed,
the carrying over of the clean concentrates past the jets of No.
2 should be accomplished and regulated by the uphill travel only.
Frequently, the sand and water on the belt will be distributed
unevenly, the sand working to one side of the belt, and making
a broad heavy corner, while the other is sloppy. This is caused
by the belt not being horizontal from side to side ; it is levelled
by raising or lowering one side, by altering the position of the
supports, N.
W. M'Dermott makes the following observations on the
appearance of the ore on the belt :
" Should the

discharge of concentrates exceed the quantity of
sulphides falling on the belt, sand or rock will be found close
up to the jets of water, and by and by passing them. If the
uphill travel be too slow, the sulphides collect below No. 2, form-
ing a great "head," extending towards No. 1, and even below,
in which latter case an increased loss of sulphides will assuredly
take place in the waste. When working properly, a small head
is always kept below the jets of clear water, and the sulphides
come over clean and regularly."
Riffle Surfaced Belts for the

Frue Vanner. The smooth belt
vanner, just described, is essentially a slime machine adapted to
the very finest material. The special value of the vanner in
recovering fine sulphides without previous sizing has been illus-
trated as follows :
*
"If a flat heavy object say a sixpence is put on a Frue
vanner while in operation, it will not pass over with the fine
material steadily delivered past the water jets. The shaking
motion, by cutting away the supporting adhesion of the belt
surface, owing to the inertia of the coin, allows the light current
of water on the edge of the coin to wash it backwards with the
sand. The same action applies to coarse particles of sand mixed
with fine particles of mineral. Thus the effect of mere mass is
such that the Frue vanner with smooth belt is not adapted to
saving coarse mineral ; an excess of wash water at the head
drives down the coarser mineral, but the finest clings to the
belts and safely passes the water jets."
In order to overcome this defect a roughened belt was intro-
duced, having a number of depressions or riffles on its surface.
In moving down this belt the coarse sulphides and gangue pass
into these depressions, and then are compelled to proceed over a
rising surface. At these points the separation of the coarse
sulphides and coarse sand is perfectly performed, the heavy par-
ticles remaining in the depressions, while the lighter are washed
away. The belt then in moving upwards carries the pyrites
with it. In other respects the machine closely resembles the
older smooth belt form, to which it is inferior in its power of
catching slimes, although better adapted for coarse material.
The riffle-belted Frue vanner is worked at a greater inclination,
with a faster upward motion, a slightly faster shake, and with
more water than the ordinary vanner, and consequently it
treats more pulp, one machine usually taking the product of
five stamps. It offers great advantages over the smooth belt
whenever the percentage of sulphides is very high, as the riffles
and carry up much more material.
collect
The Embrey Concentrator. This machine, like the Frue
vanner, consists of an endless belt with flanges along its edges ;
it differs mainly in
having an end-shake instead of a side-shake,
the motion being parallel to the length and travel of the belt,
instead of at right angles to it. It is shown in Fig. 43, where
the light wooden-framed form is shown. This is cheaper but less
compact than the iron-framed variety, and is recommended for
use wherever floor space is not of the first importance. It is
unnecessary to describe it in detail, as it differs little from the
Frue vanner. It is necessary to shake this machine at a faster
rate than the side-shaking machines, the usual speed being from
230 to 240 revolutions per minute, although 200 per minute is
*
Gold Amalgamation, p. 71. M'Dermott and Duffield, New York, 1890.
occasionally enough. If an attempt is made to treat an ore ab

too low a rate of speed, complete separation cannot be effected.
More water is used than on the Frue vanner, while the inclina-
tion and rate of upward movement of the belt are both greater.
In consequence of this, more pulp is treated, the amount varying
from 6 to 10 tons per day of twenty -four hours, and, therefore,
either one or two machines are used to each five stamps, accord-
ing to the character of the ore and the closeness of concentration
required. Fewer Embrey concentrators than Frue vanners are
in use, owing to the fact that the sale of the latter is pushed
more by the company which owns both the patents.
Among other concentrators with endless inclined rubber belts
may be mentioned the Triumph, which, like the Embrey, has a
longitudinal shaking motion applied to the belt, the crank shaft
revolving at the rate of 235 to 240 times per minute. The
Fig. 43.
forward motion of the belt is regulated by a friction roller,

instead of a cone pulley, which is used in the Frue vanner and
the Embrey concentrator. It has also an amalgam saver, con-
sisting of an iron trough containing quicksilver, which is stirred
by iron teeth attached to a slowly revolving horizontal shaft.
It is in wide use in California (where it seems to be preferred to
the Embrey machine) and in Australia.
The Liihrig Vanner. This machine was invented about
five years ago, and has already met with great success in many
instances in the concentration of auriferous tailings from stamp
batteries. It bears a considerable resemblance to the Frue
vanner and is thought by many experts to be destined to
supersede it. In the Liihrig vanner the endless travelling
india-rubber band is not flanged at the sides, and has a slight
side inclination, which is, therefore, at right angles to the
direction of travel ; the latter is horizontal, not inclined ; by
an arrangement of cams and springs, end-blows are given to the
framework carrying the belt. The pulp from the batteries is

collected in settling boxes placed overhead and delivered on to
the belt through a small distributing box situated above its-
right-hand upper corner, the belt being driven from right to left.
Clear water is supplied through a perforated pipe fixed diagonally
across the belt. The pulp moves across the belt from the higher
side to the lower, this motion being assisted by the clear water ;
the light particles of gangue are washed down in this direction
at a faster rate than the heavy particles of pyrites. On the
other hand, the travel of the belt and the end-blows move the
ore in the direction of the length of the belt. The results of
the combined motions are as follows :
1. Tailings pass off the table

nearly opposite the distributing
box at the right-hand end.
2.A middle product, containing both sulphides and gangue,
is delivered near the middle of the side of the belt.
3. Clean concentrates are delivered near the left-hand end of
the belt, having travelled the greatest distance before being
washed off at the side.
Each of these three products is delivered into a separate hopper,,
and by a simple arrangement of sliding plates the exact points
on the belt at which the delivery of the middle product is
divided from those of the headings and tailings can be altered
so that the percentage of each product can be regulated.
Details of the dimensions, capacity, &c., of these machines are
as follows:
The india-rubber belt is about 4 feet wide and 19 feet in

total length ; the belt travels from 1 8 to 20 feet per minute ;
the total quantity of clean water used is from 4J to 5 gallons
per minute; the supporting framework is 5 feet 6 inches high,
6 feet broad, and 16 feet long; the number of percussions is
from 150 to 210 per minute; the extent of motion to and fro is
from | inch to 1J inches, according to the nature of the ore.
The amount of ore capable of being treated in twenty-four hours
is stated to be 4 to 5 tons, if the concentrates are iron pyrites,
and 3 to 4 tons if they are galena and blende. At the May
Consolidated Mine, 7 to 8 tons per day were being treated on
each machine in 1893. At a test run in Glasgow, 2,898 Ibs. of
a pyritic gold ore were treated in four hours five minutes, and
the re-treatment of the "middlings" occupied one and a-half hours
more this is equal to a rate of 6 J tons per day. In the com-
;
pound system the middlings from several machines are delivered

at once to a vanner placed at a lower level, and time thus saved.
The power required is given by the Liihrig Company as y1^ H.P.
per table, and by Mr. H. D. Griffiths, A.R.S.M., of Johannesburg,
as | H.P. in the case of the machines at the May Consolidated
Mine.* Three single tables are required for each battery oi five
*
South African Mng. Journ., August 19, 1893.
stamps on the Witwatersrand, this being the same as in the case

of the Frue vanners on the same field.
In comparing the Liihrig vanners with other travelling belt
machines, the chief points to be noted are that in the former
case :
1. The direction of travel is horizontal.

2. Three or more different products are drawn off continuously
instead of only two.
3. The water and pulp flow across the belt instead of in the
direction of its length.
The machine resembles the Frue vanner in that alterations
are required in the inclination of the belt, the rate of travel,
the number of blows per minute, the force of each blow, and
the amount of clean water and pulp, with every change in the
constitution of the ore, so that the machines require constant
supervision. The regulation of these is, however, very simple
in the Liihrig vanner.
The results obtained in various countries have been excellent.
The following is the result of a trial run on an American gold
ore from the Eureka and Excelsior Mine, Oregon :
Hartz Jigs. These machines differ in principle from all. those

previously described, inasmuch as the particles are separated by
their fall through a somewhat deeper column of water than is
the case on inclined tables, while a series of blows from below,
causing waves moving upwards, continually brings the particles
into suspension, and allows them to drop again. The initial
period of the fall in water, during which the motion depends
chiefly on density, is thus continually reproduced, and the result
is a perfect separation of heavy from light particles of ore when
working on any materials except the finest pulp. Jigs consist

of sieves supporting beds of ore, which are completely immersed
in water; the ore is raised and allowed to fall by a quick
succession of currents of water caused by the sudden action of
a piston below, which is so worked that the upward movement
or pulsation resembles that produced by a blow, while the
downward movement is gradual. Under these conditions the
heavy particles work downwards and pass through the sieve,
while the lighter gangue is carried away horizontally by a
stream of water introduced either from below or from above.
Such machines are especially suitable for coarse ores.
In the Hartz jig a layer of coarse heavy particles are spread
on the sieve to prevent too much of the ore from passing
through. The stuff is fed in regularly at the head of the
jig, and the strokes of the piston raise both the bed of heavy
particles and the ore. The heaviest grains of ore find their way
during the downstroke into the interstices of the bed, gradually
pass through it, and coming to the screen, fall through into the
tank below. The lighter particles cannot descend, and are
gradually washed over the end partition by the continuous
supply of water. Two products are, therefore, given, neither
requiring further treatment if the conditions are favourable, and
the machine properly adjusted. The wire meshes of the screen
are always much larger than the ore treated, and the bed is
composed of material of as nearly as possible the same density as
the concentrates to be obtained, and is usually from J to 1 inch
in depth. The number of strokes of the piston per minute is
from 60 to 80 with coarse sand of ^
inch in diameter, and 200,
The
300, or even 400 with very fine sand, approaching slimes.
length of stroke varies under the same conditions from to inch,
and in the case of very fine, almost impalpable sand, the stroke
may be diminished till it becomes a mere tremor. Some of the
highest authorities on concentration have stated their belief that
for enriching even very fine sand, the Hartz jig is the simplest
and most economical machine yet invented, and requires the
least amount of labour.* It is obvious that it is not adapted to
save rich slimes, and up to the present it has not been widely
used for treating gold ores.
*
Gallon's Lectures on Mining. Eng. edition, vol. iii., p. 103.
Pneumatic Jig. In this machine, which was devised by

S. R. Krom, air is used instead of water for the separating
medium. The ore is fed into a tank, the discharge from the
bottom of which is regulated by a revolving wheel. The puffs of
air are delivered through a number of vertical wire gauze tubes,
which pass through the ore bed nearly to its top. No less than
450 to 500 puffs of air per minute are given, and the result is
that the heavy particles alone descend through the interspaces
between the wire gauze tubes, while the lighter material is dis-
charged horizontally. It is necessary for the success of this
machine that the ore should be quite dry, so that the particles
may be completely independent of each other as a rule the ore
;
would have to be dried in special furnaces. The air jig has not
passed into any extensive use, although it might succeed in hot
climates, where water is scarce, and where only a pneumatic
machine could be used. Careful sizing is necessary as a pre-
liminary to successful working of this machine.
Clarkson & Stanfield's Concentrator. Another dry con-
centrator, although one constructed on a totally different principle,
is the centrifugal machine, which was devised by T. Clarkson &
R. Stanfield. The action of this machine is based upon the
joint operation of the three powers centrifugal force, atmo-
spheric resistance, and gravitation and the machine some-
what resembles a " Catherine-wheel " working horizontally.
Pulverised ore is fed on to the surface of a rapidly rotating disc
about 20 inches in diameter, provided at the periphery with a
raised rim, which is perforated by a multitude of small radial holes,
through which the ore is thrown by centrifugal force. The par-
ticles of rich and heavy ore, by reason of their superior inertia,
are thrown to a greater distance, while the worthless particles,
being lighter, are more quickly overpowered by the forces of
atmospheric resistance and gravitation, and thus fall short.
The ejected ore is then collected in annular compartments
arranged concentrically round the central disc, the process going
on continuously while the ore is being shot out. Means are
also provided for regulating either the centrifugal force or the
atmospheric resistance, according to the nature of the ore.
It is stated by the inventors that in one of these machines,
5 feet in diameter, 50 tons of ore can be concentrated in twenty-
four hours, only 3 H.P. being required. In this machine, the
centrifugal force is proportional to the mass of the particles,
whilst the atmospheric resistance is proportional to their cross
section. It may be shown, therefore, that all particles for which
the product a D (diameter x density) is approximately the same
will fall into the same compartment. But the value of a D is
nearly the same for all particles that are equal-falling in air ; for
in this medium, d is so small in comparison with D (see footnote
on p. 167), that it may be neglected, and, therefore, the formula
a(D - d) becomes a D. Now, the difference in size between

equal-falling bodies of different specific gravity is less for air
than for water ; for example, the diameters of spheres of galena
and of quartz, which are equal, falling in water and in air, are in
the ratio of 1 4'1 and of 1 2-9 respectively.
: : Hence it would
follow that, according to theory, this machine would be less
effective as a concentrator than the wet sizing machine described
on p. 170. However, the inventors have obtained good results
in trial runs on both a large and a small scale.
A complete mill on the Clarkson-Stanfield system has been
erected at the Castell Carndochan Mine, near Bala, North
Wales, for the treatment of a low-grade gold ore, containing
about 3 dwts. gold per ton. The mine owners have reported
that, in the treatment of a parcel of 700 tons of ore, about
12 tons of rich concentrates of two grades were obtained, con-
taining about 72 per cent, of the gold in the ore. The total
cost of drying, milling, classifying by sieves, concentrating,
cartage, &c., was certified as being 5s. 6 -83d. per ton. The
fuel used in drying the ore was 40 Ibs. of coal per ton, and
the motive power is water, and therefore costs little. The
concentrates are sold to smelters.
Samples of ore from Coolgardie have also been successfully
treated, very rich concentrates and tailings suitable for cyanid-
irig being obtained.
CHAPTER X.
STAMP BATTERY PRACTICE IN PARTICULAR

LOCALITIES.
Battery Practice in California. The ores treated are quartz-

ose, containing from 3J dwts. to 1J oz. of free gold per ton, the
average being from 10 to 12 dwts. Some iron pyrites is also
present, sometimes containing traces of other sulphides. The
amount of pyrites varies from 1 to 5 per cent. ; it is usually
massive, and contains from 4 to 4J ozs. of gold per ton. The
pyrites is in somewhat rare instances crystalline, and is then
of little value. The gold is seldom visible, and is in such a
finely divided state that it cannot be saved by mere gravitation
methods, the use of mercury being essential.
The obsolete practice formerly pursued at Grass Valley, and in
other districts of California, has been fully described by G. F.
Deetken.* Up to a recent date it had not been entirely super-
seded,! and may be briefly described as follows The ore is fed
:
into the batteries by hand, and no rock-breakers are used. The

stamps weigh 800 Ibs., and have a fall of 10 inches, the depth
of discharge being only 3 inches. No mercury is added to the
battery. After crushing, the pulp is run over a set of blanket-
strakes, where the ore is subjected to a rough concentration.
The tailings are passed through the so-called Eureka rubber,
where the "rusty" gold, ground between sliding plates of iron,
is brightened and rendered amalgamable. Subsequently the
sand is run through narrow sluices, lined with amalgamated
copper plates, where an effort is made to catch the fine free gold,
although the depth and speed of the current renders the opera-
tion very incomplete. The tailings are then concentrated in
round buddies, tossing tubs, &c., and the concentrates subjected
to chlorination.
The blanket concentrates are treated in an Atwood's amalga-
mator, in which they are made to pass through two mercury wells
or baths, being forced under the surface by revolving paddle
wheels, and then they are treated with the ordinary tailings in
the Eureka rubber. The skimmings of the mercury wells are
treated in a small slowly revolving pan, the Knox pan, in which
*
Mineral Resources West of the Rocky Mountains, 1873, pp. 319-345.
t According to the latest reports, this method has now been abandoned
everywhere.
STAMP BATTERY PRACTICE. 1Q1
they are ground with mercury between iron mullers and dies for
several hours, and the tailings from it are roasted and chlorinated.
The whole process bears a strong resemblance to the present
Australian practice on free-milling ores containing coarse gold.
It was uneconomical, costing about $2 or 8s. per ton, without
reckoning the cost of chlorinating the concentrates, and only
from 70 to 75 per cent, of the gold was extracted.
The present method of treatment* consists briefly in rapid
crushing in narrow mortars with heavy stamps working fast with
a low drop, the depth of discharge being moderate. Mercury is
fed into the mortar box, and the gold is saved on copper plates
situated both on the inside and the outside of the mortar, whilst
the tailings are concentrated on shaking tables. Mercury wells
and the various accessory amalgamating machines formerly
attached to most mills are now falling into disuse, and do not
form part of the most approved modern machinery.
The stamps weigh from 750 Ibs. to as much as 1,100 Ibs., the
later practice being to make them of a weight of not less than
950 Ibs., whilst the height of the drop has been gradually
diminished, until the average is now only about 6 inches.
According to the most modern view, from 4 to 4J inches is
quite enough for ordinary ores ; this height was first used at
the Pacific Mill of the Plymouth Consolidated Mining Company,
some three years ago, but other companies have been quick to
follow their example. The heavy low-drop stamp, delivering over
100 blows per minute, has a duty much higher than the more
old-fashioned stamps, which still form a great majority of the
3,500 head now in use in the State. The softer the ore the
lower the drop should be, the limit being passed only when a
good splash of the pulp can no longer be obtained, or when the
momentum becomes insufficient to break down the larger lumps
of quartz rapidly.
The mortars are narrower and less roomy than those in use in
Colorado, and the depth of discharge is kept constant at about
6 inches. Below the screen is a single amalgamated copper plate
about 4J inches wide, inclined outwards at an angle of about
45. Mercury is fed into the mortar at intervals of one hour.
The Blake rock-breaker and some form of ore feeder are in
almost universal use ; among the varieties of the latter, the
Champion and Tulloch's machines are most favoured, although
Stanford's and the roller feeders are still to be seen. The
screens are inclined outwards at an angle of about 10, and
consist either of Russia sheet iron or of brass wire cloth. The
horizontal burr-slot screen is perhaps that most in use, but
others are equally suitable for certain classes of ore. The size
*
The following a brief digest of the full account given by J. H.
is
Hammond, M.E., "The Milling of Gold Ores in California,'*
entitled
published in the Eighth Report of the California State Mineralogist, 1888.
of slot most in use is No. 6, which is -027 inch in width, with

about 180 holes to the square inch. The stamping is f uer with
low grade ores, because the gold in these is in a finer state of
division, and the object of the stamping is to release the particles
of gold from the rock in which they are enclosed, without re-
ducing the ore to an unnecessarily finely divided condition. The
crushing is kept coarse if the sulphides contain an appreciable
percentage of the gold, since the finer the stamping the more
difficult close concentration becomes, and, as the sulphides con-
stitute the most brittle portion of the ore, they are always
reduced to a finer state of division than the gangue in which
they are enclosed. In general, it is better to crush too coarsely
than too finely, for if the latter error is fallen into, not only is
the output diminished and a larger proportion of the sulphides
reduced to slimes, with the result that more is lost in the tail-
ings, but the particles of gold are supposed to be subjected to
repeated hammering, and converted partly into non-amalgamable
and partly into "float" gold, and so lost in great measure. This
question of over-stamping has already been discussed, p. 140, and
an opinion there stated that the evils caused by it have been
greatly overrated. Most of the ores are crushed much finer
than might be supposed from the size of the meshes of the
screen. If the ores are made to pass through a 30-mesh screen,
it is found on trial that more than 80 per cent, of the material
will usually pass through a 60-mesh screen also, and 50 per cent,
will pass through a 120-mesh screen, much of it being impalpable
slime. The low drop, already referred to, acts in the direction
of minimising the quantity of slime, and has a most beneficial
effect in facilitating the concentration of the sulphides.
The high speed of the stamps, besides increasing the crushing
capacity, is advantageous on account of the good splash thus
created, which is beneficial to the battery amalgamation, besides
being essential to rapid discharge. The duty of the stamps is
on an average 2J tons per head per day, but is considerably
more in the most modern mills. It is lower on clayey or on very
hard ores, and cannot conveniently be increased beyond about
4 tons, since the battery is an amalgamating as well as a crush-
ing machine, and the greatest output may not be compatible
with the highest percentage of extraction. The amount of water
used is from 1,000 to 2,400 gallons per ton, most being required
with clayey and heavily sulphuretted ores. In addition to this
from 1 to 2 gallons per minute are supplied to each concentrator
for feed water, and from ^ to 1 gallon per minute for wash water.
The amalgamating plates outside the battery are in three sets,
the apron plates, the tables, and the sluice plates ; the united
length of these varies from 10 to 30 feet, according to the re-
quirements of the ore. They are almost invariably made of
copper which is electro-plated with silver. The details con-
STAMP BATTERY PRACTICE. 193
cerning the preparation and care of these plates have already

been given, pp. 122-125. The coarse gold is chiefly saved
inside the battery ; the finely divided particles, forming a fluid
amalgam, are splashed through the screen and caught on the
outside plates. From 50 to 80 per cent, of the gold is caught
inside the battery, and the amount of free gold lost in the
tailings is not more than 5 to 8 per cent, in the best mills. The
tendency is in the direction of using more feed water with a
lower inclination to the plates, as by this means the ore is
brought into more intimate contact with them, being rolled
over and over instead of swimming down them, and the danger
of scouring is diminished. Other details concerning Californian
stamp mills have been already given in the general description
in Chapters vi. and vii.
Sizing of the tailings before concentration is not attempted, and
this point deserves more attention than it has hitherto received
at the hands of Californian millmen. Moreover, means should
be adopted to get rid of the free gold and floured amalgam con-
tained in the sands before passing them over the shaking tables,
which are ill adapted for catching these materials. The two
concentrators in most general use in the State are the Frue
vanner and the Triumph concentrator, both of which have been
already described. Next to these the Golden Gate and the
Duncan machines are favoured, while Hendy's pan is still to
be found doing good work, although it was one of the earliest
contrivances used in the Pacific States.
The concentrates are treated by roasting and chlorination,
generally by the vat process. Barrel chlorination has as yet
made little headway in California, owing in great part to the
fact that only small quantities of sulphides are produced at any
one mill, whilst the distances between the establishments are
very great. Chlorination costs about $10 per ton, and an
average of 92 per cent, of the gold is thus extracted from the
concentrates, which are usually worth fr6m $80 to $100 per ton.
After concentration, the tailings are in many cases run through
a length of 100 to 200 feet of blanket sluices, where an effort
is made to catch the free gold, amalgam, and sulphides still
remaining in the tailings. These sluices are cheap and require

little attention, but only account for a small percentage of the
total gold recovered. The blanket sands are ground with
mercury in a Chilian mill or arrastra, driven by water or steam
power.
An excellent addition to a mill is afforded by the automatic
tailings sampler erected at the original Empire Mill. This
machine, by an arrangement of cogs, dips a bucket into the
falling stream of tailings at stated intervals and takes a sample ;
all these samples are kept separate. They are usually taken at
intervals of an hour. Most mills sample their tailings on very
13
rare occasions. The average percentage of extraction does not

fall below 85 per cent, in any but the oldest and worst
appointed mills in the State, or in those running on unusually
refractory ore.
The usual cost of milling in a 40-stamp mill, treating 80 tons
per day, is as follows :
Cost per Ton.

I. Labour
(a) In the mill, 20| cents.
(6)Assaying, retorting, melting, and
superintendence, . . . .
24 ,,
II. Supplies-
Castings, 7 to 10 cents.
Mercury, 1 to 4
Lubricants, screens, illuminants, and
miscellaneous, . .. . 4 to 8 ,,
Total, . . 354 to 45
or, Is. 6d. to Is. 10d.
The cost of water power, if it is purchased, must be added to this, and, if
steam is used, about 11 cents per ton must be added for additional labour,
oil, &c., besides the cost of fuel. Interest on capital, and the cost of
chlorination are not included in this estimate.
The power required in such a mill is distributed as follows :
Rock-breakers, 12 H P.
40 stamps, 66
16 concentrators, 8
8 shaking tables, 2
1 clean-up pan, 14
1 clean-up amalgamating barrel and batea, ,
2
Total, . . . 92
The poorest ore which can be mined and milled at a profit

must contain from 2 to 5 dwts. of gold per ton.
In 1889 there were about 3,000 stamp heads existing in
California, of which not more than about 2,000 were in active
operation, crushing 4,000 tons of ore per day. Of these stamps
about 60 per cent, were run by water power, and the remainder
by steam. The cost of milling varied from 39 cents to $2 per
ton, the mean cost being about $1 per ton.* In 1892, about
4,000 head of stamps were in position in California, 3,500 of
which were at work. The lowest cost of milling was that at the
Dalmatia Mine, El Dorado County, where the total cost of
mining and milling was 43 cents per ton. In 1851, when quartz
mining was first begun at Grass Valley, no ore was considered
worth treating unless it yielded $40 per ton by the wasteful
process then in vogue, f In 1873, according to Deetken, the cost
had fallen to a minimum, of $2 per ton.
*
Ninth Report Gal. State Min., 1889, p. 25.
t Eleventh Report Gal. State Min., 1892, p. 19.
Method of Working employed in the Blackhawk Mills,

Gilpin Co., Colorado. The ore obtained in this district and
supplied to the mills is often called refractory, although the
greater part of the gold contained in it is extracted by amal-
gamation. It is refractory in the sense that it yields its
gold
less readily than the typical Californian free-milling ores. It
contains, in general, from 10 to 20 per cent, of metallic sul-
phides, and from 15 to 20 per cent, of quartz, the remainder
Fig. 44.
of the gangue being chiefly felspathic matter derived from

the country rock. The metallic sulphides are chiefly copper
and iron pyrites, but antimonial grey copper, arsenical pyrites,
and galena are also present, and blende and carbonate of iron
occur sometimes. The greater part of the gold is free, being
enclosed in the quartz, but about 25 per cent, of the gold and
33 per cent, of the silver are locked up in the sulphides, which
are saved by concentration, and shipped to the neighbouring
smelters in Denver. The gold is in a very finely divided state,

the ore often giving no "colours" when subjected to ordinary
panning, although in the mill it will yield a good return. The
average milling ore contains about J oz. of gold and 2 or 3 ozs.
of silver per ton.
The method used consists briefly in slow crushing and
amalgamation in the battery, where as much gold is saved as
possible, followed by the passage of the pulp successively over
a set of amalgamated plates, then over blankets, and finally over
concentrating machines, the tailings being allowed to run into
the stream. A sectional view of the battery used is shown in
Fig. 44. It will be observed that only a shallow light iron
mortar, a, is used, the sides consisting of a housing of wood.
The water is introduced by the pipe, b, and allowed to run down
the stem of the stamp. The ore after passing the screens flows
over the amalgamated plates fixed on the tables, c. The stamps
are light, weighing from 500 to 600 Ibs. each, and fall at the
extremely slow rate of from 26 to 32 drops per minute, the
height of the drop being from 16 to 18 inches. This is greater
than that adopted in any other part of the world, being four
times as great as that adopted in the latest Californian practice.
The depth of discharge is 13 inches when new dies have just
been put in, and increases to 15 or 16 inches as they wear down.
According to the most modern view, however, this depth should
be diminished, and at the mill latest erected it is only 11 inches.
The shoes and dies are small (8 inches in diameter) owing to the
lightness of the stamps, and the screens are fine, a burr slot
being used which corresponds to a 50-mesh wire screen. The
screen surface is normal, the dimensions being 4^ feet by 8 inches
for each battery of five stamps. Under these conditions the
output is of course low, being on an average about 1 ton per
stamp head in twenty-four hours. The feeding is done by hand,
one man feeding twenty-five stamps, and there are no rock-
breakers, their absence being accounted for by the age of the
mills and the lack of fall in the ground in the rear of the
batteries, which interferes with their introduction at the present
day. Nevertheless, rock-breakers and automatic feeding would
reduce the cost of crushing considerably, as the cost of hand-
feeding alone is $450 per battery in a year, while the first cost
of automatic feeding machines would only be from $500 to
$1,000, and the cost for maintenance, power, and repairs nominal.
The amount of water added to the battery is from 1J to 2
gallons per stamp per minute. The amount of mercury used
varies of course with the richness of the ore. When ore con-
taining J oz. of gold per ton is being crushed, about 2J ozs. of
mercury are added to each battery per day in small quantities
at intervals of an hour, while about an equal amount (2J ozs.) is
required to dress the plates, both inside and outside. Of this
quantity one- third is put on the back inside plate, one -fourth on
the front inside plate, and the remainder on the outside plates.
The loss of mercury from all sources is about 20 or 25 per cent,
of that added each day, and, with ^ oz. ore, is about -1 oz. of
mercury per ton crushed.
Two sets of amalgamated plates are used inside the battery,
each 4^ feet long, the front one being 6 inches wide, and the
back one 12 inches. The front plate is nearly vertical, but the
back one is set at an angle of 40. There is only a single set of
plates outside, which are 4 feet wide by 12 feet long, and have
the high grade of 2J inches per foot. The plates are dressed
every twelve hours with a weak solution of cyanide of potassium
(2 ozs. in 3 gallons of water), this being the only chemical used
on the mill. The consumption of cyanide is about 2 ozs. per
battery of five stamps in three days. Its use is necessitated by
the acid nature of the water, which already contains sulphates
and carbonates when it comes from the mines, and is further
contaminated by ferrous sulphate, formed by the oxidation of the
pyrites in the ore and then dissolved. This water soon forms a
film on the surface of the copper plates, and also rapidly corrodes
the screens, which otherwise, from their unusual height above
the dies, would last a long time. Nevertheless no effort is made
to counteract the bad influence of this water by adding alkalies,
limestone, or lime water to it or to the ore.
After leaving the plates the ore passes over the "blanket-
strips," which are 3 feet long and 18 inches wide. They are
washed every four hours, or every two hours if the ore is rich.
The blankets serve to catch any escaping mercury, amalgam,
"rusty" gold, and the heaviest pyrites, together with pieces of
rock to which gold is attached. Probably this work would be
done equally well by the concentrators, by which an operation
would be saved. The blanket-sands are sold to the smelters with
the other concentrates. The concentrators consist of shaking
tables, constructed in the locality ; they have been already
described at p, 175. The concentrates usually contain about
15 dwts. of gold and 5 ozs. of silver per ton.
The amalgam obtained is divided fairly equally between the
front inside plate, the back inside plate, and the outside plates.
In retorting it the interior of the iron retort is chalked or lined
with paper to prevent the gold sticking to it, the latter method
being considered the better. The balls of dry squeezed amalgam
are put into the retort, broken with an iron rod and then
pressed down until hard and uniform. A large bolt with
a nut at the end is often used for the purpose. The cover is
then put on and luted down with clay. The variety of retort
used is similar to the bell-shaped one mentioned on p. 133. One
hundred parts of amalgam yield from 33 to 50 parts of retorted
meta], the average being about 40 ; the amount depends on the
richness of the ore and the coarseness of the gold. The bullion
contains from 750 to 850 parts of fine gold, and almost all the
remainder is silver, about 10 parts per 1,000 being base metal.
According to T. A. Rickard,* the late manager of the New
California Mine and Mill, from whose descriptions many of the
details already given have been taken, the results obtained by
his mill on an ore containing 10 dwts. of gold and 2J ozs. of
silver per ton, were as follows
:
The use of the slow drop is intended to enable the gold and
pyrites to settle a little by gravity through the lighter particles
of gangue between each blow, and so to keep them longer in the
battery, and thus increase the chances of amalgamating the gold.
The wide roomy mortar prevents a violent splash against the
screens and plates, so that little scouring of the latter takes
place. There can be no doubt that, on the whole, the method of
working in Gilpin County, which was slowly and painfully
elaborated about twenty years ago, is the best that could be
applied to the particular ores which occur in the district.
Battery Practice on Free- Milling Ores in Australia and
"New Zealand. Australian practice as a whole, if some of the
new mills in Queensland are excepted, owes little to the experi-
ence gained in the United States, and pursues widely different
methods from those in use there. Modern methods in both
countries are based almost entirely on the processes in use in
Central Europe in the first half of this century, and have been
modified to some extent in different directions in various
localities, the changes being in most cases beneficial, and
especially suited to the class of ore in course of treatment. The
process employed on the free-milling coarse-gold ores in Yictoria,
New South Wales, and New Zealand has perhaps undergone less
variation from the original Transylvanian type than that
employed in any other part of the world, chiefly owing to the
extreme simplicity of these ores, and the ease with which a high
percentage of the gold can be extracted.
The ores consist of white quartz, containing large grains and
plates of remarkably pure gold, which is thus often visible,
and sometimes occurs in big pieces weighing several pennyweights.
Inclusions of country rock (slate, <fec.) are common, and in these
cases the sulphides often slightly increase in quantity, but rarely
form more than from ^ to f per cent, of the rock, and consist
chiefly of iron or arsenical iron pyrites. The free gold is not
intimately associated with these pyrites, and the ore may even
fall off in value when the
quantity of sulphides increases. The
poorest ores treated contain 4 or 5 dwts. of free gold per ton, an
amount which includes from 85 to over 99 per cent, of the total
gold contents, and the concentrates contain from 10 dwts. to
4 ozs. or more of gold per ton. These concentrates are saved in
some districts and treated by roasting and chlorination, while in
other places they are allowed to run to waste, especially if they
contain less than 1 oz. of gold.
The method of treatment consists in crushing the ore somewhat
coarsely in rectangular mortar boxes, with stamps of medium
weight, running at a medium rate of speed, and then in passing
the pulp through mercury wells and over blanket strakes. The
mortar sand, which is collected periodically, is amalgamated in
revolving barrels, together with the blanket concentrates, but no
mercury is added to the mortars, and amalgamated plates are
conspicuous by their absence. The following details concerning

the machinery, <fcc., employed at certain mills in Victoria and
New Zealand are chiefly derived from T. A. Packard's
writings* on the subject, to which the reader is referred for a
more complete account.
Rock-breakers and automatic feeding machines are seldom
used, and the hand-feeding is often done very carelessly. The
mortar boxes are of rectangular shape, have no amalgamated
plates inside them, and are considered by American experts to be,
in many cases, unsuited for internal amalgamation, as the mer-
cury and amalgam do not collect out of reach of the falling
stamps, which would, therefore, cause flouring. It is in this
way that the fact may be explained that a smaller saving of
gold is effected when mercury is fed into the mortar boxes, than
when the customary practice is adhered to, since the mercury and
amalgam are subjected in the former case to conditions under
which excessive flouring takes place, and the mercury is lost in
the tailings, together with such gold as it may have taken up.
On the other hand, Australian millmen point out that the
mortar boxes are spacious enough for the retention of the coarse
gold which collects by gravity in them, without being pounded
to fragments by the stamps, and is removed at intervals by a
scoop, and that the collection of this gold is retarded and not aided
by the addition of mercury. The mortar boxes are from 14 to
16 inches wide, and spaces are usually left between the dies,
and between these and the ends and sides of the boxes, which
would certainly be far too roomy if the battery were intended
only for a crushing machine and not as an apparatus for collecting
coarse gold also. Australian managers prefer blankets to copper
plates, their reason for this preference, as well as for the reten-
tion of their mortar boxes, being based, no doubt, on the coarse-
ness of the gold to be saved.
The weight of the stamps is usually about 800 or 900 Ibs.,
8 cwts. being a favourite weight ; the diameter of the shoe is
10 inches, the height of the drop is 7 J or 8 inches, and the speed
about eighty blows per minute. The depth of discharge is low,
being usually about 4 inches, and this is kept constant by packing
up the die with sand or by putting in thin iron plates below it,
as it is worn down. The screens are round-punched Russia iron
or sheet copper, the latter lasting much longer they are pierced
;
with from 81 to 180 holes per square inch. The screens are of
greater area than those used in America, being about 12 inches
in vertical height instead of 8 inches. Double discharge is often
used and would always be advantageous, the duty being from
2J to 3J tons per stamp per day at the mills where double dis-
*
"Variations in Gold Milling," Eng. and Mng. Journ., Jan., Feb., and
March, 1893.
charge is used, and more than half this amount in the

little
single discharge batteries. The amount of water used is very

large e.g., 8 gallons per minute per stamp head at Clunes,
where there is a double discharge and where a low inclination
isgiven to the blankets. This quantity of water is four or five
times as much as that supplied in California.
A
double discharge battery in use at the South Clunes United
Mill is shown in section in Fig. 45 ;
A
and B are the screens,
and the launder, C, conveys the discharge from the back to mix
with that coming from the
front screen. The perfor-
ated iron plate, D, serves
to distribute the ore evenly
over the apron, E, over
which the pulp Hows ;
the
latter subsequently passes
through the mercury wells,
F and G, being compelled
to pass through the bath of
quicksilver by the vertical
boards, H and K. The
wells are made of iron,
which keeps the mercury
lively, and are 3 inches in
width and 4 inches deep ;
the ore has a fall of 10

inches upon the surface of
the quicksilver. Each well
contains about 50 Ibs. of
mercury, and is usually
cleaned up once a week by
passing its contents through
canvas and so separating the
amalgam the wells are also
;
skimmed at short intervals.

They are omitted altogether Fig. 45.
in some mills, gold being
saved only by gravity in the mortar and by concentration on
blanket strakes.
The blanket tables, L, should have a united width equal to-
that of the screens they are usually divided into widths of 18
;
inches, and consist of three or four lengths of 6 feet each,

over
which the ore passes in succession the grade at which they are
;
baize is the material

placed is from 1 to 1 inches per foot. Green
most commonly used, the colour enabling specks of gold to be
readily seen. The blankets in the top row are washed in tubs-
every half hour or hour, according to the richness of the ore, and
those in the lower rows at longer intervals. The blankets are
sometimes succeeded by concentrators, which usually consist of

some modification of the Cornish huddle. At some mills the
experiment has been made, with considerable success, of replacing
the lowest row of blankets by amalgamated plates. The blankets
are well adapted to catch coarse gold, but the fine gold escapes
them, and is caught on the plates, which are efficient for this
purpose, though perhaps of no greater utility than blankets
in arresting coarse particles.
In cleaning-up, the screens are taken out, the contents of
the mortars are passed through a No. 4 sieve, and the material
that remains on this is returned to the battery and used to pack
round the dies. The fine material one or two buckets full from
each mortar box at each fortnightly clean-up contains a quantity
of free gold, and is treated with mercury in the amalgamating
barrel, together with the blanket concentrates. The skimmings
from the wells, and occasionally the blanket concentrates also,
are treated in Berdan pans, and the concentrates from the
buddies are either roasted and amalgamated in Chilian mills,
or, more frequently, roasted and chlorinated in revolving barrels.
The amalgamating barrel, mentioned above, is hung on trun-
nions, and revolved afc the rate of about 16 revolutions per
minute. From 5 to 20 cwts. of concentrates, battery sand, &c.,
according to the size of the barrel, are charged in together with
75 Ibs. of mercury and some cold water, the chemicals used
being sulphate of copper and wood ashes. After running for
about ten hours, the barrel is discharged into a tank and its
contents allowed to settle, the process being sometimes aided by
passing the pulp through a series of wells, as, for example, at the
South Clunes United Mill where there are three drops, the
heights of which are 12, 9, and 6 inches respectively, the
material falling each time into a bath of mercury. The battery
sand is usually kept separate from the blanket concentrates, as
it is much richer than the latter. The tailings from the barrel
and the Berdan pans are sometimes run over a few feet of copper
plates to catch any finely divided amalgam that may be con-
tained in them, and are sometimes concentrated on Rittinger
shaking tables.
The percentage of gold recovered varies greatly according to
the nature of the ore and to the method of treatment adopted.
At many of the mills no steps are taken to ascertain the extent
of the loss, and, in the best mills, only from 80 to 85 per cent, of
the gold is saved. More than half of the gold which is recovered
comes from the mortar boxes, where it is merely collected by
gravity, and the remainder is derived mainly from treating the
blanket concentrates. At Clunes, where mercury wells are used,
the relative amounts of gold obtained from the various appliances
are approximately as follows : From the mortar box, 45 per
-cent.; from the wells, 30 per cent.; from the blankets, 15 per cent.;
and from the skimmings of wells, 10 per cent. At Otago, where

mercury wells are not used, the distribution is as follows :
Mortar box, 60 per cent.; blankets, 33 per cent.; copper plates,

7 per cent.
The loss of mercury would be very small, if it were not for
the flouring that takes place in the treatment of the skimmings,
<fec., in Berdan pans. Loss by flouring also takes place if the
amalgamating barrel is run too rapidly. Owing chiefly to this
latter cause, the loss per ton of ore crushed is 8 dwts. of mer-
cury at Otago, while at Chines it is about 3 dwts. At the
South Clunes United Mill the loss for the eight years (1884-92)
averaged only 5J grains of mercury per ton, this being probably
the smallest loss on record in a stamp mill. This fact proves the
often-repeated assertion that a bath of mercury can be mani-
pulated with very little loss. If metallic copper be present in
the ore, as is sometimes the case, the copper amalgam formed
floats on the surface of the mercury, and is readily carried away
with the tailings, the loss
being thus increased to 2 or 3 ozs. of
mercury per ton of rock crushed. The cost of treatment is, in
some well appointed mills, only from 2s. to 2s. 6d. per ton.
Method of Working in the Thames Valley, N.Z.* The
ore in this district may be regarded as on the border line
between "refractory" and "free-milling" ores. It varies greatly
in hardness and composition, and consists of "stringers" of
quartz running through a more or less decomposed and brecci-
ated hornblende-andesite. The gold is often in visible specks
and threads in the quartz, but is also largely associated with
copper and iron pyrites, blende, galena, stibnite, &c., the quan-
tity of sulphides varying from j to 10 per cent, of the ore, and
averaging from 2 to 3 per cent. Silver occurs as argentite, fcc.,
while tellurides of both gold and silver are sometimes found.
The average amounts extracted from the ordinary ore are gold,
5 to 10 dwts. per ton, silver, 2 to 5 dwts.; but a considerable
percentage of the gold recovered comes from "specimen" ore,
which is often worth some thousands of pounds per ton.
No rock-breakers are used, grizzlies or sizing screens being also
unknown, and the feeding is done by hand. The small stuff is
shovelled into the mortar, and big pieces of rocks are rolled into
the feed opening, and rammed in by a few blows of the sledge
hammer, if they stick there. The results of this primitive prac-
tice are seen in the excessive rate of wear of the shoes and dies
i.e., 22 ozs. of iron per ton of ore crushed, 14'5 ozs. for the shoe,
and 7 -5 ozs. for the die. The feeding is very high, the mortar
boxes being choked up with ore, so that the output is still further
reduced from this cause. The mortar box is spacious, approaching
in design the Gilpin County pattern, but, as no attempt is made
*
The following account is partly based on the description given by T. A.
Rickard in the Engineering and Mining Journal, Dec. 3 and 10, 1892.
204: THE METALLURGY OF GOLD.
to amalgamate the ore before it leaves the battery, a narrower

shape might be used with advantage. The stamps weigh from
6*40 to 840 Ibs., the faces of the shoes and dies being 9 J inches in
diameter, while the dies are 4 inches deep. Both are made
of white cast iron, only the shoes being chilled. The height of
the drop is 8 or 9 inches, and the number of drops per minute
is 75. The depth of discharge, though varying as the dies wear,
is on an average only from 2 to 3 inches, the bottom of the screen
being placed on a level with the surface of the pulp in the mortar,
or even below it, the bottom being on a level with the top of the
dies in one mill. The screens are placed in a vertical position,
and are made of round-punched Russia sheet iron, having from
148 to 180 holes per square inch.
The wearing of the screens is very rapid, owing partly to the
small depth of discharge, the ore being flung almost horizontally
from the surface of the dies and striking the screens with great
force, but still more owing to the acidity of the battery water.
A grating usually lasts about six days, or while about 50 tons of
ore are crushed. Copper gratings might be used with advantage.
The mine waters contain an unusual quantity of the proto-
sulphates of iron, copper, manganese, and aluminium, and,
consequently, when the millstufF is very wet, the screens are
corroded with great rapidity, while the soft surface ore, in which
the sulphides have been largely decomposed, wears out the
screens faster than the hard deep-level ore. It is unfortunate
that lime water and limestone are not readily obtainable in the
district, as an addition of either of these would undoubtedly
lengthen the life of the screens. The output is from 1 to 1J
tons per stamp per day, which is low, considering that no battery
amalgamation is attempted. A fast-running heavy stamp, with
double discharge, would probably double the capacity of the
mills, and leave the gold in a better condition for amalgamation.
The pulp, after being ejected from the mortar, is run over
amalgamated tables, 7 feet long and 4J feet wide. On the
upper part of these tables, next the battery, is a plate 2J feet
long, succeeded by a well 2^ inches wide arid 2 inches deep, filled
with mercury. Below this is another length of 18 inches of
amalgamated plate, and then a succession of three more "ripples"
or wells, not filled with mercury. The total length of the plates
is thus only 4 feet, and the only other means adopted to catch
the gold consist in the use of the ripples and some blanket
strakes, by which some of the free gold and pyrites are caught.
After passing over the blankets, the tailings are allowed to run
into the sea. The "blind ripples" i.e., those not containing
mercury are cleaned with a scoop every half hour, the heavy
sand and pyrites so obtained going to the pans together with the
material caught on the blankets, which are washed every hour
or even oftener. The blankets cost 6 shillings per square
yard, and last for three months. The plates consist sometimes
of copper, but more usually of Muntz metal ; they are cleaned,
wiped, and dressed every four hours, the wells being skimmed
with a cloth at the same time. The mercury in the wells is
squeezed through wash leather once a week.
The use of the mercury wells serves but little purpose as, in
accordance with general experience elsewhere, the sulphides
cause a scum to form over the surface of the bath after a few
minutes, and no further amalgamation takes place until it is
skimmed off.
The pans used for the treatment of the blanket concentrates
are chiefly Berdans, with a few Watson and Denny's. The
Berdan pans are furnished with a drag, fixed to one side
of the sloping bottom, instead of a ball, as is usual. This is a
useful modification, as grinding and amalgamation are kept
separate, while in the ordinary Berdan pan, the ball remains at
the bottom of the cone where the mercury collects, and is the
cause of a considerable loss by flouring. The drag consists of two
parts, the lower part being the slipper or shoe, weighing 196 Ibs.,
to which the boss or top, weighing 230 Ibs., is rigidly keyed. The
shoe wears out in four months. The Berdan pans are 4^ feet in
diameter, with a depth of 9 inches, the bottom having an inclina-
tion towards the centre of 1 in 2f ; they work at 23 revolutions
per minute, and the amalgam is removed and strained every
twenty-four hours.
The greater part of the amalgam is obtained from the plates,
the percentages being from the plates about 75 per cent., from
the wells 5 to 10 per cent., from the pans about 15 per cent.
One of the features of the district is the " specimen " stamp
attached to each mill, by which the valuable specimen ore is
crushed, the gold being caught chiefly in the mortar box, while
all the tailings are saved and treated in the pans.
The amalgam yields about 40 to 50 per cent, of retorted metal,
which on melting yields bullion about 600 to 675 fine in gold,
the remainder being almost entirely silver. The loss of mercury
is high, owing to excessive flouring in the pans ; it varies from
15 dwts. to 1 oz. per ton of ore crushed. The cost of working
is about four shillings per ton, water power being used and paid
for.
The absence of concentrators to save the sulphides renders
the process somewhat wasteful. The sulphides from these tail-
ings often contain from 15 to 25 dwts. of gold per ton, and a
much greater amount of silver. Even the free gold is stated to
be by no means completely extracted, nearly 50 per cent, passing
away in the tailings. There is certainly an accumulation of
tailings on the foreshore estimated to amount to over 1,000,000
tons, which are believed to average about J oz. of bullion per
ton. A few years ago a large number of samples were collected
from various parts of the foreshore and sent to England, where

they were assayed by the author, and were found to contain on
an average about 2 dwts. of gold and 6 dwts. of silver per ton.
A large proportion of the sulphides was in the state of slimes
and difficult to save by concentration, but, on
reducing the
samples in the ratio of 10 to 1 by panning, the residue was
found to contain J oz. of gold per ton, while if reduced in the
ratio of 28 to 1 the residue of pure sulphides contained 1 oz. 6
dwts. of gold and 14 dwts. of silver per ton. These results
seem to point to the fact that the extraction of the free gold is
not so incomplete as is usually believed, although the total loss
is considerable, owing to the value of the sulphides.
Battery Practice in Dakota. The method adopted in the

Black Hills consists in battery and outside plate amalgamation,
followed by treatment by a few mercury traps and riffles, and
by blanket sluices, no systematic concentration being attempted.
The reasons for this are that the ore is free milling and the
greater part of the gold is sufficiently coarse to be caught inside
the battery, whilst the sulphides from most of the lodes are
not rich enough to pay for treatment, and so are allowed to
run to waste, although they could be saved at a trifling cost.
The sulphides from the Homestake Mine contain gold and silver
to the value of only $5 or $6 per ton, whilst the poorest ores
which can be treated at a profit, either by the pyritic smelters or
the barrel chlorination works in the neighbourhood, yield from
$8 to $12 in gold per ton. At the Caledonia Mill, however,
the ore treated contains about 4 per cent, of pyrites, which assay
about $90 to the ton, and concentration is absolutely necessary for
successful working. There are 740 stamps running continuously
in the Black Hills, most of which are under the management of
the Homestake Company ; 3,000 tons of ore are crushed per day.
The yield in bullion is from 3 to 4 dwts. of gold and nearly 1
dwt. of silver per ton. When surface ore was being treated the
tailings often contained less than ^ dwt. of gold, and even as
late as the year 1889, from 70 to 80 per cent, of the total contents
were extracted, but the increase in the amount of pyrites in the
ore, due to the greater depth of the workings, has of late years
resulted in an increased loss in the tailings and a diminished
yield on the plates, so that the extraction probably does nob
average more than from 50 to 70 per cent., and the tailings
often contain as much as 2 to 2 J dwts. of gold per ton.
The Homestake Mill is arranged in such a way that the ore
need not be touched by hand after it is let fall into the gigantic
ore-bins; consequently eight men can do all the work in con-
nection with the battery of 200 head of stamps. The ore
is passed over grizzlies, thence through Blake's or Gates
crushers, each Blake (of which the dimensions of the mouth
are 9 x 15 inches) being set to supply 20 stamps with ore
passing through a 1^ inch ring. From the crushers the ore

passes through Challenge ore feeders into the batteries, which
are arranged in two rows set back to back with the ore
feeders between them, the distance between the faces being 44
feet. The stamps weigh 850 Ibs., their fall is 8 to 9 inches, and
they make 90 drops per minute. The screen used is No. 7, and
the output is about 4 to 4J tons per stamp per day. There is
one copper plate inside the mortar below the screens, and about
1 oz. of mercury per day per stamp head is added with the ore.
The shoes and dies are of cast iron and last about two months.
It has been shown by experiments in the Father de Smet Mill*
that for crushing low grade ores a square-bottomed die is the
best, and that there is no economy in using the dies after their
surfaces have become perceptibly irregular by wearing. It is
accordingly desirable to cast them as thin as practicable and to-
replace them often at least once a month.
The amalgamated plates placed outside the mortar are about
10 feet long and 4J feet wide ; they have an inclination of 2
inches to the foot, the grade being kept as high as is practicable
consistent with good amalgamation, owing to the scarcity and
high cost of the water. At the lower end of these plates the
ore is passed through a mercury trap or well, and the discharge
is then narrowed and the pulp is passed through sluices from
60 to 100 feet long and 2 feet wide, lined with copper plates,
and broken at intervals by mercury traps. Hungarian riffles
are also used, and at the Caledonia Mill, where the sulphides
are valuable, an imperfect concentration is effected by means
of blankets. The battery plates are dressed every day, and any
excess of amalgam then found is removed; the sluice plates
are dressed every four days. The battery is cleaned up fort-
nightly, and the sluices, mercury wells, riffles, &c., monthly.
The cost of milling is 83 cents (3s. 8d.) per ton, the water alone
costing from 50 to 57 cents per stamp per day, or 12 to 13 cents
per ton of ore. The pulp is run through settling pits and the
water used over again during the winter months, when the
scarcity of water is felt most severely. It is then warmed before
use. The fineness of the bullion is, in general, 820 in gold and
170 in silver.
Stamp Battery Practice in the Transvaal, t On the
Witwatersrand the ores consist almost entirely of the now
familiar "banket" formation, a conglomerate consisting of
water-worn pebbles of translucent quartz set in a matrix
composed mainly of oxides of iron and silica. In the ore near
the surface the gold is almost entirely free, existing in the
cement, the pebbles being generally barren. At a moderate
*
Trans. Am. Inst. Min. Eng., vol. x., p. 97.
t This account is mainly based on information kindly supplied by Mr.
S. H. Farrar, M.I.C.E., F.G.S., M.I.M.E., of Johannesburg.
depth, however, the oxides of iron gradually give place to

sulphides, but even then the greater part of the gold can be
extracted by amalgamation. The ores vary greatly in richness,
some containing less than 1 dwt. of gold per ton, and others as
much as 2 or 3 ozs., the average being about 15 dwts.
The method of treatment consists in crushing through some-
what coarse screens with a small depth of discharge, thus
obtaining a very large output; the gold is caught partly on
plates inside the mortar boxes, partly on copper tables placed
outside. The pulp is concentrated in various ways, blanket
strakes, Frue vanners, and spitzluten being in common use.
The Luhrig vanner is being introduced, and has already met
with a considerable amount of success. The concentrates, if
they contain little or no sulphides, are ground in pans with
mercury ; if pyritic, they are either roasted and chlorinated or
treated with cyanide. The tailings from the batteries or the
concentrators are treated by the cyanide process.
There are two types of stamp mills in existence on the Rand,
both of which have been very successful, although differing
somewhat in their construction. These are the American mills,
constructed chiefly by Messrs. Eraser and Chalmers, and the
English mills sent out by the Sandycroft manufactory, and by
other well-known English firms. A brief account of some of the
details of these mills is given below.
Rock-breakers are in use at almost all the mills on the Rand,
"both the Blake-Marsden and the Gates crushers being in wide
use. In general, one Blake-Marsden machine with a feed-
opening of 15 by 9 inches is required for every twenty stamps.
The rock-breakers are generally contained in separate rock
houses, placed at the mine. The advantages are that the rock-
breakers can be placed nearer the ground, and are less expensive
to erect, while the stuff in the battery ore-bins is more completely
mixed. Self-feeders to the batteries are universal, the Challenge,
Tulloch, and Sandycroft machines being chiefly used they are
;
arranged so as to keep the ore at a depth of 1 to 2 inches on the dies.

The mortar blocks are usually of 14-inch square timber, from
9 to 15 feet long, and are laid on concrete. It has been found
that the usual practice of filling the space between the blocks
under one mortar, and those under the next, with well-tamped
earth is less satisfactory in giving stability than that of filling
the space with solid masonry. The latter course has accordingly
been adopted at the best mills. The mortar blocks are covered
with blanketing or sheets of india-rubber or lead before the
mortars are bolted to them. The English mortars are cast about
4 feet high and 4 feet 8 inches long ; the width inside at the
level of the dies is lOf inches, that of the American mortar being
11 inches. Both mortars widen considerably towards the upper
part, the walls not being vertical at any point. The thickness
of the mortar at the bottom is 6 to 9 inches, and the inside is

usually lined with wrought-iron plates to protect it from wear.
The American mortars are furnished with amalgamated copper
plates inside, both on the feed and discharge sides, but the
earlier English mortars on the discharge side only. The inclina-
tion of these plates to the horizon is usually from 70 to 80.
Steel wire mesh screens are almost invariably used, punched
Russia iron plates being rare. There are from 100 to 1,200
holes per square inch of the screen, the usual number being 600
to 900 (i.e., 25 to 30 holes to the linear inch). The depth of
discharge is about 3 to 6 inches, and the scouring action on the
front inside plates is consequently so violent that but little
amalgam accumulates there, and, in the opinion of many
experts, these plates might advantageously be dispensed with.
Fig. 45a.
A section of the mortar used at the Croesus mine is shown in
Fig. 45a. a is a cast-steel lining plate with slots or recesses in
it for
collecting the amalgam; b is the feed opening; c the
wooden blocks for carrying e, the copper plate ; d is the screen
opening ; and f a steel plate.
The weight of the stamp in the earlier mills was usually
about 850 Ibs. in the American, and from 700 to 750 Ibs. in
the English mills. In the later mills, which represent in every
respect the very best modern practice, the stamps are of 1,000
Ibs. weight or upwards, while in the Modderfontein
40-stamp
battery the weight is 1,250 Ibs., the heaviest gravitation stamp
yet in action. The actual weights of the parts of two English
stamps in use were found to be as follows :
H
Nominal weight of Stamp, . . 700 Ibs. 750 Ibs.
Actual ,, Stem, . . 238 287 ,,

Head, . . 197 197
Tappet, . . 115 115
Shoe, ,. . 180 180
Total actual weight of Stamp, . 730 779 ,,
The usual height of drop varies from 6 to 9 inches, and the

speed is usually from 90 to 95 drops per minute. The shoes
and dies are 9 inches in diameter, and the shank of the shoe in
English mills is 5 inches long and 3 inches wide. The shoe is
often kept in work until the butt is worn to about 1 inch thick.
The dies are now sometimes made in two pieces, a new boss
being fitted into the footplate as soon as the old one has been
worn out. Steel is almost exclusively used for shoes and dies,
manganese steel being preferred. The order of drop is usually
1.3.5.2.4 in the English mills, but by some authorities the orders
1.4.2.5.3 and 1.5.3.2.4 are considered better.
The cam-shaft is from 4| to 6 inches in diameter, according
to the weight of the stamps. A
separate cam-shaft is used for
each battery of five stamps. The cam-pulleys in the American
mills resemble that described on p. 108 ; in the earlier English
mills they were often of wrought iron, 6 or 7 feet in diameter,
with round arms. Many of these have proved defective, be-
coming crystalline and breaking after running for a short time
only. Wooden pulleys are now generally used.
The water for ten stamps is supplied through a 3-inch gas
pipe, from which two IJ-inch pipes branch to each mortar. The
pipes are sometimes suspended from the roof instead of being
attached to the battery posts ; the objection to the latter course
is that the pipes work loose owing to vibration. The amount
of water supplied varies in the De Kaap gold field from 100 to
200 gallons per stamp per hour, or from 1,600 to 3,200 gallons
per ton of ore crushed. On the Witwatersrand the amount of
water per ton of ore varies from 1,600 to 3,500 gallons, and,
according to Hatch and Chalmers, averages about 2,000 gallons
per ton crushed. Mercury is added to the mortar boxes in small
quantities every few minutes, the total addition amounting to
a spoonful in two hours.
The arrangement of the amalgamated copper plates outside
the mortars differs somewhat in different mills. In some, a
single plate of best Lake Superior copper is used, 4^ to 5 feet
wide, and from 8 to 10 feet long, with a mercury well at the
lower end. In others, the lip of the mortar box is covered with
a sloping copper plate from 12 to 14 inches wide, running the
whole length of the mortar. The pulp flows over this "lip-
plate" on to a second copper stiip, the "splash-plate," 14 inches
wide, and as long as the lip-plate, but inclined in the reverse
direction so as to throw the pulp back towards the mortar. It

then falls into an iron trough or distributor of semi-circular
cross-section, perforated with small holes for the purpose of
distributing the pulp over the plates. The latter are about 4}
feet wide and | inch thick ; the pulp flows over 4 or 5 plates in
succession, each 2J feet long. They consist of copper, which is
seldom coated with electro-deposited silver. 'There is a shallow
riffle at the lower end of the first plate, and a drop of 1 inch
between each pair of the rest ; sometimes the plates simply
overlap, so that the drops are of only J inch. At the lower end
of the series of plates is placed a mercury trough. The tables
for the support of the plates are made of heavy timber so as to
be steady these timbers are not fastened to any of the battery
;
posts or sills, as the vibration caused by the stamps would inter-

fere with amalgamation. The grade of the plates varies from
1 to 2 inches per foot, and can be altered by means of wooden
wedges or of adjusting screws.

The plates at the Sheba mill are treated as follows, according
to an account furnished by Mr. H. ISichols, A.R.S.M. : The
outside plates are usually dressed every hour if the rock is of
ordinary richness (i.e., containing from 10 to 15 dwts. of gold
per ton). In dressing, the amalgamator brushes all the quick-
silver and thin amalgam up to the top of the plate and distributes
it there. If the rock crushed has been rich (containing say
1 oz. per ton), quicksilver must be sprinkled on the plates and
worked into the hard amalgam with a brush, but if it has been
only 10-dwt. ore, the plates only require fresh mercury once in
two or three hours. The outside plates may require scraping
once every day or every other day if the ore is'rich the inside
;
plates are scraped only on the occasion of the monthly clean-up.

The yield from each of the two sets of plates is about the same.
A retort of 2,500 ozs. of squeezed amalgam yields about 750 ozs.
of gold. Cyanide of potassium is added in the mortar boxes in
small quantities when the plates are discoloured by salts of
copper ; and soda is used to remove grease. When a screen is
badly broken, the five stamps connected with it are at once hung
up, as much loss may be incurred by allowing coarsely-ground
rock to pass over the plates.
On the Rand, the amount of gold caught on the plates is
the ore, and
usually between 60 and 70 per cent, of the gold in
the average is about 65 per cent. A higher percentage can be
easily caught, but only if the output is decreased,
and this
sacrifice is less worth making, for the reason that most of the
is extracted
gold which is allowed to escape from the tables
from pyritic
subsequently by cyanide. The percentage extracted
ore is but little less than that from the oxidised banket, the
more refractory minerals being in general absent from the ore.
The amount of gold caught inside the battery is from 10 to
40 per cent, of the total saved by amalgamation if inside plates

are used, and 4 or 5 per cent, if there are none, and the greater
part of the remainder is retained on the first two feet of tables.
This part is scraped every day, but the lower parts are left
undisturbed for longer periods, varying up to a week. At the
Hobinson mine the plates are dressed every four hours, but
amalgam is removed only every other morning (Hatch and
Chalmers).
The method of concentration at different mills varies; the
following details refer to the Simmer <fe Jack mill of 100 stamps,
when the ore was still free-milling, and a rough method of con-
centration sufficed. Here blanket tables, 16 feet long and of the
same width as the copper plates, are placed immediately below
the latter. The blanket tables are divided longitudinally into
three parts by wooden " fillets." The blankets are washed
every two hours and the concentrates collected in wooden boxes
running on wheels. One division of each table is washed while
the pulp is deflected so as to run over the other two divisions.
The blanket-sands amount to about 6 per cent, of the tailings ;
they are treated by grinding with mercury in four pans, which

together treat 24 tons per day of twenty-four hours. The
charge for each pan is 2,600 Ibs., five charges being treated in
twenty-four hours. The pans are of cast iron, 5 feet in diameter,
and have double bottoms into which steam can be passed for
heating purposes ; the mullers make from 50 to 60 revolutions
per minute. The discharge takes place through a pipe at the
bottom of the side of the pan, and the pulp is run into settlers,
of which there are two, each 7 feet in diameter. The pulp is
here diluted with water and stirred by four arms revolving at
the rate of fifteen turns per minute. There are a number of
holes at different levels in the side of the settler and the diluted
pulp is run off through these as soon as the mercury has settled
sufficiently.
Where the ore is pyritic, this method does not suffice, and
vanners or hydraulic classifiers (Spitzluten, see p. 170) are
employed. It is usual to set three Frue vanners to treat the
pulp coming from five stamps. There has been much discussion
on the respective merits of these two classes of machines in
treating battery pulp on the Hand. Where vanners are used,
the concentrates have usually been roasted and chlorinated, and
the tailings treated by cyanide. The growing tendency to treat
even concentrates by cyanide, however, throws doubt on the
advantage of using vanners at all, for if both classes of materials
into which the pulp is separated are ultimately treated by the
same process, there is no prima facie reason for effecting the
separation at all, and, on the other hand, spitzluten afford a
product in an excellent condition for cyaniding. The pulp
coming from the plates has about 40 per cent, of its gold locked
up in the pyrites, and 60 per cent, in the form of finely-divided

free gold which cannot be removed by the
vanning machines
but is readily extracted by cyanide. It follows that concentra-
tion and chlorination are not sufficient of themselves,
although
the cyanide process may be, and for this reason a Committee of
the Johannesburg Chamber of Mines reported in May, 1895,
that close concentration is usually quite unnecessary. Doubtless
the best percentage extraction would be obtained in almost every
case by classifying the pulp, concentrating the various classes
separately, chlorinating the concentrates, and cyaniding the
tailings. It would appear, however, that, under present con-
ditions, the greatest profit is gained by rough classification in
spitzluten, followed merely by cyanide treatment.
The treatment of amalgam presents no unusual features. At
the Simmer & Jack mill there is a cast-iron clean-up pan, 3J feet
in diameter, with mullers making 60 revolutions per minute.
The hard amalgam, skimmings, and headings from the battery
are ground in this pan, with fresh mercury, for three or four
hours ;
after this the mercury is strained through chamois
leather, and hard balls of amalgam, which contain from 33. to
35 per cent, of gold, are thus obtained ready for retorting.
The life of the various parts of the mills is approximately as
follows : The shoes last for two to four months, the dies from
three to four months, and the screens about fourteen days ; when
banket is being crushed the wear is less than if quartz is in
course of treatment. Tappets have been in use for three years
without showing any signs of wear. The stems of the stamps
become crystalline and the ends usually break off after about
twelve or eighteen months' use, but this can be avoided by
annealing at regular intervals. The mortar boxes last at least
four years, the cam-shafts about three years, and the guides
twelve months.
The amount of ore crushed is usually over 4 tons of banket
per stamp per diem. In 1891 the average output was from 2J
to 3J tons per day; but is now greater (viz., about 4J tons)
owing to the greater average weight of the stamps. At the
May Consolidated Mine the 1,150-lb. stamps crush 5^ tons per
day, and Hatch and Chalmers consider that, in a few years time,
from 5 to 6 tons will be a common result. The large output is
due partly to the nature of the rock, banket being softer than
ordinary quartz, and partly to the small depth of discharge and
the coarseness of the screens.
In present practice there is really no difference between
English and American mills as regards design. It is generally
acknowledged that stamps of less weight than 950 Ibs. should
not be used, and the tendency is to increase this weight. The
fittings, such as shoes and dies, are almost universally
made of
English steel. It may safely be said that English mills made
by the leading makers compare favourably with the best

American mills.
Taking an average of the whole of the operations on the
Rand, the yield of gold on the amalgamated plates is gradually
diminishing^ having been 13-5 dwts. per ton in 1890, 11-3 dwts.
in 1891, 9-77 dwts. in 1892, 9-59 dwts. in 1893, 9-52 dwts. in
1894, and less than 8 dwts. at the end of 1895. In the last-
named year the thirty-seven leading companies on the Rand
crushed 2,485,311 tons of ore, and obtained 1,182,794 ozs. of
gold on the plates (or 64 per cent, of the total gold contents),
37,138 ozs. by concentration and 486,082 ozs. from the tailings
by cyanide treatment.
The cost at the best mills is from 3s. to 5s. per ton for
crushing and amalgamating, and about 4s. to 6s. per ton for
treatment of tailings by the cyanide process. Concentration
and chlorination probably cost about 3s. to 3s. 6d. per ton of
ore. The following details of the work done in two mills in
twelve months in 1894-95 are given as examples of what is
possible with good management :
CHLORINATION. 215
CHAPTER XL
CHLORINATION :THE PREPARATION OF ORE
FOR TREATMENT.
The Plattner Process. The value of chlorine, as an agent for
the extraction of gold from certain ores and from almost all
concentrates, has now been recognised for many years, and its
use is gradually extending, although it is doubtful whether its
application will ever be as general as appeared probable before
the introduction of the cyanide process. Its use was first
suggested by Dr. John Percy, F.R.S., at the Swansea meeting of
the British Association, held in August, 1848, in a paper * em-
bodying the results of experiments carried out in the year 1846.
Simultaneously, in 1848, Prof. C. F. Plattner, Assay Master at the
Royal Freiberg Smelting Works, applied chlorine gas to the assay
of the Reichenstein residues, and proposed that a similar method
of treatment should be adopted on a large scale. These
residues were the result of treating the Reichenstein ore with
the object of extracting the arsenic. They consisted chiefly of
oxides of iron and oxidised arsenical compounds, and had been
roasted in the course of the process for the extraction of the
arsenic. The residues had been accumulating for more than a
century, and contained from 15 dwts. to 1 oz. of gold per ton ;
they were considered too poor to smelt, while they could not be
made to yield the gold contained in them by amalgamation.
Prof. Plattner's suggestion was followed up by investigations
made by Dr. Duflos in 1848,f and by Lange in 18494 Dr. Duflos
compared the results obtained by treating the residues with
chlorine water by percolation in a stationary vat, and by agita-
tion in a revolving barrel ; and as these results were the same,
he recommended the stationary vat as being more economical.
He also obtained identical results with chlorine-water and with
dilute solutions of chloride of lime and hydrochloric acid mixed
together. On the other hand, Lauge found that gaseous chlorine,
applied to the ore in the same manner as had been used by
Plattner in assaying, was a more efficient agent than a solution
of chlorine in water, and it seems to have been in accordance
with his advice that the first chlorination works, that at
Reichenstein, was established in 1849. The chlorinating vessels
*
Phil. Mag., 1853, vol. xxxvi., pp. 1-8.
t Erdm. Journ. Prak. Chem., vol. xlviii., 1849, pp
Karsten's Archiv., vol. xxiv., 1851, pp. 396-429.
were small earthenware pots, and the precipitant employed was

sulphuretted hydrogen. Plattner subsequently recommended
wooden vats coated with pitch, and ferrous sulphate as a pre-
cipitant, and although these were not at first used at Reichen-
stein, they were adopted by Mr. G. F. Deetken in 857, when he
1
introduced the system into California. The system ascribed to

Plattner consists of the following operations :
1. The concentrates or residues are subjected to a

" dead " roast
in a reverberatory furnace.
2. The roasted ore is slightly damped with water and
charged
into wooden vats, holding from 1 ton upwards. The vats have
false bottoms consisting of filter beds of gravel or of cloth.
Chlorine gas, generated in another vessel, is introduced at the
bottom of the vat, and rises through the ore; permeating every
part of it. The vat is then closed up and left undisturbed for
twenty-four hours or more, by which time all the gold is con-
verted into soluble chloride of gold.
3. The soluble salts are then washed out by water, which is
allowed to flow on to the surface of the ore, and, passing through
it, drains through the filter bed. When all the gold has thus been
removed in solution, the tailings are thrown away.
4. The solution of gold chloride is acted on by ferrous
sulphate, or some other suitable reagent, by which the gold is
precipitated the particles of the precious metal are allowed to
;
settle, and then are collected and melted down.

Plattner Process at Reichenstein. The following descrip-
tion of the original process employed at Reichenstein is an
abstract of the account given in Kerl's Huttenkunde, vol. iv.,
p. 372, 1865. The material treated consisted of the residues
obtained by roasting arsenical iron pyrites for the production of
white arsenic, which was volatilised and condensed in brick
chambers. There were forty-eight earthen chlorination pots,
each holding 150 Ibs. of ore. These pots were strengthened with
iron hoops, and suspended on two journals, so that they could
be discharged by inverting them.
The lower part of the pots was of a conical shape, and this
part was filled with pebbles and sand covered by a perforated
earthen plate, the function of which was to prevent the ore from
mixing with the filter bed. The ore filled the cylindrical part of
the pot above the earthen plate. The chlorine was generated by
the action of hydrochloric and sulphuric acids on manganese
dioxide in earthenware vessels, and was conveyed thence to the
ore-pots through leaden pipes. The gas was introduced below
the filter bed, and passed upwards through the ore for an hour ;
a wooden cover was then fitted on, but not luted down until
chlorine had been passed for from six to seven hours longer,
after which all joints were luted down with dough, and the vat
left until the next day. The cover was then removed and water,
CHLORINATION. 217
at a temperature of from 64 to 77 F., poured on, and allowed

to percolate through the ore and filter bed by gravity. The
liquid coming from twenty-four pots was conveyed to four vats,
the first one being filled with solution before the second was
used, and so on ; the contents of the fourth vat, being too poor
for precipitation, was used over again for leaching. The leaching
was stopped when 90 cubic feet of water had passed through the
total charge of 3,600 Ibs., this being at the rate of 312 gallons
per ton of 2,000 Ibs. The liquid from the first three vats was
drawn off into twenty glass globes, which were heated on a sand
bath so as to raise the temperature to 77 F. Sulphuretted
hydrogen, obtained from fused sulphide of lead and sulphuric
acid, was passed through until the saturation point was reached,
when the liquid was left to settle until the next day ; after this
the clear supernatant liquid was passed through sawdust filters
to catch all the sulphide of gold, which might still have been in
suspension. The sulphides were refined by dissolving them in
acids, precipitating the metallic gold by ferrous sulphate, and
melting it in clay crucibles with nitre and borax. The
amount of arsenides treated daily was 3,600 Ibs., containing
about |- oz. of gold per ton, so that only about 250 ozs. of gold
were extracted yearly, and it is difficult to believe that the
enterprise could have been a commercial success.
MODERN PRACTICE IN CHLORINATION.

The original vat process as described above has been subjected
to great changes, many of which originated in the United States.
The most notable alterations are an increase in the size of the
vats and the substitution of wood for earthenware in their con-
struction, and the use of ferrous sulphate instead of sulphuretted
hydrogen to precipitate the filtered liquors. This process, as
altered, is still in very extended use for dealing with roasted
concentrates, but in a few special cases, where large quantities
of material are available for treatment, the barrel process has
been adopted. The barrel process of chlorination, which was
first used on the large scale in the United States, differs from
the vat or Plattner process chiefly in the fact that the chlorination
is effected in
revolving barrels, instead of in stationary vats,
while the operations of chlorination and lixiviation are usually,
though not invariably, performed in different vessels. The
two processes will be described under the following headings :
1. Crushing.
2. Roasting.
3. Chlorination and Lixiviation.
4. Precipitation of the gold from its solution and production of bullion.
Vat and barrel chlorination differ essentially only in the

methods described under the third heading, although the differ-
ences in the conditions under which they are applied involve

variations in the treatment under the other headings. The
sections devoted to crushing and roasting include descriptions
of the methods used to prepare ore for chlorination both in the
vat and the barrel. The remaining operations are described
separately for each of the two processes, and a general account
of the precipitation of gold is added.
CRUSHING.
In those cases in which gold ores are treated by crushing and
amalgamation, and the whole of the tailings, or only the con-
centrates obtained from them, are subsequently chlorinated,
the method of crushing will be determined by the considerations
already discussed in the chapters on amalgamation and concen-
tration, and will depend partly on the state of aggregation of
the free gold. If, on the other hand, ores are to be treated in
the first instance by chlorination, special regard may be paid to
the method of crushing as affecting the suitability of the crushed
product for leaching. There are two somewhat opposite con-
ditions to be fulfilled, viz.:
1. The crushed product should be fine enough to admit of the
whole of the gold being laid open to the attack of the chlorine.
2. It must be coarse enough to allow all the soluble chloride
of gold thus formed to be washed out of the ore rapidly and
easily.
It is quite impossible to fulfil completely both of these con-
ditions on a large scale, although in the laboratory some ores,
when treated with the greatest care and patience, may be made
to yield the whole of their gold. In practice it is better to aim
at an extraction of from 80 to 95 per cent. (98 to 99 per cent,
being attained in some rare cases), and to do this at as low a
cost as possible. With this object in view, the ore is kept as
coarse as possible, and is usually passed through screens with
only from 8 to 30 holes to the linear inch. The chief point to
be attended to is the attainment of uniformity in the product,
as any considerable proportion of slimes enormously increases
the difficulties of leaching. Moreover, since the crushed ore
must almost invariably be roasted, it is a great advantage to
adopt some method of dry crushing. Proposals have been made
to subject the ore to wet-crushing by stamps or other machines,
and to collect the pulp in settling-pits, and dry it, either in
the roasting furnace or in a separate furnace. It appears that
this method has never been tried, and there are several practical
objections to it.
Dry crushing by means ot stamps has been found to answer

well at Park City, Utah, and at other places in preparing silver
CRUSHING. 219
ores for treatment by the hyposulphite leaching process, but

many ores would form too much impalpable powder for this
system to be applicable. In order to obtain a perfectly uniform
product in any machine, it would be necessary for every particle
that had been reduced to the required degree of fineness to be
instantly separated from those which were still too large. If
the screening is not instantaneous and thorough, the small
particles are ground still finer while the larger pieces are in
process of reduction to the required fineness. Of all the
machines for dry crushing which have been designed with a
view to effect this object, rolls have come into the most general
use, and give the most satisfaction. They are used, among
other places, at Mount Morgan, the largest chlorination mill in
the world, and at the Dakota Mills. Rock-breakers are now
always employed to reduce the ore to the size of a hen's egg, or
smaller, before feeding it to the rolls.
The following description of Krom's rolls, the best known
modern make of high-speed fine-crushing rolls, is chiefly derived
from Mr. A. H. Curtis' paper on "Gold Quartz Reduction."*
Crushing rolls made by several other manufacturers in various
countries might be described almost in the same words, and
give nearly if not quite equally good results in practice, while
they are much lower in price.
The rolls are placed with their faces a short distance apart,
this distance varying with the degree of fineness to which the
ore must be reduced. The main driving power is applied to the
shaft of only one roll, by means of belt pulleys, the other roll
being driven only with sufficient force to ensure that the rollers
will always take'hold of the ore, and also to keep them in motion
when no ore is passing between them. If all the power is
applied to turn one roll only, the feeding must be continuous, or
the free-running roll would come to a standstill. On recom-
mencing the feeding, a blow would then be given to the stationary
roll, in the attempt made by the other roll to
set it instan-
taneously in motion at full speed, and this jar would have a

disastrous effect on the machinery.
In the older forms, tooth-gearing was used instead of belt
the two rolls were
pulleys for the application of the power, and
compelled to revolve at an equal rate of speed by gear-wheels,
connecting together, placed on the axles. The advantages
of
belted rolls are, that a higher speed can be easily attained, and
also that if the rolls were to become jammed from any cause, the
belts would slip or be thrown off, while the tooth-gearing would
be broken. Geared rolls, however, are still largely used for
coarse crushing. The rolls have crushing tires made of forged
steel, and are
firmly and secured to the shafts by means
simply
of cone-shaped heads. Chilled-iron rolls are still often used,
ii.
*Proc. Inst. Civil Eng., session 1891-92, vol. cviii., part
being much cheaper; but the wearing of the faces is more rapid
and less uniform than in the case of steel rolls. Emery wheels
for levelling the unevenly worn faces of chilled-iron rolls are
recommended by some makers. These crushing tires can be
taken off and replaced when they are worn out. In the older
forms of Krom's rolls the crushing strain is taken up by powerful
springs, which press the rolls towards one another when par- ;
ticularly hard fragments are passing through the rolls, they are
forced apart against the action of the springs. These springs are
now dispensed with. It is desirable, in order to keep the wear
of the faces even, that the rolls should always be kept parallel,
and special appliances, such as Krom's swinging pillow blocks,
have been introduced by some makers to ensure this. The hopper
is specially designed to spread the ore evenly across the crushing
face, and the rolls, screens, elevators, &c., are all securely boxed
in with a wooden housing. This last precaution is necessary in
order to prevent loss by floating dust, which otherwise may be
large, the richest part of the ore thus passing off, and not only
making the atmosphere of the mill insupportable, but having a
disastrous effect on the bearings of the machinery. Rolls are
usually from 12 to 16 inches across the face, and from 22 to 36
inches in diameter.
Corrugated rolls are used to some extent in Australia, but
though the crushing surface is increased in this way, they are
considered in the United States to be of little value, wearing
unevenly and soon getting out of order.
The method of crushing usually adopted in mills where rolls
are employed may be described in general terms as follows. It
may be supposed that the ore is to be passed through a 20-mesh
screen :
The
ore is put through a rock-breaker, and passed at once to
classifying screens, by which it is divided into four classes of
material. One of these consists of ore that will pass a 20-mesh
screen, and this amount (usually small) is separated as finished
product another small portion is returned to the rock-breaker
;
by some form of elevator, being too coarse for the rolls, and the
greater part is passed to one or other of two pairs of rolls. Each
of these pairs of rolls is occupied in crushing material of a
particular degree of fineness, reducing it to a further degree of
fineness. The material fed to them is classified accordingly.
After each passage through the coarse rolls, a certain amount of
finished product is obtained, and the remainder is classified, part
being returned to the same rolls, and part being sent on to the
fine rolls, the product from which is also classified. Revolving
screens are in general use for classification. The use of two
rock-breakers, one for coarse crushing, and the other to take its
product and reduce it to the size of broad-beans, is often recom-
mended. The number of pairs of rolls to be used in succession
CRUSHING. 221
depends on circumstances two is the usual number, a third pair

:
being occasionally added. One rock-breaker and two pairs of

rolls will suffice for the reduction of most ores to a size sufficiently
fine to pass through a 30-mesh screen, but the crushing can be
done more economically in the long run by still more gradual
reduction.
Mr. John E. Rothwell, late manager of the Golden Reward
Chlorination Mill, Dakota, gives it as his opinion* that "two sets
of rolls are sufficient, but three will do better. The ore should
come to the coarse rolls not coarser than j-inch mesh, and these
rolls should be set about f inch apart. The middle rolls are set
about ^ inch or less apart, and the fine rolls about as far apart
as the size to which the ore has to be crushed. If only two sets
are used, the coarse are set a little closer than with three, and
the fine remain the same.
" The
springs should be set up so tight that they will not give
to the hardest pieces of ore, but will allow a piece of steel or
iron to pass through without throwing the belts. The periphery
speed of the rolls should be about the same as, or a little faster
than, the falling speed of the ore, and the ore should be fed in
an even sheet across the surface of the roll ; little trouble will
then be experienced in keeping the surfaces true and in pro-
ducing a granular pulp, carrying but a small percentage of dust.
If rolls were made of larger diameter and narrower, the result
would be a still more gradual reduction, and possibly a greater
capacity. I have used those of 39 i inches (1 meter) in diameter
and 12 and 15 inches face."
"It is stated by users of the Krom rolls (which claim to be
essentially fine pulverisers) that the faces of the tires wear
evenly and do not become grooved, and that they have a long
life. At the Bertrand Mining Coy.'s Lixiviation Mill, Nevada,
where Krorn's rolls are used for pulverising silver ores with a
quartz gangue, 15,000 tons of ore were crushed before it was
found necessary to put new tires on the finishing rolls ; while
after a further crushing of 5,000 tons of ore, the tires of the
roughing rolls were still expected to be good for two or three
months' work. In neither case were the tires found to be at all
grooved, the reason for their renewal being that they had
become worn too thin for further work. The rolls in the
Bertrand Mill are stated to crush 50 tons of hard quartzose
ore in twelve hours to pass through a 16-mesh screen while in ;
the Mount Cory Mill, Nevada, 50 tons are reduced to 30-mesh

in the same time." f In these mills silver ores are crushed for
treatment by roasting and lixiviation with a hyposulphite solu-
tion. It is possible to crush quartz to pass through screens of
from 30- to 70-mesh by rolls, the degree of fineness to which the
*L ng. and
J
Mnrj. Journ., Feb. 7, 1891.

t Curtis on "Gold Quartz Reduction." Proc. Inst. Civil Eng., 1892.
222 t
'THE METALLURGY OP GOLD.
rock can be pulverised depending on the number of times the

product, after screening, is returned to the hopper. The very
fine dust, which is inevitably produced in
greater or less quantity,
according to the character of the rock, may be separated by
exhaustion with a fan, driven at carefully regulated speed, and
in this way the sliming is reduced and the resulting loss of gold
in the subsequent operations minimised. The average speed of
fine-crushing rolls is from 80 to 100 revolutions per minute, the
two rolls being driven at equal rates of speed.
Crushing by rolls at Rapid City, Dakota. The following
details concerning the crushing plant at the Rapid City Chlor-
ination Works, S. Dakota, which was fitted up in 1891, have been
supplied by Mr. D. Dennes, formerly the foreman of the mill.
" Gates crushers were used for some
months, after which they
were replaced by a large Blake crusher, which is found to w ork r
more economically and with less cost for repairs. The crushed
material from the Blake is passed through a rotating cylindrical
drying furnace, and then screened, part being passed to the ore-
bin set aside for finished product, a little going back to the rock-
breaker, and the remainder going to the first pair of rolls. These
have steel tires and are set at ^ inch apart; they are driven
at a speed of 90 revolutions per minute by means of toothed
gearing, both rolls being driven independently. The objections
to the use of belts for the coarse rolls are stated to be that a
specially hard piece of ore may throw the belt off and so
interrupt the work. The peripheral speed of the rolls is 14 feet
per second. The product of the coarse rolls is classified by
screening, and the bulk of it sent at once to the fine rolls, which
are steel-faced belted rolls driven at 155 revolutions per minute.
The two faces just touch, but each roll is driven by a separate
belt, and one is revolved at the rate of two revolutions per
minute faster than the other. This arrangement has a remark-
ably beneficial effect in keeping the wearing surfaces even. The
tires are replaced alternately, so that there is always one new
and one old tire working together. Almost all the product of
these rolls passes at once through a 16-mesh wire screen, which
is the finest screen used.
" The
revolving screens of Messrs. Fraser & Chalmers were
used for six months, and then discarded. Flat rectangular
screens are now used, measuring 6 feet by 12 feet, set in a
wooden frame which is subjected to a reciprocating motion,
striking against a rubber pad placed close to one of its long
sides. The screen is slightly inclined, so that the ore, which is
fed evenly across the screen, travels down it. This apparently
retrograde step in screening was justified by a considerable
improvement in efficiency and a reduction of expense. The
revolving screens were found to wear out very rapidly, and the
repairs were costly. The weight of ore causes the metal filter-
CRUSHING. 223
cloth to sag down between the bars constituting the framework

;
when this part of the cylinder reaches the top of its course, the
cloth sags in the opposite direction from its own weight. The
result of this double motion is that the sieve breaks
along a
line close by and parallel to the framework bars. "When this
breakage occurs, repairs are tedious and expensive, and the mill
must be stopped while they are being done, while flat screens
can be lifted out, and a new one put in, by two men in a few
minutes."
From 100 to 120 tons of ore are stated to have been crushed
in this mill per day, but as the mill was in continuous operation
for a short time only, and is now shut down, it is not possible to
draw definite conclusions from the results obtained. The speed
of the rolls seems to have been excessive, and the difficulties-
with the revolving screens do not appear to have been felt in
other mills. Probably they were due to the fact that wire-cloth
was used for the screens, instead of the more serviceable punched
sheet-iron.
Comparison between Rolls and Stamps. As the subse-
quent treatment of an ore determines its method of crushing, no
accurate general comparison of stamps and rolls can be made.
A comparison is only possible in the special cases where both
methods of crushing are applicable. Wet crushing by rolls need
not be considered, as it is not practised ; even where the advo-
cates of rolls wish to replace stamps by rolls for wet crushing and
amalgamation, they propose that the ore should be crushed dry
and then wetted down. Dry crushing by stamps is usually
about one-third more expensive than wet crushing, as the-
capacity falls off to that extent, in spite of double discharge,
assisted occasionally by currents of air given by blowers, which
are designed to carry the crushed ore against the screens, 'i he
amount of slimes made is also large. Advantages in the use of

rolls for dry crushing have been stated to be " the fewness of the
wearing parts, and consequent small cost of repairs ; the great

efficiency of the process, in that the ore escapes from the rolls
immediately it is crushed, so that over-crushing is unlikely to
occur, and the great capacity of rolls, eifective work being
constantly done, and the amount of crushing surface brought into
contact with the ore per minute being very large. The prime
cost of the rolls is considerably less than that of stamps of the
same capacity." *
The capacity of rolls has perhaps been frequently over-
estimated owing to the assumption having been made, that the
product of equal crushing surfaces must be the same. Thus r
Mr. Curtis observes in the paper already cited "As an :
index of the capacity of Krom's rolls, it may be stated

that two sets of 20-inch (diameter) rolls, with faces 15 inches.
*
Curtis, loc. cit.
long, give rather more effective crushing surface than fifty-

gravitation stamps, each 8 inches in diameter, falling at the rate
of ninety drops per minute. In making this calculation, the
average diameter of the rolls is taken as only 24 inches, so as to
allow for their gradual wearing, while their speed is taken at
100 revolutions per minute."
Although the calculated crushing surface is as stated, it does
not follow that the capacity of the two sets of machinery com-
pared with one another is the same, since it will depend on the
pressure as well as the crushing surface. No doubt the pressure
at the moment of impact of a 900-lb. stamp is enormously greater
than that exercised by the faces of the rolls.
On this point Mr. T. Richards, M.E., writes:* "The co-
efficient of (crushing) effect in such machines (as revolving rollers
and reciprocating machines) is as the area of the acting surfaces,
and the speed with which they approach each other. The
mistake in respect of the crushing power of rollers comes from
confounding their circumferential velocity with the working one
that is, with the parallel velocity at which the surfaces approach
and leave each other," that is, the velocity at right angles to
the crushing surface. Mr. Richards elsewhere shows that a
Blake machine in this way actually outruns a revolving roller,
although its crushing face travels more slowly than the periphery
of the roll. He proceeds " The same
:
principle holds good in
respect to stamps, the crushing surfaces having a parallel
approach, while with rotary machinery the approach is not
parallel, except on an imaginary line at the centre." As a
matter of fact, the product of two pairs of rolls amounts to only
about 40 to 50 per cent, of that which would be given by a fifty-
stamp mill on the same ore. In the absence of exact comparative
data regarding the same ore, no precise figures regarding the
relative capacities can be given.
A point to be noticed is that, in machines which act by
pressure applied on the principle of the lever, such as recipro-
cating-jaw crushers or rolls, the whole force necessary to crush
the lumps of quartz is transmitted to the fulcrum, this fulcrum
being represented in the case of rolls by the bearing surface at
the axle. The consequence is that the frame must be made very
strong and heavy, and the axle bearings attended to with great
care, if rapid wear is to be avoided.
Drying the Ore. Although rolls are essentially dry crushers,
nevertheless, if the ore is nearly pure quartz, a small percentage
of moisture in it is not a serious disadvantage. If, however,
the moist ore is argillaceous, the product from the rolls will not
readily pass through the screen, and the latter soon becomes
clogged, frequent stoppages for cleaning being thus necessitated.
It has been stated that 5 or 6 per cent, of moisture present in an
*
Prod, of Gold and Silver in the United States, 1880, p. 369.
ROASTING. 225
ore does not seriously interfere with crushing, but this does not
accord with the experience in some mills. At any rate, in every
mill where dry crushing is used, some means of artificially drying
the ore is adopted.
The oldest method was to spread the ore, after the large lumps
had been removed by a grizzly and crushed to 1 J inch size, on
large flat areas heated from below by flues from the roasting
furnace. The floor was usually covered with iron plates. After
being dried, the ore was shovelled up and passed to the crushing
mill. This plan involved much additional handling of the ore,
and was a source of ill-health among labourers, besides requiring
great floor space. It has been superseded by the adoption of
inclined continuous-discharge, revolving iron cylinders, similar
to the Howell-White furnace, but not lined with bricks. The
ore is passed through these, and is dried by the products of com-
bustion of a fire, which are also passed through it. One such
cylinder, of 3 feet in diameter and 18 feet long, will dry from 30
to 40 tons of ore per day, at a small cost for fuel and power.
An alternative furnace viz., Stetefeldt's shelf drying kiln
was described in a paper read by the inventor at the meeting of
the American Institute of Mining Engineers, held at Eoanoke,
Virginia, in June, 1883. In principle it resembles the Hasen-
clever furnace, a number of shelves being arranged zig-zag above
each other in a vertical shaft, down which the ore slides, falling
from shelf to shelf, while the products of combustion from a
furnace rise through it. It is 21 feet high, and dries from 30 to
50 tons of ore per day. It is in wide use in the United States,
and although its first cost is considerable, its working expenses
are said to be lower than those of the rotary furnaces.
BOASTING.
The operation of roasting, as a preliminary to chlorination,
has for its the expulsion of the sulphur, arsenic,
object
antimony and other volatile substances existing in the ore,
and the oxidation of the metals left behind, so as to leave
nothing (except metallic gold) which can combine with chlorine
when the ore is subsequently treated with it in aqueous solution.
For this purpose the ore is heated in a furnace, through which a
current of air is passed, salt being added if oxide of copper, lime,
magnesia, &c., are present. Ores containing much pyrites might
be freed from most of their sulphur by pile roasting, and then
subjected to fine crushing and a dead roast in a reverberatory
furnace, but the extra cost of handling would probably exceed
the saving due to the smaller consumption of fuel. This system
has not been tried in chlorination mills on an extensive scale.
The ordinary reverberatory furnace, worked by hand labour, is
lo
in wider use than any other, especially where only a few tons,
or less, of concentrates are to be treated per day. Various
mechanical furnaces, capable of handling large quantities of ore,
have been devised to supersede the old-fashioned contrivance,
and some of these will be described in the sequel.
Reverberator?/ Furnaces. The construction of the ordinary
reverberatory furnace is too well known to need detailed descrip-
tion here.* It consists of a vaulted chamber, containing the
ore ; through this chamber, the flames and products of combus-
tion from a furnace and a current of air are made to pass in a
horizontal direction above the ore, which is thus heated. The
ore is also stirred by hand with iron rakes, which are passed
through small working doors. The hearth of the vault (also
called the "laboratory" of the furnace) is formed of bricks placed
on edge (not flatwise, except where economy is studied rather
than durability), as close together as possible. No mortar is
used, but a little clay is plastered between the bricks. The
height of the furnace hearth is about 3^ feet above the floor of
the building, on which the labourers stand, and the space under-
neath the hearth is either occupied by vaults or filled with well
tamped rubble. The arch is usually one course of bricks
(8 inches) thick ; the height between it and the hearth is, in
long furnaces, about 24 inches near the bridge, and gradually
diminishes towards the other end. This height is less in
short furnaces. The best fire-bricks are used for the fire-box and
bridge, and for the hearth and arch of the first few feet of the
"
laboratory." The remainder is made of common brick. It is
necessary to have a damper in the flue to regulate the draught;
the aperture of the flue should not be on a level with the hearth,
as in that case the loss by dusting is increased. The brickwork
of the furnace is supported by longitudinal and transverse iron
braces. The working doors have cast-iron frames, and are about
15 inches wide and 9 inches high. The fuel used must of course
be a long-flame coal, or wood; short-flame coal and coke are
inadmissible.
For the particular purpose of roasting pyritic ores before
chlorination, the temperature on the working floor of the furnace
must be low when the ore is first charged in, and high in the
later stages. If a single small floor is used, the fire must be
alternately checked and urged to secure these conditions. More-
over, when the roasting is nearly complete, the high temperature
required renders the gases passing into the flue very hot, and so
a corresponding waste of fuel results. To prevent this waste, it
is customary for roasting furnaces to be built with a very long
hearth, or to have several successive hearths, so as to utilise the

*
See the Introduction to the Study of Metallurgy, by Prof. Roberts-
Austen, 1894, p. 207. Also Kustel's Roasting of Gold 'and Silver Ores, 1880,
pp. 80-89 ; and Griiner's Traitt de Metalluryie, p. 265.
ROASTING. 227
waste heat from the portion of the working floor next the fire.
Furnaces with three floors at slightly different levels are much
favoured ; in these a charge remains for a few hours on each of
the floors in succession. It is first placed on the floor farthest
removed from the fire, and, after a time, is raked down on to the
middle hearth, and thence to that nearest the fire, fresh charges
being put on the spaces just cleared, so that there are always
three charges in the furnace in various stages of oxidation.
The most usual form in the Western States of America is the
"
4-hearth," in which the length of the hearth is four times its
width, so that the dimensions are, say, width 15 feet, length
60 feet. In this case there should be eight working doors on
each side. Instead of three floors at different levels, a single
continuous floor, gently sloping from the flne towards the fire, is
in use at many works in Australia, Mexico and the United
States. At Suter Creek a continuous-hearth furnace is 12 feet
wide and 80 feet long, and the mineral is worked in three
distinct parts, as though there were three floors. The angle of
slope is made large in some Australian furnaces, so that in the
"
course of the "rabbling or stirring, the ore continually travels
towards the fire-box. Furnaces with two or three superposed
floors are also used to a limited extent ; the lowest floor is next
the fire-box, and communicates by a vertical flue with the floor
above, and so on. The ore is charged-in on the top floor, and
after a time is raked down through the vertical flue on to the
next floor. In this case the floors are heated by the gases passing
below them as well as above them, and fuel is economised, but
tho furnaces are costly to build and to keep in repair.
It was proposed many years ago to insert a drop of 10 feet
between the finishing floor and the floor next to it. The charge,
when already red hot, would thus fall vertically downwards
in a thin shower against a current of hot air. Some furnaces
are said to have been built in this way, but it seems that none
are now in existence. The principle is excellent, and is utilised
in the Stetefeldt furnace.
When the pyrites to be roasted are rich, it may be an ad-
vantage to build dust chambers to ordinary reverberatory
furnaces. The amount of gold contained in the dust thus
recovered is usually only 1 or 2 per cent, of that contained in
the ore, so that in some cases it may be a long while, before the
dust chamber pays for itself, even if that point is ever reached.
The operation of roasting pyrites in an ordinary furnace with
three floors may be described as follows: The furnace being
hot, and the flame from the fire-box reaching completely across
the first floor, the ore is charged-in on the third floor and spread
out by the rabbling tool. The weight of the charge may be
taken as from 12 to 18 pounds per square foot of floor space,
varying according to the nature of the ore, a high percentage
of sulphur necessitating small charges. The layer of ore is 2

or 3 inches deep. It is not spread quite evenly, but made
to form a series of parallel ridges by means of the rabbling tool,
so as to increase the surface exposed to the air. The working
doors may be closed at first to heat the ore quickly. Moisture
is at once
given off in great quantities, and the sulphur soon
begins to burn with a blue flame. When this is seen to take
place, all the working doors are opened, and the charge is ener-
getically rabbled, with little intermission, until the sulphur
name disappears. If this is not done, clots are formed which
are afterwards difficult to break up. The air for the combustion
of the sulphur is supplied by holes in the fire-bridge and from
the working doors. The flames and heated products of com-
bustion from the fire tend to rise above the colder air, and
move along next to the arch of the furnace, while the air forms
a sheet between these gases and the ore. In practice, although
the air is introduced below them, nevertheless the reducing
gases from the fire partially mix with the air, and greatly
reduce its oxidising power. Moreover, the combustion of the
sulphur in the ore on the first two floors further reduces the
amount of free oxygen present in the current of air, and
roasting on the third floor is, therefore, largely dependent on
air derived from the working doors.
When the sulphur flames have abated, and the charge has
been heated almost to redness, it is transferred to the middle
floor, where it is raised to a dull red heat and most of the oxida-
tion is performed ; during this stage the ore swells considerably,
so as to occupy much more than its original bulk. All the lumps
previously formed should be broken up on this floor. Rabbling
is continued until the ore is uniformly dull throughout, so that,
on turning it over, the fresh surfaces appear but little brighter
than that which has been exposed for some time. The charge is
then transferred to the floor next the fire. There is now little
risk of the formation of lumps, and the charge may be allowed
to reach a bright red heat. Rabbling is of less importance than
before, as little oxidation takes place, the chief reaction which
occurs being the decomposition of the sulphates already formed.
As long as this is still going on, the ore emits the odour of
sulphur dioxide. When no further odour can be detected, and
the ore can be piled up so as to maintain a vertical face, shows
no bright specks on its glowing surface, emits no sparks if
some of it is tossed up by the working tool, and is inclined to
become black very readily from cooling, the charge is said to be
" dead " or "
sweet," and is ready to be withdrawn. It should
be observed that, when the ore contains much sulphur, its
particles at a low red heat appear less coherent than when cold,
and flow almost like water so that the charge cannot be made
to form a heap with steep sides. Care must therefore be taken
ROASTING. 229
when the ore is on the middle floor to prevent any

part of the
charge from flowing out of the working doors, which it is very
liable to do when being rabbled.
Kiistel states* that the best means of rapidly
ascertaining
whether a charge is completely roasted is to throw a little of
the ore into some water, and then to plunge a bright iron rod
into the liquid. If the rod remains bright the ore is ready
for withdrawal, but if sulphates still remain undecomposed,
the surface of the iron will be darkened. This is not a safe test
with all classes of ore, as the presence of sulphate of iron in the
water would not be detected in this way. A more trustworthy
and equally simple test is to add a few drops of chloride of
barium to the water. A white cloud, which consists of BaS0 4 ,
indicates the presence of soluble sulphates. In most cases

the water need not be filtered before it is tested, but even
when nitration is necessary the whole operation can be per-
formed in two or three minutes. The charge is withdrawn
by a scraper, and falls by gravity through a hole in the floor
of the furnace near the working door into a pit below. This
hole is covered by a plate while roasting is being performed.
The time of roasting depends chiefly on the ore, but may be
shortened by more continuous rabbling than most workmen can
perform, the work being somewhat exhausting. In a three-
floor furnace, concentrates with 15 per cent, of sulphur usually
remain eight hours on each of the three floors. The fuel used
is either wood or flaming-coal. If the flame from the coal is not
long enough to reach across the first floor, this will not be heated
uniformly in that case the part of the charge next the fire-bridge
;
is finished first and must be moved away, while that from

the other end of the floor is brought up nearer to the fire.
This causes a great increase in the labour, besides occasionally
leading to the withdrawal of a charge of which a part is not quite
"dead." The draught is regulated by the damper in the flue
leading to the dust chamber, and by opening and closing the
working doors.
Chemistry of Oxidising Roasting. Professor Koberts-
Austen discusses as follows f the roasting of a " mixture
consisting of sulphides mainly of iron and copper, with some
sulphide of lead, small quantities of arsenic and antimony as
arsenides, antimonides, and sulpho-salts, usually with copper as
a base. The temperature of the furnace in which the operation
is to be performed is gradually raised, the atmosphere being an
oxidising one. The first effect of the elevation of the tempera-

ture to distil off sulphur, reducing the sulphides to a lower
is
stage of sulphurisation. This sulphur burns in the furnace

* San Francisco, 1880.
Roasiinfj of Gold and Silver Ore*.
t Presidential address to the Chemical Section, British Association,
Cardiff Meeting, 1891.
atmosphere to sulphurous anhydride (S0 2 ), and coming in

contact with the material undergoing oxidation is converted into
sulphuric anhydride (SO 3 ). It should be noted that the material
of the brickwork does not intervene in the reactions, exceprt by
its presence as a hot porous mass, but its influence is, neverthe-
less, considerable. The roasting of these sulphides presents a

good case for the study of chemical equilibrium. As soon as the
sulphurous anhydride reaches a certain tension, the oxidation of
the sulphide is arrested, even though an excess of oxygen be
present, and the oxidation is not resumed until the actions of
the draught change the conditions of the atmosphere of the fur-
nace, when the lower sulphides remaining are slowly oxidised,
the copper sulphide being converted into copper sulphate, mainly
by the intervention of the sulphuric anhydride, formed as in-
dicated. Probably by far the greater part of the iron sulphide
only becomes sulphate for a very brief period, being decomposed
into the oxides of iron, mainly ferric oxide, the sulphur passing
off. Any silver sulphide that is present would have been
converted into metallic silver at the outset were it not for the
simultaneous presence of other sulphides, notably those of copper
and of iron, which enables the silver sulphide to become con-
verted into sulphate. The lead sulphide is also converted into
sulphate at this low temperature (viz., about 500). The heat is
now raised still further with a view to split up the sulphate of
copper, the decomposition of which leaves oxide of copper. If,
as in this case, the bases are weak, the sulphuric anhydride
escapes mainly as such ; but when the sulphates of stronger
bases are decomposed the sulphuric anhydride is to a great
extent decomposed into a mixture of sulphurous anhydride and
oxygen. The sulphuric anhydride, resulting from the decom-
position of this copper sulphate, converts the silver into sulphate,
and maintains it as such, just as, in turn, at a lower temperature,
the copper itself had been maintained in the form of sulphate
by the sulphuric anhydride eliminated from the iron sulphide.
When only a little of the copper sulphate remains undecomposed,
the silver sulphate begins to split up (viz., at about 700) . . .
partly by the direct action of heat alone, and partly by reactions

such as those shown in the following equations :
Ag S0 4 + 4Fe 3
2 4 = 2Ag + 6Fe 2 3 + S0 2
Ag 2 SO 4 + Cu 2 = 2Ag + CuS0 4 + CuO
The charge contains lead sulphate, which cannot be com-
still
pletely decomposedat any temperature attainable in the roasting

furnace except in the presence of silica. . The elimination
. .
of arsenic and antimony gives rise to problems of much interest,

and again confronts the smelter with a case of chemical equi-
librium. For the sake of brevity it will be well for the present
to limit the consideration to the removal of antimony, which
ROASTING. 231
may be supposed to be present as sulphide. Some sulphide of

antimony is distilled off, but this is not its only mode of escape.
An attempt to remove antimony by rapid oxidation would be
attended with the danger of converting it into insoluble anti-
moniates of the metals present in the charge. In the early
it is, therefore, necessary to employ a very
stages of the roasting
low temperature, and the presence of steam is found to be useful
as a source of hydrogen, which removes sulphur as hydrogen,
sulphide, the gas being freely evolved. The reaction
Sb 2 S 3 + 3H 2 = 3H 2 S + 2Sb
between hydrogen and sulphide of antimony is, however, endo-

thermic, and could not, therefore, take place without the aid
which is afforded by external heat. The facts appear to be as
follows Sulphide of antimony, when heated, dissociates, and
:
the tension of the sulphur vapour would produce a state of

equilibrium if the sulphur thus liberated were not seized by the
hydrogen, and removed from the system. The equilibrium is
thus destroyed, and fresh sulphide is dissociated. The general
result being that the equilibrium is continually restored and
destroyed until the sulphide is decomposed. The antimony com-
bines with oxygen and escapes as volatile oxide, as does also
the arsenic, a portion of which is volatilised as sulphide.
" The main
object of the process which has been considered
is the formation of soluble' sulphate of silver." The reactions,
however, are precisely similar in an ordinary oxidising roast.
The following remarks on the decomposition of the various
minerals present in complex ores may be of use in assisting the
student to understand the reactions which proceed in the roasting
furnace :
1. Iron Pyrites, FeS

2
On heating this compound, sulphur is
.
volatilised, the reactions being probably expressed thus :
3FeS 2 = Fe 3 S 4 + S2
7FeS 2 = Fe 7 S 8 + 3S 2
The sulphur burns to S0 2 which is partly converted by the

,
heated quartz, &c.,* into SO 3 uniting with the free oxygen

,
present. The ferrous sulphate formed by this sulphuric acid is

split up by the heat and the ferrous oxide (FeO)
converted into
ferric oxide (Fe 2 3 ) which gives the ore a red colour when cold.
If the temperature of the part of the charge nexfc the fire-bridge
has been too high, or if the charge is kept too long in the
furnace, some magnetic oxide is formed, thus :
3Fe 2 3
= 2Fe 3 4 +
This an undesirable change, as the magnetic oxide
is is acted on
by chlorine far more readily than the sesquioxide.
* Plattner's
Metallurgiache Rostprozesse, Freiburg, 1856.
2. Copper Pyrites. The decomposition of the copper sulphate

formed in the furnace leaves a mixture of cuprous and cupric
oxides, both soluble in chlorine.
3. Galena, PbS. The presence of this mineral in any but
small quantities is very detrimental, as both lead sulphate and
lead silicate (formed by its decomposition in the presence of
silica) are very fusible, and, at the temperature required to split
lip copper sulphate, cause the ore to become pasty and form
lumps. Roasting must be performed very slowly and cautiously
to avoid this effect.
4. Arsenical Pyrites, FeAsS. Arseniates of iron, copper, lead,
&c., when formed are not easily decomposed, as they resist a
high temperature, and are only slowly converted into sulphates
by sulphuric acid at a red heat. It is, therefore, desirable to
avoid their formation, and with this end in view the precautions
which have been already mentioned in the extract from Prof.
Roberts- Austen's address are taken.
5. Antimonial Sulphides are still more difficult to deal with,
the antimoniates formed being less easily decomposed than ar-
seniates. Their formation is avoided in the manner already
described.
6. Blende, ZnS, forms oxide and sulphate of zinc, of which the
latter can only be split up by a very high temperature. At a
bright red heat a basic sulphate is formed which is converted
to oxide at a white heat. If blende is roasted at a high tem-
perature and with a plentiful supply of air, sulphate of zinc is
not formed to a large extent.
Elimination of Arsenic and Antimony. Mr. H. M. Howe, in
explaining how this is effected, distinguishes three horizontal
zones in the ore :* (1) the upper surface, where oxidation is only
slightly hindered by sulphurous and sulphuric acids and by the
products of combustion of the fuel ; (2) the middle layers, where
oxidation proceeds to a very limited extent (3) the lowest
;
"a
layers, where pellet of ore is simply exposed to the action of
the other pellets with which it is in contact, of volatilised
sulphur, and of sulphurous and sulphuric anhydrides generated
by the action of sulphur on previously formed metallic oxides.'*
He proceeds "The expulsion of arsenic and antimony as
sulphides is favoured in the middle and lower zones by the
presence of volatilised sulphur, mixed with sulphurous acid and
at most a very limited supply of free oxygen and sulphuric acid.
In the upper part of the middle layer, to which a small amount
of free oxygen penetrates, we have the gently oxidising conditions
favourable to the formation of arsenious acid and trioxide of
antimony. In the upper zone the stronger oxidising conditions
rather favour the formation of fixed arseniates and antimoniates,
though, even here, part of the arsenic and antimony may
*
Copper Smelting, U.S.A. Oeol. Survey, Washington, 1885.
ROASTING. 233
volatilise and escape while passing througli their intermediate

volatile condition of arsenious acid and trioxide of antimony."
On stirring the mass, these arseniates and antimoniates, being
exposed to the reducing action of volatilised sulphur and unde-
composed sulphides in the lower zones, may again be converted
into volatile oxides. Protoxide of iron, suboxide of copper, and
sulphurous acid are also efficacious in reducing arsenic acid r
higher oxides of iron and copper, and sulphuric acid being
"
formed. Thus, every individual atom of arsenic may travel
forth and back many times through the volatile condition, being
oxidised at the surface and reduced below the surface, . . .
and every time it arrives at this volatile condition an oppor-

tunity is offered it to volatilise and escape." If a small quantity of
coal or coke dust is mixed with the ore, after it has been completely
oxidised, and the air excluded, the arseniates and antimoniates
are again reduced to the lower oxides, and, if they are "carried
past the volatile state," i.e., reduced to metals, they may be
" Of course
again passed through it by an oxidising atmosphere.
the expulsion of arsenic and antimony is favoured by the presence
of a large proportion of pyrites, both because the sulphur distilled
from the pyrites tends to drag them off as sulphides, and because
the presence of the pyrites prolongs the roasting, and thus
increases the number of times which the arsenic and antimony
pass back and forth past their volatile conditions ; hence, it is
sometimes desirable to mix pyrites with impure ores to further
the expulsion of their impurities."
The Use of Salt in Roasting. Certain ores require the
addition of salt in roasting in order to chloridise material which
would otherwise absorb chlorine when the ore came to be
"
gassed," and so cause additional expense as well as incon-
venience. If silver as well as gold is to be extracted from
the ore, the addition of salt is necessary in order to form
chloride of silver in the furnace, since metallic silver is not
attacked by chlorine at the highest temperature ever employed
in the leaching vat. The silver chloride is then dissolved out
by hyposulphite of soda or some other solvent either before or
after the extraction of the gold.
Even if no silver is present, an ore must be roasted with salt
if it contains much copper (as sulphide, or as an oxidised salt),
lime, magnesia, or other substance which, after being subjected to

an oxidising roasting, is rapidly attacked by chlorine at ordinary
temperatures. The salt is usually added towards the
end of
the operation, when no sulphides and only a small percentage of
ore
sulphates are left undecomposed ; sometimes, however, the
and salt are mixed before charging in. To some sulphides only
5 pounds of salt per ton of ore are added, but others require as
much as 90 pounds per ton. The weight of salt added must be
at least six to eight times that of the silver present in the ore.
If a large amount of salt is used, it is desirable to leach the

roasted ore with water, before treating it with chlorine gas, in
order to remove the coating of soluble sulphates and chlorides
remaining on the surface of the granules of ore.
The chemical action of the salt is due to a double decomposi-
tion between it and the sulphates of the heavy metals, by which
sulphate of soda and the chlorides of the heavy metals are pro-
duced. The following general equation approximately represents
the reaction :
2NaCl + RS0 4 = RC1 2 + Na 2 S0 4

Chlorineis also set free by the action of sulphuric anhydride
on and the presence of water vapour induces the formation
salt,
of much hydrochloric acid. These gases act directly on the
several constituents in the ore, forming chlorides and oxy-
chlorides. The metallic chlorides and oxychlorides formed are
in many cases volatile (e.g., the compounds of copper, iron, lead,
arsenic, antimony, <kc.), and, in passing off, the volatile compounds
carry away with them varying proportions of gold and silver,
which, as a rule, are not recoverable in the dust chambers. The
chloride of copper is especially active in causing these losses.
Other reactions which probably take place are as follows :
1. Ferrous sulphate, acted on by salt at a red heat in presence
of air, yields hydrochloric acid and chlorine, which act on the

gold and silver, while ferric sesquioxide and sodic sulphate are
produced.
2. Ferric chloride, Fe 2 Cl 6 is also produced at the same time.
,
This is volatile, and chloridises silver with great energy at a red

heat, sesquioxide of iron being produced.
3. Cupric chloride, CuCl 2 is easily decomposed into cuprous
,
chloride, Cu 2 Cl 2 and free chlorine, or into the oxychloride,

,
Ou 2 C1 2 and free chlorine. The vapours of CuCl 2 thus give

.
,
rise to further supplies of nascent chlorine available for the

chlorination of the silver.
4. Arsenic and antimony form volatile chlorides which are
decomposed by means of oxygen and water vapour, yielding

arsenious and antimonious acids and nascent chlorine or hydro-
chloric acid.
It is thus obvious that the presence of base minerals is
advantageous in that they may cause nascent chlorine to be set
free in the presence of silver in all parts of the furnace. On the
other hand, the loss of gold is increased by any increase in the
quantities either of silver or of the base metals, since in the
former case the time of the roasting is prolonged. The best
chloridising effect is obtained in a highly oxidising atmosphere,
so that very little sulphur is required in the ore, and, if much
is present, the practice of eliminating the greater part before add-
ing the salt is not likely to be attended with any diminution in

ROASTING. 235
the percentage of silver chloride formed. Moreover, the water of

crystallisation in the salt promotes the formation of hydrochloric
acid. When salt is used in roasting, the ore should be allowed
to cool slowly in heaps after being withdrawn from the furnace.
If treated in this way, a higher percentage of the silver, &c., is
found to be chloridised than if the ore is wetted down at once,
or even spread out to cool in a thin layer. Chlorine continues
to be evolved for a long time after the withdrawal of the charge
has taken place, the heaps smelling strongly of the gas.
Losses of Gold in Roasting. Plattner proved in 185 6* that
in the oxidising roasting of ordinary auriferous pyrites, a loss of
gold can take place only when the operation is carried on so
rapidly that fine particles are carried off mechanically by the
draught. This conclusion, as far as sulphides and arsenides are
concerned, has been confirmed by Kiistel,f and by Prof. S. B.
Christy,| but the latter adds that it is extremely difficult to
prevent all mechanical loss by dusting, which is caused by even
a moderate draught. Kiistel records the loss of 20 per cent, of
the gold present during the oxidising roasting of certain
tellurides of gold and silver, and states that this is not a
mechanical loss, but is due to volatilisation. The effect of
tellurium on the volatilisation of metallic gold is shown on p. 7.
The losses of gold which are sustained when salt is added to
the furnace charge have been fully investigated by Prof. S. B.
Christy, and the following account is mainly derived from his
paper on the subject. Kiistel had previously found that a
telluride ore, on being roasted with 4 per cent, of salt, lost 8 per
cent, of its gold before the ore was red hot. Aaron found that ||
certain ores, consisting of simple pyrites, suffered great loss of

gold in roasting with salt which had been added at the com-
mencement of the operation ; only a small part of this gold was
condensed in the flue, in which was found a yellowish fluffy
precipitate, consisting largely of chlorides of copper and iron,
and containing nearly 30 ozs. of gold to the ton. He found
that the. loss was greatly reduced by diminishing the quantity
of salt, and by reserving it until the dead roasting was nearly
complete.
In the chloridising roasting of a Mexican ore, consisting mainly
of magnetite and pyrites with 3 -5 to 7 per cent, of chalcopyrite,
Mr. C. A. Stetefeldt found the losses of gold to be from 42-8 to
93 per cent, of the total gold contained. He statesf that "there
is no doubt that the volatilisation of the gold takes place with
*
Metallurgische Rostprozesse, Freiburg, p. 128.
t Roasting of Gold and Silver Ores, 1880, p. 56.
J Tram. Am. Inst. Mng. Eng., 1888.
Loc. cit.
II
Leaching of Gold and Silver Ores, 1881, p. 121.
IF Trans. Am. Inst. Mng. Eng. y vol. xiv., p. 339.
that of the copper chlorides. The loss increased with the

quantity of these chlorides formed and volatilised." He further
shows, however, that the presence of copper chloride is not the
only possible cause of loss, since an ore consisting of hard white
quartz, intimately mixed with about 7 per cent, of calcite and a
little pyrites, lost 70 to 80 per cent, of its silver, and 68 to 85 per
cent, of its gold, when roasted with 5 per cent, of salt. When
subjected to an oxidising roast, no loss of gold took place. The
reason for the extraordinary behaviour of this ore was not
discovered.
Prof. Christy found that, in the ores on which he experimented
on a small scale in a muffle furnace, a greater loss was sustained
by adding the salt near the end of the roasting operation, than
by mixing the same weight of salt w.ith the ore at the start. He
explained that this is due to the fact that the amount of gold
volatilised varies with the amount of chlorine which comes in
contact with it. When the salt is added at the start, the
chlorine is at first removed by the sulphur as fast as k is formed,
escaping as chloride of sulphur, and thus the gold is protected
from attack. When the salt is added after a long oxidising
roast, the chlorine is rapidly generated (the ore being* red hot
and containing large quantities of sulphates), and the gold
is no longer protected from attack by the sulphur. The loss of
gold is also in all cases increased by working at a higher tern-
perature, owing to the larger amount of chlorine generated, and
to the increase in the volatility of the gold. It is apparent from
the results given on p. 21 that the temperature used in chlorid-
ising roasting must be very carefully regulated, the loss of gold
being increased far more by high temperature than by a length-
ening of the time in the furnace. Moreover, the salt must be
reduced to the least possible quantity.
The advantage found to be gained in practice by adding salt
near the end of the operation is due to the fact that, in the
continuous roasting of ores in long-bedded furnaces, the gases
given oft' from the finishing floor pass over a great length of
comparatively cold, unsalted, and unoxidised ore before reaching
the tlue. The quantity of gold chloride mixed with the chlorine
which is evolved from the red-hot ore as soon as the salt is added
is no doubt large, but the S0 2 from the colder ore, and the steam
from the fuel, "offer excellent means for the reduction of the
chloride of gold right within the furnace, while the most efficient
means probably is the pyrites themselves," which have been
proved to be readily capable of condensing gold on their surface.

If all the salt is added at the start, there is a continued vola-
tilisation of chloride of gold throughout the furnace, and a less
favourable opportunity for it to condense. The difference
between the results in the muffle and in the reverberatory
furnace is thus explained.
MECHANICAL FURNACES. 237
At Nevada City, at the Merrifield Mine, and in other works

in the neighbourhood, the old-fashioned long furnace, with a
single step separating the finishing hearth from the rest of the
furnace, was still used in 1888.* These furnaces are from
55 to 65 feet long, holding from 6 to 9 tons, and producing
about 3 tons of roasted ore per day, so that the ore remains
in the furnace from two to three days. The custom there
was to give the ore a long oxidising roast at a low red
heat, ending at a low cherry-red heat, arid then, when the
ore reached the finishing floor, the temperature was slightly
lowered, and the salt added. The salt was stirred thoroughly
into the ore, and as soon as it was " dissolved by the roasted
"
ore i.e., in about half an hour the charge was drawn into
the cooling pit. This lowering of the temperature is evidently
of great importance in reducing the loss, while the dura-
tion of the roasting is regarded as less material, so long as
no salt is present. These mills are on custom work, charging
$15 to $17- per ton of ore for treatment, and guaranteeing a
yield of 90 per cent, of the gold and 60 per cent, of the silver.
Their method of roasting seems to be considered in California as
thjjt best suited to concentrates containing a high percentage of
sulphur, but their loss in roasting has not been ascertained.
The best method of roasting any particular ore, however, cannot
be determined by any general rule, and exhaustive experiments
must be made in every case before a definite course of procedure
is finallyadopted.
At one of the Californian chlorination mills it was found by
experiment in 1882 that nearly 50 per cent, of the gold and
28 per cent, of the silver was being lost by volatilisation. In
this case the pyrites was roasted on two hearths for thirty-six
hours, 1 per cent, of salt being added four hours before the
charge was drawn. The reason for the great loss was thought
by Professor Christy to be the high temperature of roasting,
particularly on the charging-in floor.
The variation of the loss in different ores which are treated
precisely alike is doubtless due partly to the presence or absence
of metals forming volatile chlorides which carry oft* the gold, and
partly to the physical condition of the latter, the volatilisation
being greater if it is in a state of minute subdivision.
MECHANICAL FURNACES.
The furnaces which have been designed with the object of

saving the labour necessary to work the reverberatory furnaces
may be divided into four classes, viz. :
*
Trans. Am. Inst. Mng. Eng., vol. xiv., p. 340.
1. Stationary hearth furnaces, supplied with iron hoes moved

by machinery by which the ore is rabbled. The O'Hara, Spence,
and Pearce Turret furnaces are examples of this class.
2. Rotating- bed furnaces, in which the hoes or stirrers are
stationary, while the bed supporting the ore revolves, so that

the latter is stirred by the hoes. An example of this class used
to roast gold ores as a preliminary to chlorination is afforded by
the furnace used at the Bunker Hill Mine, California.
3. Rotating cylindrical furnaces, which consist of brick-lined
iron cylinders capable of being rotated, so that the ore is
tumbled over and over by their motion while it is being roasted.
Examples are the Bruckner, the White-Howell, and the Hof-
mann furnaces.
4. Shaft furnaces, in which the powdered ore falls by gravity,
in a shower, through an ascending column of hot air, the oxida-
tion being effected in the course of the fall. The Stetefeldt
furnace, which is the only one based on this principle, is not
used for dead roasting, as it is not adapted for the purpose. It
is used for the chloridising roasting of silver ores, and will not
be described in this volume.
In mechanical furnaces the consumption of fuel is often much
greater than that in the long-bedded reverberatory furnace,
where it is usually from 10 to 20 per cent, of the weight of the
ore if flaming coal is used.
1. Furnaces with Mechanical Stirrers. O'Hara Furnace.
This is the oldest mechanical furnace, and it bears a great
resemblance to the old-fashioned reverberatory furnace. It has
two superposed hearths, in each of which the arch is very low,
so as to confine the heat close to the ore. An endless chain, set
in motion by suitable machinery, passes through the furnace,
resting on the upper hearth, and returns along the lower hearth.
Attached to the chain at proper intervals are iron frames of
a triangular shape on these frames are a number of ploughs
:
or hoes set at an angle, so that one set of hoes turns the ore to
the centre, and the next set turns it in an opposite direction
towards the walls. The ploughs thus stir the ore thoroughly,
and at the same time move it gradually towards the fire. The
ore falls from the upper to the lower hearth by gravity,. and
similarly falls from the lower hearth into a pit when it arrives
at the hottest place in the furnace. The ore is from five, to ten
hours in the furnace, according to the amount of sulphur contained
in it. In the modern form the hearths are each 8 feet wide and
90 feet long, and the capacity is about 35 tons per day, at an
expenditure of about 2J H.P. The ore is not roasted dead,
however, in this case, about 6 per cent, of sulphur remaining
in it. No less than twenty-three of these furnaces are now in
operation in the United States, although none are used in
chlorination mills. *
*
Eng. and Mng. Journ., May 20, 1893, p. 463.
Spence Furnace. This is perhaps the best shelf furnace yet

devised, and has been used successfully in the roasting of copper
sulphides in the United States. It consists of a series of four
or five superposed hearths, communicating with one another
by means of vertical passages at alternate ends of the furnace.
The rabbling is done by rakes having a reciprocating motion
longitudinally in the furnace and armed with teeth of triangular
section having the apices pointing in the opposite direction to
that in which the ore travels. When the rakes move in the
direction in which these apices point, the ore is only stirred, but
when the return movement is made, the flat sides of the teeth
push some of the ore along the floor of the furnace, and a part
falls through the vertical passages on to the lower hearths. The
ore is charged in through a hopper and discharged roasted into-
a pit. The number of floors may be varied according to the
character of the ore to be treated. The motion of the rakes
should not be continuous, as in that case wearing of the teeth
becomes very rapid. It is better to set them in motion (by
racks and pinions) for a few minutes, and then to withdraw
them for a while in order to allow them to cool. It is claimed
for this furnace that if once made hot, the combustion of the
sulphur in pyritic ores supplies the place of other fuel, so that
the roasting is perfectly performed with no further cost than
that of power for the rakes. It is of course obvious that
without an extra fire, the lowest hearth, which is not heated
from below and which is the place where the air is admitted,
would tend to become the coldest, and would then be in no way
adapted for the decomposition of the sulphates formed on the
upper hearths, a process which involves endothermic reactions,
and, therefore, requires external heat. The Spence furnace
without a fire may, however, be valuable for the production of
sulphates of iron or copper.
At the Treadwell Mine, Alaska, Bruckner cylinders were
formerly used,* but discarded on account of the amount of dust
made and carried into the flue, and the large amount of fuel
consumed, and Spence furnaces introduced. Each furnace had
four hearths, the lowest being strongly heated to decompose the
sulphates. The ore remained in the furnace for sixteen hours,,
and 3 per cent, of salt was added on the hearth next above the
lowest one. The rakes of the upper hearths lasted a long time,,
but those of the finishing hearth were worn out in three months,
and were often broken sooner. These furnaces were not found
to be of any use until converted from muffle into reverberatory
furnaces so that the products of combustion of the fire passed
directly over the ore to be roasted. Six of the double Spence
furnaces were built at a great expense, and the cost of roasting
in them was found to be less than half that in the Briickner, but
*
Eng. and Mng. Journ., April 11, 1891.
their capacity was small and the amount of fuel required was
found to be very great. An ordinary reverberatory furnace
was, therefore, built, and the results obtained were so satisfactory
that three others were added, and the Spence furnaces thrown
out of work. They are apparently not now used at any chlorin-
ation mill, but might possibly be adapted to some ores.
One of the chief causes of difficulty and expense in working
furnaces with mechanical stirring apparatus is that the iron hoes
gradually become heated, and they are then rapidly corroded by
the sulphur in the ore. In hand stirring, the rabbling tool is
withdrawn as soon as it is hot, and allowed to cool, while
another is substituted for it meanwhile, thus prolonging its life.
To imitate this action, the Spence hoes are sometimes arranged
to work for a while and then to be withdrawn completely to
cool. In the O'Hara furnace, also, it is better to have only one
hearth, the hoes passing completely outside the furnace on the
return journey. In spite of such arrangements, however, the
trouble and expense caused by the wearing of the hoes are very
great.
Pearce Turret Furnace. This consists of an ordinary rever-
beratory hearth built in an annular form. In the centre of the
circular space surrounded by the hearth is a vertical iron column
carrying four hollow horizontal arms projecting through a slot
into the reverberatory hearth which they cross transversely.
The column revolves and the arms carry rabble blades which
traverse the hearth, stirring the ore and moving it round the
" Air is forced
circle by degrees. through the hollow arms and
is discharged against the rabble blades, performing the double
duty of cooling the iron work and of furnishing heated air for
the oxidation of the ore." The ore is discharged automatically
after passing once round the furnace. Two or more fireplaces
are used. These furnaces are very economical, and are now pre-
ferred to any others in Western America for roasting.
2. Rotating Bed Furnaces. Rotary Pan Furnace. This is
used at the Bunker Hill Mill, California, to desulphurise concen-
trates containing much sulphur and small quantities of arsenic,
antimony, and lead. The stationary hearth is 7 feet wide and
18 feet long, and has two working doors. At the end of the
stationary hearth is a drop of 6 inches on to a horizontal revolv-
ing hearth, made of iron lined with fire-brick, 12 feet in diameter,
with a discharge hole in the centre. This is next the fire-place.
The hearth revolves by means of gear-wheels placed beneath it,
at the rate of one turn per minute. The charge remains on it
for eight hours, and is then discharged through the central
aperture. The capacity of the furnace is 2 tons per day, the
fuel required being \\ cords of wood. In Fig. 46, a similar
but larger furnace is shown.
3. Revolving Cylindrical Furnaces. These furnaces have

come into more general use than other mechanical furnaces, but
are employed at very few gold chlorination mills, since their
capacity is too large for the modest requirements of most of these
establishments, many of which treat only a few tons per week.
Revolving cylinders treat from 10 to 50 tons of ore per day.
The Bruckner Cylinder. This furnace was first introduced in
Colorado in 1867, and is now, after alterations and improve-
ments, extensively used in the United States, chiefly in the
chloridising roasting of silver ores, although it is also suitable
for dead roasting. In its latest form it consists (Fig. 47) of a
cylinder of boiler plate iron, lined with fire bricks. The ends
are partly closed, leaving central apertures 2 feet in diameter.
There are also four receiving and discharging openings (a, a),
closed by hinged doors. The discharge is effected into a pit, or
into hot-ore cars placed underneath. The cylinder revolves on
two chilled-iron friction rings (6, 6), resting upon four chilled-
iron carrying rollers (c, c). The rotation is caused \>y friction
between the rings and the rollers, which are driven by a belt
and pulley. The older method of turning the cylinder by means
of a gear-wheel, connecting with teeth set on the cylinder, has
been abandoned.
The lining is one brick (2^ inch) thick in the middle of the
cylinder, but additional layers are added near the ends until the
circle is contracted down to the size of the openings in the ends,
which are also lined ; each layer i'alls short of the preceding one
by about 2 inches, so that the lined ends have a conical form.
The mortar used consists of one-third fire-clay and two-thirds
crushed fire-brick. In the old form six iron pipes passed through
the cylinder in a plane at an angle of 15 to the axis, and per-
forated plates uniting them formed a diaphragm which assisted
in stirring the ore. The circular flange surrounding the opening
at one end of the cylinder connects loosely with the fire-box, d't
the other end connects with an opening leading to dust chambers
and the stack. The dust chambers (not shown in the figure) are
large, and must be carefully attended to, as the amount
carried
over is usually 10 per cent, of the ore, and may rise to 25 per
cent, when the mineral is light. The heavier particles collected
in the first chamber are charged into the furnace again ; the
lighter dust is unsuitable for this purpose, and must
be subjected
to special treatment. In the improved Bruckner the diaphragm
is discarded, as it is rapidly corroded by the sulphur, and
increases the quantity of material carried into the dust chambers.
The furnaces are now usually made 7 feet in diameter and 18
feet long, a full charge being from 6 to 8 tons. This large size
has given greater satisfaction than that formerly employed viz.,
6 by 12 feet the full charge of which was from 3 to 4 tons of
ore. As a charge can be kept in the furnace as long as is
necessary, the Bruckner is particularly useful when very base

ores are to be roasted, or when the ores vary greatly in com-
position, so that the duration of roasting is not constant. The
mixing of the ore is rendered more perfect by the conical shape
of the ends, which causes the ore to be thrown backwards and
forwards, changing its position frequently and exposing new
surfaces to the action of the fire.
Bruckner's cylinder has been found suitable for the dead oxi-
dising roasting of pyritic ores, a little charcoal being sometimes
added towards the end of the operation to reduce the sulphate of
copper formed. It has the disadvantage of being too hot at
first, and not hot enough at the finish if the fire is kept uniform,
and consequently pyritic ore tends to form into balls. These
may contain sulphates, which are not then easy to decompose
except with great waste of time and fuel. The consumption of
fuel is also rendered greater by the shortness of the furnace, as
a large amount of waste heat passes out into the dust chamber,
and is not utilised, a state of things exactly similar to that
which occurs in the reverberatory furnace with only one floor.
An important improvement in the Bruckner furnaces in use
at the Portland mill, Deadwood, Dakota, has lately been made
by the Manager, Mr. Hickock. Between the fire-box, which is
mounted on wheels, and the revolving cylinder is placed a short
iron cylinder, 18 inches long, rigidly connected with the fire-
bridge and closely fitting the throat of the revolving cylinder.
In the lower side of the small cylinder is an aperture, from 8
inches to 10 inches square, closed by a sliding door. This is
useful in regulating the draught, without cooling down the fire.
By its use a sheet of cold air can be made to pass immediately
above the ore and below the products of combustion of the fire.
It is stated that a considerable saving of fuel has been effected
since the introduction of this simple device, and that a more
perfect roast can be obtained in one-third less time than was
formerly possible.
The Ilofmann Furnace. This furnace was designed by H. O.
Hofmann an improvement on the Bruckner cylinder, which
as
has the disadvantage of being much hotter at one end than at
the other. The result of this is that the ore near the fire is
exposed to a higher temperature than that near the flue, and,
consequently, it is finished much sooner, so that fuel is wasted
while finishing the ore in the colder part of the furnace, and, in the
case of antimony ores, if the temperature is high enough to roast
the ore next the flue, the portion close to the fire becomes caked
and suffers a considerable loss of silver by volatilisation. The
Hofmann furnace (Fig. 48) has a fireplace and flue at each end.
The flues are between the fireplaces and the cylinder, and open
downwards into dust chambers, C, built beneath, which are con-
nected with the main stack. The arrangements at each end are
the same, and are such that, by means of dampers, the current of
the air and gases can be made to pass through the furnace in
either direction. The fire is lighted first at one end, A, and
the dampers arranged so that the draught passes through the
revolving cylinder and down the flue, B, at the other end. After
a few hours the fire is lighted at the other end, in the fireplace,
D, and the position of the dampers is reversed. The alternate
heating from the two ends is regularly performed until the
charge is completely roasted. In this way more uniform heating
is obtained, both halves of the charge being raised to the
required
temperature without any portion being overheated. It is stated
that the formation of balls is diminished by this system, and,
in particular, the furnace is found to answer well when treating
ores which require either a very low roasting temperature, or a
very high one. Thus, for example, antimonious ores, which cake
readily, are said to be successfully treated by the use of moderate
fires ; and ores containing very little sulphur, and so requiring a
higher temperature, can readily be made as hot as is necessary

by the alternate use of targe fires, without using much fuel.
When one fire* is in operation the other is allowed to go down.
By closing one of the large dampers near the main flue, and
opening the damper of the corresponding descending flue, and
the " plug door" in connection with it, a current of fresh air can
be introduced into the furnace beneath the flame from the fire,
thus greatly increasing the rate of oxidation of the ore. This
arrangement is especially useful when ores carrying a high
percentage of sulphur are being treated, their tendency to form
balls being in this way greatly diminished.
It is clear that the use of much larger cylinders than those of
the Bruckner furnace is rendered possible by the presence of a
fire at each end. Nevertheless, the evil of heating the ore
unequally is only palliated; the ore in the middle of the furnace
is never heated so strongly as that at the two ends, whilst in the
case of continuous-discharge inclined cylinders, every particle of
ore is heated similarly. Moreover, the use of two fires, which
are alternately checked and urged, must cause a considerable
waste of fuel. The principle is obviously inferior to that of the
White-Howell furnace.
One incidental advantage of this furnace over all other
revolving cylinders lies in the absence of loss from the carrying
away of dust at the moment of charging-in. In other cylinders
the ore falls in a shower into the furnace through a strong
draught, by which much of the dust is carried off into the flue.
When charging the Hofmann furnace, however, the dampers in
both the descending flues are left open, so that the draught
passes through both fireplaces direct to the dust chambers,
without entering the furnace at all.
The White and White-Howell Roasting Furnaces. The White
furnace consists of a long cast-iron revolving cylinder, lined with

fire-brick, and inclined towards the fire end. The cylinder
is bound by four cast-iron rings which rest on friction wheels,
serving as supports, and it is revolved by other friction wheels

driven by a shaft and pulleys. Crushed ore is fed into the
furnace continuously at the upper end, passes through it by
gravity, and is continuously and automatically discharged by
falling into a pit through an opening in the floor close to the
fire. The time occupied by the ore in passing through the
furnace depends on the angle of inclination of the cylinder
which can be changed, so that the time of roasting can be
shortened or lengthened according to the nature of the ore.
Ores containing a high percentage of sulphur require to be
subjected to heat for a longer time than those with little
sulphur, and the angle of inclination is reduced in such cases.
The average inclination is about one in twenty. The cylinder
is lined with fire-brick throughout, and projecting bricks raise
portions of the ore and drop it through the flames, thus assisting
the oxidation.
The advantages claimed for this furnace are that it is con-
tinuous in its operation, discharging its product regularly into
the pit at the lower end, and this roasted ore can be allowed to
accumulate and be withdrawn as required. The ore is submitted
to a gradually increasing temperature, the most favourable con-
ditions for dead roasting being thus obtained. The usual size
of this furnace is about 4J feet in diameter and 27 feet long,
the capacity being usually stated at from 20 to 30 tons per day.
When employed for the dead roasting of pyritic ores its capacity
is less, and in some cases the passage through a 27-foot cylinder
is insufficient to eliminate the whole of the sulphur. In such
cases the ore is made to traverse successively two similar
cylinders, the combined length of which is sometimes over
60
feet. Great care should be exercised in keeping the ore supplied
to the furnaces as uniform as possible, so that when once the
rate of feed and the proper angle of inclination for the ore have
been determined, no further alterations are needed in order to
continue to give a perfectly roasted product.
In the White-Howell furnace (Fig. 49) only the enlarged part
next the fire-box is lined with fire-brick, the remainder being
left unlined. Cast-iron spirally arranged shelves assist in raising
and showering the pulp through the flames. To both White
and White-Howell furnaces an auxiliary fire is often added for
roasting the dust which escapes from the main furnace.
The
dust, when it has been completely roasted, is shovelled out and
mixed with the main bulk of the ore by hand. This arrangement
is decidedly inferior to that suggested by John E. Rothwell,*
who uses a hopper-shaped dust chamber with its bottom con-
* Mineral
Industry for 1892, p. 234.
sisting of an inclined
cast-iron plate project-
ing about 8 inches into
the upper end of the
cylinder. The dust car-
ried out of the cylinder
settles in this chamber,
and, as it accumulates,
slides down the sides
and mixes with the fresh
ore. The ore is thus
kept more uniform than
if re-mixed by hand, and
some saving in labour
is also effected. Roth-
well uses a cylinder 36
feet long and 5 feet in
diameter, with an in-
clination of 14 inches
only, rotating once per
minute. The lining is
of fire-brick, six inches
thick.
Use of Producer
Gas in Roasting. The
use of producer gas in
roasting may be men-
tioned, as it bids fair to
completely replace solid
fuel for the purpose at
some future time, great
saving of expense being
thus effected. It was
introduced at the Hoi-
den Mill, Aspen, Color-
ado, in 1891, and has
now completely dis-
placed other fuel there
for both drying and
roasting.* The gas
plant consists of two
"Taylor" revolving-bot-
tom gas producers, one
6 feet and the other 7
feet in diameter. The
* W. S. Morse in Trans.
Am. Inst. Mng. Eng., Mon-
treal meeting, 1893.
CHLORINATION 1 THE VAT PROCESS. 249
of one-third " "

coal used is a mixture Sunshine nuts and two-
"
thirds " Newcastle nuts and pea-coal, the analyses of which are
as follows :
1
surface. The thickness of the filter bed (which is not shown in

the figure) is usually from 6 to 12 inches. It is supported on
boards (A, Fig. 50), 1 inch thick, in which numerous J-inch
auger holes are drilled ; these boards rest on wooden strips (not
shown in the figure), 3 inches wide and 1 inch thick, which do not
reach the edge of the vat, and so keep a clear space 1 inch deep,
just above the true bottom of the vat, in which the solution can
accumulate. The solution is drawn off" by a leaden pipe fitted
with a stopcock, preferably of stoneware; the pipe should be
level with the bottom of the vat, which may with advantage be
made with a slight fall towards the outlet to prevent any liquid
*
being left in it. Deetken states that fine sea-shells (consisting
of carbonate of lime) have been used instead of quartz pebbles

for the filter bed without any prejudicial result. Talcose rocks,
and particularly silicates of alumina, must not be used on
account of their power of absorbing the chlorine. For the same
reason sulphides, magnetic iron oxide, metallic iron, fragments
of wood or other organic matter, or, briefly, any substances
capable of being acted on by chlorine or of reducing the chloride
of gold must be carefully excluded from the filter bed.
The surface of the filter bed may be covered with boards, not
fitted closely together, but made into a framework by cross-pieces
and pierced with many auger holes. This cover is useful when
* Mineral Resources of the, Rocky Mountains, 1873, p. 342.
CHLORINATION ! THE VAT PROCESS. 251
the tailings are being cleared out, otherwise, in shovelling away

the ore, the surface of the tilter bed is partly removed also.
Messrs. MacArthur & Forrest's devices to avoid this are given
on p. 310. Filter cloths of canvas, burlap, or cocoa-nut fibre
matting are also frequently used above the filter bed, stretched
tightly over a framework of wood which accurately fits the inside
of the vat. The space between the canvas and the wall of the
vat is packed with hemp or other material closely tamped down.
Filter-cloths of every material, except asbestos, are soon rotted
and demoralised by the action of the chlorine, and their use is
frequently dispensed with. Wool lasts longer than cotton.
Charging-in the Ore. When the vats are ready to be
charged, a layer of dry ore is spread over the false bottom, and
time given for the water from the filter bed to be drawn up into
this layer by capillary attraction. If attention is not paid to
this point, the lowest layer of ore becomes too wet from the
combined effect of the water added to it before charging-in and
that absorbed from the false bottom. The result is that the
passage of the chlorine through the mass is resisted, and there
is a great increase in the consumption of the gas. Deetken
states that the whole of the usual charge of gas may be thus
consumed, not rising more than a few inches above the bottom.
The greater part of the charge is damped by sprinkling with
water and thorough mixing. The amount of the water added
varies with the nature of the ore, but the usual amount is from
6 to 12 per cent, for roasted ores. If it is made too wet, dry
ore to the required amount is mixed with it. A
good rough
method of ascertaining when it is of the proper degree of damp-
ness is to compress some in the hand ; balls of ore should
be readily formed in this way, but should be just dry enough
to crumble up again. The reason for the addition of water
is that perfectly dry chlorine has scarcely any action on metal-
lic gold at ordinary temperatures, and up to a certain point
an increase in the amount of water present raises the rate of
solubility of the gold. The limit of the amount of water that
can be added is, however, determined by physical conditions, as
the mass must be of loose porous texture in order to permit the
gas to readily permeate through every portion of it.
In order to
promote this porous texture and uniform dampness, the ore
is
shovelled upon, and made to pass through, a sieve of four holes to

the linear inch. This sieve may be conveniently made to slide
on rollers on iron rails placed above the vat, and the ore, shaken
left
through it, falls into the vat in a light shower. Although
undisturbed as far as possible, the charge must be levelled off
with a rake occasionally.
When the vat is filled to within 6 inches of the top, the
surface of the ore is made concave or saucer-shaped, higher at
the sides than in the centre. The cover, usually of wood, is
then lowered on to the vat by means of a chain and pulley, and

the rim luted with a mixture of wet clay and sand, or, more
usually in former times, with dough. These joints are kept
moist during the "gassing" process by wet rags. The gas is
introduced through a lead pipe, which is shown on the left-hand
side of Fig. 50, passing into the vat below the false bottom.
A small hole is left in the cover through which the displaced
air may escape, and the issuing gases are tested from time to
time by means of a rag tied to a stick and moistened with dilute
ammonia. As soon as chlorine is found to be coming off freely,
the hole is plugged, but the current of gas is not stopped until
after the lapse of one or two hours more, when the charge is
supposed to be saturated with the gas, the total time required
for the impregnation being usually from five to eight hours.
Generation of the Chlorine. The chlorine is generated in
air-tight vessels of lead litted with a stirring apparatus passing
through the lid and worked from the outside. The gas neces-
sary for a 3-ton charge of roasted concentrates may be generated
in a leaden vessel of 20 inches in diameter and 12 inches deep.
The joints of this generator and of all pipes traversed by the gas
or by liquids carrying it in solution must be made by melting
the lead, no solder being used except pure lead. The " burning
together" of the joints is usually effected by an air-hydrogen or
coal gas blowpipe jet. The cover of the chlorine generator is
made gas-tight by a water joint, 2 inches deep, as are also the
apertures in the lid for the passage of the revolving stirrer
(which is usually made of hard wood) and for the delivery
tube. Heat is applied by placing the generator on a sand bath
standing on a perforated arch over a fire-place. The sand bath
is often replaced with advantage by a water bath, as the heat
required is not more than 90 F., and sudden heating causes

inconveniently tumultuous generation of gas. The charge for
3 tons of ore consists of 20 to 24 pounds of rock salt, 15 to
20 pounds of manganese dioxide, containing 70 per cent, of
available material, and 35 pounds of oil of vitriol of 66 B.,
diluted with half its weight of water. The cover is usually
removed to introduce the solids and the water, but the acid is
added, half a gallon at a time, through a siphon. At Hey wood's
Works, California, the acid is contained in a lead vessel furnished
with a glass stopcock, from which a small continuous stream is
allowed to fall into the siphon.
The outlet tube is of lead, but connections in pipes are often
made by short pieces of indiarubber-tubing, well greased on the
inside. These resist the action of chlorine fairly well. The gas
is passed through the wash- bottle shown in Fig. 51. It is
usually a large glass bottle or carboy with its bottom removed,
supported in a lead-lined box filled with water. The gas is made
to pass through about half an inch of water. The use of the
CHLORINATION : THE VAT PROCESS. 253
wash bottle is partly to free the chlorine from hydrochloric acid

or other impurities with which it is contaminated, but mainly
to give an indication of the rate of flow of the gas it is desirable
;
that this should be as uniform as possible, as otherwise leakage

is more difficult to prevent. As soon as the current of gas falls
off, fresh acid is added and the vessel stirred. The wash-bottle,
as usually constructed,
would be quite inade-
quate to free the gas
completely from hydro-
even
chloric acid, whilst,
if itwere approximately
eliminated, some more
would be speedily formed
by the decomposition of
water by the chlorine.
It is, therefore, fortunate
that this elimination is
Fig. 51. not absolutely necessary,
Scale, 1 in. = 1 ft. in spite of the customary
declaration to the contrary which generally appears in the descrip-
tions of the process. Thus it is frequently stated that hydro-
chloric acid will act on any sulphides left undecomposed in the
ore, generating sulphuretted hydrogen which would precipitate
the gold already dissolved in the impregnation tank. This
statement would not need any criticism, if it were not for the
fact that it has often been made, but has apparently never been
contradicted. No doubt matters might be arranged for the
above reactions to take place if sufficient care were taken, but
in practice they are not to be feared for the following reasons :
(1) If undecomposed sulphides were present

in the roasted
ore, they would be attacked with much greater violence by
chlorine than by hydrochloric acid, and before any gold could
be dissolved, the whole of the sulphides present would be con-
verted into sulphates, according to reactions, the final effect of
which is shown in the following equation :
R2 S 4C1 4H 2 R 2 S0 4 8HC1.
In some cases, no doubt, chlorides would be formed. (2) If the

sulphides were not oxidised by the chlorine, they would
be
almost equally efficacious with sulphuretted hydrogen in preci-
acid
pitating the gold, so that even in this case hydrochloric
would do no harm. (3) If an excess of chlorine is present, sul-
phuretted hydrogen could scarcely be said to be formed
at all
under any circumstances, as, if it were formed, it would be
instantly decomposed by the chlorine. (4) If gold
were precipi-
tated by sulphuretted hydrogen in the impregnation vat, it would
be in such a finely divided state that it would be re-dissolved in
254. THE METALLURGY OP GOLD.
chlorine in a very short time, provided the gas were in excess.

The only disadvantage due to the presence of much hydrochloric
acid in the gas lies in the fact that certain metallic oxides (oxides
of iron, copper, &c.) are much more readily soluble in the acid
than in chlorine, chlorides and water being formed, and the re-
sulting solution will be contaminated with these chlorides, so
that special precautions are necessitated to prevent the bullion
from becoming base.
Impregnation of the Ore. The ore is allowed to remain
impregnated with the gas for from twenty-four to forty-eight
hours, the continued presence of a strong excess of gas being
ascertained at intervals by removing the plug from the cover
and applying the ammonia test. When, as is usually the case,
there is a large excess of gas when the impregnation is at an
end, it may be disposed of in one of several ways. It may be
dissolved by adding water before raising the cover (the usual
method of procedure), or it may be withdrawn by aspiration and
discharged outside the building, or stored in a gasometer for use
in a subsequent charge. If the cover is raised before getting rid
of part of the excess of the gas, the atmosphere of the mill is
rendered unbearable for several minutes.
The time of impregnation varies according to the size of the
particles of gold, the fineness of the metal, and
the tempera-
ture employed. Chlorine has a very slow action on pure gold,
the rate increasing gradually with the temperature up to 100.
In order to obtain some data for the rate of solution of gold by
chlorine at different temperatures below 100, the following
experiments were conducted by the author in the laboratory of
the Royal Mint. Comparisons were made at the same time
between chlorine, bromine, and cyanide of potassium, in order
to determine their relative efficiency. The gold used consisted
"
of " cornets weighing about a half gramme each. These cornets
offered a large surface to attack, being porous in texture, and con-
sisted of gold 9U9-3 parts, and silver 0-7 part. They had all
been prepared together by cupellation, parting, and annealing,
so that their physical state must have been very similar. The
results are given in the table on the next page.
The results obtained simultaneously by the use of potassic
cyanide are given at p. 336. The amount of liquid used was in all
cases 30 c.c. These results tend to show that both chlorine and
bromine dissolve gold more rapidly at 50 to 60 C. than at ordi-
nary temperatures, that bromine is more rapid in its action than
chlorine, and that both are considerably more rapid than potas-
sium cyanide, particularly at the higher temperatures employed.
In all cases an excess of the reagent was present at the end of
the operation, but the strength of the solutions obviously fell off
during the experiments. The results of these experiments point
to the desirability of further investigations on the subject.
Solvent.
into chlorides and sulphuric acid is set free, the reactions being
influenced by the mass of the reagents present. Protosulphates
or any other protosalts present are converted almost instantan-
eously to persalts by the chlorine, as follows :
6FeS0 4 + 3C1 2 = 2Fe 2 (S0 4 )3 + Fe 2 Cl 6
It is obvious from these reactions that great waste of chlorine

in the impregnation vat is caused by imperfect roasting, 1 per
cent, of unoxidised sulphur present in pyrites converting 8 9 per
-
cent, of chlorine (or about 200 Ibs. per ton of ore) into hydro-
chloric acid. This simple calculation is sufficient to show the
impolicy of neglecting to roast the ore dead and then trying to
retrieve the error by increasing the allowance of chlorine.
Moreover, it demonstrates the uselessness of eliminating the
hydrochloric acid from the chlorine before mixing it with the
ore, and expecting in that way to prevent the ill effects pro-
duced by sulphides. The fact that many sulphides are almost
instantly oxidised by even very dilute solutions of chlorine has
been proved by a series of laboratory experiments by the author.
These experiments would have been quite unnecessary if it were
not that some chemists engaged in chlorination still appear to
doubt the rapidity of the reaction. The oxidation of protosalts
is almost as rapid, although the percentage waste of chlorine is
considerably less ; thus 1 per cent, of sulphur present in the ore

as ferrous sulphate will convert 1*1 per cent, of chlorine (or 24'6
Ibs. per ton of ore) into hydrochloric acid.
Sulphate of copper (CuS0 4 ) does not appear to be acted on by

chlorine, but, nevertheless, whenever it is present in a roasted
ore, chlorination seems to be rendered impracticable. This is
possibly due to the fact that some sulphate of iron accompanies
it. Whenever sulphates of these metals are left in the roasted
ore by accident or design it is necessary to remove them by a
preliminary leaching with water before the chlorine is introduced.
Of course if the ore is chlorinated in tubs by gas, it must be
partially dried and sieved back into the tub before impregnation
can be attempted.
Organic matter is also oxidised by chlorine, although much
more slowly. At ordinary temperatures, pitch and tar are
almost unaffected, and the fibres of matting, canvas, <kc., are
acted on very gradually. Pieces of decaying wood or dried
leaves must not be introduced with the water into the leaching
vat, and if surface water is used it should always be carefully
strained before being run in. A
rough analysis of the water
employed will often be serviceable, as it is frequently strongly
alkaline in dry countries, and may be softened with advantage.
The absorption of chlorine by metallic oxides is the most
frequent cause of waste, and, in the vat process, there are
usually no efforts made to prevent this. Well roasted sesqui-
oxide of iron (Fe O 3 ) is scarcely attacked by chlorine, especially

if the temperature attained in the furnace has been high. If
any magnetic oxide (Fe 3 O 4 ), however, has been formed from
over heating, or has been originally present in the ore, the
absorption of chlorine is considerably greater, ferric chloride
being formed and dissolved. Protoxide of iron is instantly
converted into a mixture of chloride and sesquioxide of iron.
Hydrochloric acid acts more rapidly than chlorine on all these
oxides, but is nevertheless very slow in dissolving the well
roasted sesquioxide. Metallic iron, which is sometimes acci-
dentally introduced, is dissolved at once by both chlorine and
HC1. The oxides of copper and zinc are quickly dissolved by
chlorine, and still more readily by HC1. Lime and magnesia
also readily absorb chlorine, forming hypochlorites, chlorates
and chlorides, but hypochlorites are decomposed by any acid
which may be present.
If any appreciable quantity of oxides capable of absorbing
chlorine are present, it is cheaper to dissolve them by adding
dilute sulphuric acid to the ore, and then, if possible, to leach out
the soluble sulphates formed, before subjecting the ore to the
action of the gas.
Amount of Chlorine required. The amount of chlorine
required varies greatly, both with the nature of the ore and
the manner in which it is roasted. In order to roast pyrites
dead, a long time in the furnace terminating at a high tempera-
ture is necessary, and the addition of salt may be desirable in
order to chloridise oxides which would otherwise absorb the
more expensive chlorine in the impregnation vat. These con-
ditions in the furnace, however, may cause enormous losses by
volatilisation, the endeavour to save a few pounds of chlorine
in the vat causing the loss of 30 or 40 per cent, of the gold in
the furnace. In ores where the percentage of copper, tfec., is not
large, and where, in consequence, salt need not be used in the
furnace, the roasting may be finished at a high temperature
without any disadvantage, and the consumption of chlorine may
be thus reduced to a very low point. Thus certain ores from
Dakota, containing only 1 or 2 per cent, of sulphur, and con-
sisting chiefly of silica, were chlorinated by the author in
revolving barrels, using only 3J pounds of chlorine per 2,000
pounds of ore. Even in this case, there was a strong excess of
chlorine in the ore after the solution was complete, and the
amount used could probably have been still further reduced
without lowering the percentage extraction of gold. This ore
contained 10 dwts. of gold to the ton, and over 80 per cent, was
extracted. This was an extreme case, and it is seldom that so
little chlorine is sufficient. Mr. Butters states* that at his mill
at Kennel, California, where all descriptions of concentrates and
*
Eng. and Mng. Joum., Dec. 20, 1890.
pyrites were treated by the vat process, the average consumption

of chlorine was 12 pounds per 2,000 pounds of ore. At Deloro,
Canada, the amount used in the barrel process was from 12 to
18 pounds per 2,000 pounds of ore, but at the Haile Mine,
South Carolina, only about 3 or 4 pounds per ton. In giving
the amounts of chlorine which are used both in the vat and
barrel processes, side by side, the intention is to show that the
quantity absorbed depends on the nature of the material treated
and not on the process used and the amount of water present,
which are immaterial within certain limits as far as this point is
concerned.
Leaching the Charge. When it is judged that the impreg-
nation has lasted long enough for all the gold to be dissolved,
the excess of chlorine gas is removed, the lid is taken off, and
water is added to the charge to wash out the soluble chloride of
gold. The water may be added from below, and is then either
allowed to overflow at the top, or is subsequently drawn off again
at the bottom, the inflow being suspended. It is far more usual,
however, to pour on water at the top, and let it flow out at the
bottom.
The water must be added carefully, as otherwise the ore may
pack unevenly, and channels may be formed through the mass,
and the leaching thus rendered imperfect. Water is usually run
from a tap on to a layer of gunny-sacking placed over the ore,
by which it is distributed in a fairly even manner. It has been
proposed to attach a coil of lead pipes, pierced with small holes,
underneath the cover, and so sprinkle the water all over the
ore in fine jets. In any case, water is added until it forms a,
layer 2 or 3 inches deep above the surface of the ore, and it is
then allowed to stand until gas bubbles have ceased to rise
through it, which happens in about half an hour. The stopcock
below the false bottom is then opened, and the yellow- or blue-
coloured solution (coloured by salts of iron, gold and copper),
which should have a strong odour of chlorine, is run slowly
through a filter, consisting of a canvas bag, into a small barrel
about 18 inches in diameter and 2 feet deep, the overflow of
which passes, by means of a launder or by rubber hose, to the
precipitating tanks. Some of the ore and sand, escaping with
the solution, is deposited in the canvas bag and barrel, but, if
much slimes are present in the charge, either the canvas bag
becomes clogged or the solution still remains turbid when it
enters the precipitating vats. Water is supplied on the top of
the ore as fast as it is drained away below, care being taken not
to let the surface of the ore emerge from the liquid. The leach-
ing is continued as long as any trace of gold can be detected in
the issuing liquid by protosulphate of iron. As has been ex-
plained elsewhere (p. 25), a reaction is visible in clear solutions
BO long as the liquid contains more than two-thirds of a penny-
CIILORINATION : THE VAT PROCESS. !259
weight of gold per ton of water, or one part of gold in one

million of water. The strongly-coloured'turbid solutions usually
encountered in mills, however, are not capable of yielding distinct
reactions, unless they contain much larger amounts of gold than
this. In particular, when large quantities of copper salts are
present in the solution, their strong bluish tints mask the
slight discoloration due to a precipitate of a small quantity of
metallic gold, and, moreover, they appear to interfere with the
precipitation itself, in some cases at least preventing it from
taking place.* It is always advisable to filter the solution by
asbestos or filter paper before testing the liquid. Filter paper
may be used, since it is very slow in precipitating gold from
dilute solutions, even if they are neutral, and this action is com-
pletely stopped if free chlorine is present. It is better to test
the clear liquid with stannous chloride under the conditions given
at p. 26, since the presence of salts of copper does not seem to
interfere with the reaction in this case, and, moreover, the
amount of gold present can be determined in very dilute solutions
with much greater accuracy than if ferrous sulphate is used.
Since a few minutes longer time occupied in leaching is of
small moment, while the extraction of a few more grains of
soluble gold from a ton of ore may be of the utmost importance
in the long run, it is advisable to continue to leach at any rate
until the water contains less than 1 part of gold in 5,000,000
(about 3 grains per ton), a point which can easily be determined
by means of stannous chloride properly manipulated. The last
charges of wash-water should not be mixed with the strong
solution, but stored in other vats and used again for the first
washings of other charges. In this way the amount of wash-
water does not become excessive, although the tailings are
cleaned more effectually than is usually the case. Re-precipi-
tation of dissolved gold in storage vats or impregnation vats is.
not to be feared, so long as there is an excess of chlorine present
in the liquid, and this can easily be ensured by adding a small
quantity to any solutions not smelling strongly of the gas.
The amount of water used varies according to the richness of
the ore and the method of leaching adopted. It is usually about
2 tons of water to 1 ton of ore, but in most cases part of the
water is used again in the next charge.
Precipitation of the Gold. The precipitating vat is of the
same materials as the leaching tubs, and may be from 5 to 7 feet
in diameter and 3 feet deep. There is no false bottom, and the
vat is often made wider at the bottom than at the top to prevent
any adherence of the gold to the sides. The wood is protected
by a coating of pitch or paraffin-paint, or is left without paint of
any kind. The vat receives a smooth finish inside to facilitate
perfect cleaning, and is set perfectly level to avoid loss of gold
*
Letter from Mr. Butters, Eng. and Mng. Journ., Dec. 20, 1890.
while the waste liquor is being drawn off. The precipitating

solution of protosulphate of iron (the reagent which has been
is usually introduced
chiefly used in practice in the vat process)
into the precipitating vat at the beginning of the filtering opera-
tion. Care must be taken not to introduce a wasteful quantity,
as it is better to make up for any deficiency when the gold-
solution has all been run in. This operation is conducted in
such a way as to impart a circular motion to the contents of
the vat, so that the solutions are mixed without hand-stirring,
but the latter is often resorted to in addition, in order to make
the precipitate settle better; flat wooden staves with round
handles are used for the purpose. The contents of the vat may
be tested with solutions of gold chloride and of ferrous sulphate
to determine whether the precipitation is complete, and the pre-
cipitant present in excess.
If lead or lime is present, dissolved in the solution, it will be
precipitated as an insoluble sulphate
on the addition of the ferrous
sulphate, and thus render the gold-precipitate impure
and less
easy to treat. The amount of lead in the solution is usually
small, unless hot water has been used for leaching, and most of
the lead chloride is, in any case, separated by the canvas filter.
The usual method of removing the lime is to add sulphuric acid
to the gold solution, and to let it stand for a few hours, when
calcic sulphate crystallises out, forming a crust on the sides and
bottom of the vat. The liquid is then drawn off and transferred
to another vat for the precipitation of the gold. Instead of this
method, Nelson A. Ferry, E.M., recommends* the addition of
molasses to the leach, before the addition of the sulphate of iron.
He dissolves 1 gallon of molasses in 30 or 40 gallons of water,
and keeps it for use, determining the quantity to be added
by a laboratory experiment. He states that in this way the
precipitation of the lime is prevented, but
a large excess of the
ferrous sulphate should be avoided, and the liquid kept slightly
acid. The gold then sometimes comes down nocculent at first,
but soon changes to its normal condition.
The ferrous sulphate is usually prepared on the mill by dis-
solving iron in sulphuric acid. When precipitation is complete
the liquid is allowed to remain at rest for some time, in order to
allow the gold to settle to the bottom. The old practice was to
leave it " overnight," but the length of time allowed has of late
years been greatly extended. Thus Mr. Chas. Butters f states
that forty-eight hours is usually sufficient, but that sixty hours
is better, and the determination of the extent to which the
settling has progressed may be made by tapping the solution at

various heights and filtering the liquid thus obtained. When
liquid, drawn from a point
a quart of 2 inches above the bottom
*
Eng. and Mng. Journ., Nov. 28, 1885.
t Eng. and Mng. Journ., Dec. 20, 1890.
of the vat, gives only a slight dark stain to a No. 7 Swedish

filter paper, on being passed through it, the
settling may be
regarded as complete. Mr. C. H. Aaron quotes instances*
\vhere, after forty-eight hours settling, as much gold remained
in suspension in the liquid which was drawn off as was equivalent
to 50 cents per ton of the ore treated. The conclusion may be
drawn from these statements that a certain amount of gold is
inevitably lost by being carried away in suspension, but with
care and patience the loss may be reduced to a low percentage,
and even at the present day, in spite of the introduction of
many other precipitants, ferrous sulphate is probably as widely
used as it has been at any time in the past.
When the waste liquid has been drawn off by a floating siphon,
more ferrous sulphate and fresh solutions from the leaching vat
are poured into the vat, and the process repeated until enough
gold has accumulated at the bottom to warrant a clean-up. This
may take place at intervals of from a fortnight to three months.
The clear liquid is drawn off as closely as possible, and the slime
scooped out and filtered through paper, or, by means of a press,
through canvas. Finally, the vat is thoroughly cleaned out by
rinsing it with water which is run off through a plug-hole, level
with the bottom, into a wash-tub. The gold precipitate is then
dried carefully and fused in graphite pots, with salt, sand, nitre,
borax, &c., as fluxes, according to the requirements of the case.
If the precipitate contains any considerable amount of impuri-
ties (such as oxides and basic salts of iron), which is usually the
case, it may be treated with hydrochloric acid before fusion. The
bullion produced varies from 920 to 990 fine, the alloying metals
consisting chiefly of iron and lead.
Cost of Working. The cost of treating concentrates or ores
by the Plattner process depends chiefly on the cost of roasting.
In 18G7, the total cost in California was stated by Kiistel to be
$14.55 per ton, but in 1872 it had been reduced to $11, the
expense of roasting being in each case about two-thirds of the
whole. At the works of the Plymouth Consolidated Mining
Company, California, in 1886,f the cost of treating 100 tons per
month was $9.40 per ton (the roasting accounting for $4.60 per
ton, or nearly one half), and at the Providence Mine, in the same
State, it was about $6.30, without including the expenses of
general supervision, interest on first cost, and depreciation of
plant.
It was estimated in 1888 J that, generally, throughout Cali-
fornia, a plant, capable of treating 6 tons of concentrates daily,
cost from $6,000 to $7,000 for its erection, while the cost of
extraction was about $10 per ton, and the proportion of the gold
*
Cal. State Mineralogist, 1888, p. 836.
t Trans. Am. Jnst. Mng. Eng., 1886.
$ Eighth Report, Cal. State Min., 1888.
extracted was from 90 to 92 per cent. It has, however, been more

recently stated by G. F. Deetken that the cost of
vat chlorina-
tion under favourable conditions need not exceed from $3 to $4
per ton, but he gives no details or facts in support of
his state-
ment, and it seems probable that the conditions required are too
favourable to be expected.
CHAPTER XIII.
CHLORINATION : THE BARREL PROCESS.

THE use of revolving barrels for chlorinating ores was perhaps
suggested by the old Freiberg method of barrel amalgamation.
It has already been mentioned, p. 215, that Dr. Duflos used a
revolving barrel in some of his experiments at Breslau, in 1848,
and obtained results almost identical with those given by the
vat percolation method. He, therefore, preferred the latter as
being cheaper. The next mention of revolving barrels seems
to have been contained in a patent taken out by Mr. De Lacy
in Victoria, in 1864. The process thus patented appears to
have been tried there, probably on a small scale, but certainly
never passed into general use and was soon forgotten. It need
not be described here.* In 1877, Dr. Howell Hears, of Phila-
delphia, patented a process which had some points of resemblance
to that described by De Lacy. This process was gradually
improved in practice, after having been adopted by several
mines in the United States, and in particular the improvements
introduced by Mr. A. Thies, of South Carolina, which were
everywhere adopted, caused the name of the Thies process to be
applied to the amended method of procedure. In 1887 the
barrel process was re-introduced with some modifications into
Australia by Prof. J. Cosmo Newbery and Mr. C. J. T. Vautin,
who applied it to the ore of the Mount Morgan Mine, where it
was worked for some years with conspicuous success. Messrs.
Newbery and Yautin obtained patents in many countries for
their improvements, and the method of treatment of ores by
barrel chlorination is now known in Australia as the Newbery-
Vautin process. Considerable improvements have been effected
in the practice during the last lew years, these improvements
having been largely due to the attention which was called to
the subject by the efforts of Mr. Claude Yautin.
*
For description vide O'Driscoll on Gold O/w, where is to be found a
curious history of the patent literature of chloriuation. This work must
be read with caution, as it appears to attach too much importance to one
of the later patent processes.
CIILORINATION : THE BARREL PROCESS. 263
The Mears Process. In this process the roasted ore is

charged into cylindrical barrels, together with enough water to
make an easily flowing pulp chlorine is then forced in under
;
pressure, and the barrel, which must be air-tight, is revolved

until the gold has been dissolved. The barrel is then opened
and discharged by gravity into a leaching vat below, where' the
soluble gold is washed out and precipitated by any of the known
methods. Dr. Mears became aware by an accident of the increase
caused by the use of compressed chlorine in the rapidity with
which gold is dissolved by the gas. He was experimenting on
the action of chlorine on roasted pyrites, when the discharge
pipe of his apparatus becoming stopped up, the gas accumulated
in the vessel until it was burst by the pressure. He then found
that the ore was perfectly chlorinated, although it had been
subjected to the action of the gas for a few minutes only.
The Mears process as practised on a large scale may be de-
scribed as follows:
Cylindrical iron barrels lined with sheet
lead and mounted on hollow trunnions are used, most of the
general arrangements being similar to those shown in Fig. 52,
p. 265, in which the Thies barrel is depicted. An aperture
of from 6 inches to 1 foot in diameter, situated in the middle of
the length of the barrel, serves for charging-in and discharging
the ore ; the aperture is covered by an iron plate which is kept
tightly screwed down while the chlorination is in progress.
The ore and water are introduced through this aperture, and
the plate then screwed on. The chlorine is generated in vessels
similar to those used in the vat process, and stored in a large
gas-holder made of lead. From this gas-holder the chlorine is
pumped direct into the barrel through a leaden pipe passing
through the hollow trunnion, and continued into the barrel as
the " gooseneck," which is a pipe passing vertically upwards,
close to the end of the barrel, starting from the centre, and
terminating in a hook-shaped curve near the top. The gooseneck
remains stationary and vertical when the barrel is revolved j
it is made of iron so as to be strong enough to withstand the
weight of ore which presses against it on revolving the barrel,

and is lined inside and outside with lead to protect the iron
from the action of the chlorine. The object of the gooseneck is
to prevent the ore and water from flowing into and clogging
the pipe intended for the introduction of the chlorine.
When it has been charged, the barrel is exhausted of air by a
steam-jet exhaust pump, and chlorine is then pumped in until a
pressure of 40 or 50 pounds to the square inch is attained.
Another method of introducing the chlorine is to fill a large
receiver with the gas at high pressure, and then to connect it
with the pipe leading to the gooseneck, when the gas rushes
into the barrel. The effect of the high pressure in the barrel,
however it has been obtained, is to make a strong solution of
chlorine in contact with the ore. The previous exhaustion of

the air is of little importance, as the quantity present in the
barrel is small, at least three-quarters of the space inside the
barrel being usually occupied by the charge of ore and water,
while it is doubtful if such air as is present does any harm.
Messrs. Newbery and Vautin, indeed, subsequently claimed that
the presence of a large quantity of air is actually beneficial.
The barrel is then revolved at the rate of six to ten revolutions
per minute by a belt and pulley (the latter being fixed on a
sontinuation of the trunnion), or by means of cog-wheels, or best
of all by a friction clutch. The ore is kept constantly stirred
and tumbled about by tho revolution of the barrel, the advan-
tages gained from this course being as follows :
1.
Every particle of ore is exposed equally to the action of
the chlorine.
2. Native gold is usually alloyed with more or less silver. If
this metal is present in large excess, the dissolution of the gold
is stopped after a certain point has been readied, by the formation
over it of an insoluble coating, usually supposed to consist of
chloride of silver. Such alloys cannot be dissolved by aqua regia
unless this silver chloride is removed at intervals by friction or
by solution in ammonia. It is doubtful whether chloride of
silver isformed to any extent in the cold by free chlorine,
although no doubt the nascent chlorine produced in aqua regia is
more potent, but, whether the coating formed consists of silver
or of chloride of silver, it must in either case be in a powdery
condition, and so is readily removed by the mechanical attrition
of the particles of ore against one another, caused by the rotation
of the barrel. A clean surface of gold is thus continually offered
to the action of the chlorine. Coarse particles of gold may also
be covered and protected by a coating of undissolved chloride of
gold in the vat process, where so little water is present, but, in
the barrel, the larger amount of water present instantly dissolves
all such soluble salts.
The pressure of gas inside the barrel is tested from time to
time by opening a valve to which a pressure-gauge is connected.
A pop-valve may also be used for the same purpose. When
chlorination is believed to be complete, the excess of gas is
drawn off by the exhaust apparatus, and stored up or discharged
outside the building. The barrel is filled with water, and the con-
tents then discharged into a filtering vat, where the solution is
separated from the ore and precipitated as usual.
This description applies in great part to the other barrel chlori-
nation processes which differ little. The process was formerly in
use at the Phoenix and Haile Mines, in Carolina, and at Bunker
Hill Mine, California. The chief disadvantages in the process
were the rapidity with which the gooseneck wore out, and the
great strength and corresponding cost of the barrels rendered
CHLORINATION I THE BARREL PROCESS. 265
necessary by the high pressure used. It was also found to be

difficult to
keep the stuffing-boxes in the hollow trunnions from
leaking, and the cost of repairs was excessive. Modifications
were made by Mr. Adolph Thies, at Bunker Hill, with a view
to remove these objections to the
process, and his modifications
have been everywhere adopted. The amended method of
pro-
cedure may be called by the name of its inventor.
The Adolph Thies Process. Mr. Adolph Thies was in
Fig. 52.
Scale. 1 in. = 2 ft. 6 ins
charge of the Hears process at the Bunker Hill Mine for nearly
four years before he made the improvements, in. the year 1881,
which have been associated with his name. He found that the
hollow trunnions, the gooseneck, and the pressure pumps could
all be dispensed with, and the chlorine gas generated inside the
barrel itself to any required extent by the use of bleaching
powder and sulphuric acid. This method had been mentioned
by Hears in one of his earlier patents, but had been abandoned
in favour of pumping the gas into the barrel. Thies proved it
to be cheaper and better, as all joints liable to leakage are
LeadLined \
Precipitating :
Tank
Fig. 53.
dispensed with, and the machinery is simplified and rendered

much cheaper. Although there is no doubt that chlorine
under very high pressure acts more efficiently and rapidly than
at the ordinary pressure, still Thies showed that very good
results could be obtained at a reduced cost by using a moderate
pressure of chlorine of only a few pounds per square inch. The
barrel used is similar to that already described as suitable
to the Hears process, except that the trunnions are solid. It
is shown in Fig. 52 in transverse and
longitudinal section,
while a complete Thies plant is shown in Figs. 53 and 53a.
CHLORINATION : THE BARREL PROCESS. 267
An enlargement of part of the end of the most modern type of

chlorination barrel is shown in A, Fig. 54 ; in B, the construc-
tion of an older form of barrel, which is stated to have been in
use as late as the year 1891 at the Golden Reward Mill, is
shown for purposes of comparison. The shaded parts of A and
B are made of iron (the barrel-head and the flange being cast-
iron, and the cylinder, boiler-iron) ; the lead lining is not shaded.
In A the lead lining is burnt-on to the flanged cylinder and to
the head respectively, and the head then firmly bolted to the
flange. The lead-joint C is made tight by the blows of a
hammer on a blunt chisel directed on the surface of the lead,
in the direction shown by the arrow. When repairs are needed
inside the barrel, the cast-iron head is removed and the work
thus easily done. This barrel never suffers by leakage, and the
cost for repairs is usually nominal. In the other barrel the
lead lining of the end is burnt-on to the lining of the cylinder at
D, leaving a hollow space, E. In consequence of the existence
Fig. 54.
of this space, the lead-joint is often broken when the pressure

in the barrel is great, and leakage then occurs; moreover,
repairs must be effected by a workman crouching inside the
barrel.
In charging-in, the water is added first, being run in by a hose
through the manhole, until it reaches a certain mark on the
inside of the barrel. The amount of water to be used varies
with the nature of the ore roasted pyrites absorbing much
more water than siliceous ores the quantity required is usually
;
from 40 to 60 per cent, of the weight of the ore i.e., from 80 to

120 gallons of water per 2,000 Ibs. of ore. This is enough to
make an easily flowing pulp. It is much more than that
employed in the Plattner process, where it is necessary to keep
the mass porous, so as to enable the gas to pass through it. In
the barrel, however, this reason for limiting the amount of water
does not exist, and the ore is really treated by a solution of
chlorine water. If too small a quantity of water is used, so that
CHLORINATION : THE BARREL PROCESS, 269
the pulp is not quite free-flowing, lumps are formed which are
not broken by the revolution of the barrel, and these lumps are
not perfectly chlorinated.
The ore is let fall into the barrel down a shoot from an over-
head hopper, which may conveniently be made to contain the
exact quantity of ore required for a barrel charge. The ore
should be perfectly dry (not cooled after roasting by too much
"
wetting down "), as otherwise it sticks in the hopper, instead
of sliding freely down the shoot. The latter may be conveni-
ently made of canvas, so that it can be looped-up out of the way
when not in use. The chemicals (bleaching powder and sul-
phuric acid) may be added in one of two ways. Either the lime
is thrown into the water before the ore is added, and the acid
subsequently poured upon the upper dry surface of the latter,

just before the manhole is closed ; or the acid is poured into the
water, through which it sinks to the bottom without mixing, the
ore let fall next, and the lime added on the top. This latter
system is perhaps the better, as, if it is used, the generation of
chlorine is never begun before the barrel commences to revolve,
provided ordinary care is taken.
The amounts of bleaching powder and sulphuric acid to be
added depend on the nature of the ore, and the care with which
it has been roasted. If there is nothing in the ore which is open
to attack except gold, the amount of chlorine actually absorbed
in the course of the chemical reactions is extremely small, 1 oz.
of gold requiring only 0'54 oz. of chlorine to convert it into the
soluble trichloride.
The bleaching powder used should be of the finest quality
obtainable, to prevent the introduction of an unnecessarily large
amount of sulphate of lime into the charge. Bleaching powder
usually has assigned to it the formula Ca(OCl)Cl, or Ca(OCl).,
+ CaCl 2 and may be, as the latter formula would indicate, a
,
mixture of hypochlorite and chloride of lime. The reaction,

with acid is usually expressed thus
CaCl 2 + Ca(OCl) 2 + 2H 2 S0 4 = 2CaS0 4 -f- 2H 2 -f 2C1 2
This equation certainly does not accurately represent what

happens, as much less chlorine is liberated than is indicated by
it. The amount of "available" chlorine (i.e., that which is
liberated by the action of acids) contained in commercial bleach-
ing powder varies from 20 to 35 per cent. Bleaching powder is
gradually decomposed by the carbonic anhydride in atmospheric
air (chlorine being liberated), and must consequently be kept
separated from it as completely as possible. Even if preserved
in hermetically sealed vessels, however, it suffers a slow change,
by which some chlorate of lime is formed and the amount of
available chlorine reduced. Under these circumstances, it is
necessary to re-deteriuine the value of the bleaching powder at
270 . THE METALLURGY OF GOLD.
short intervals of a few days. The shortest and best method of

effecting this is to grind a sample in a stoneware mortar under
water, and add the emulsion to an excess of a solution of potassic
iodide. The whole of the available chlorine instantly displaces
an equivalent quantity of iodine, which is set free and may be
readily estimated by a standard solution of hyposulphite of
soda in the usual manner. From this the amount of available
chlorine in the sample is calculated. The whole operation can
be performed in from ten to fifteen minutes when the standard
solution has been prepared.
In calculating the quantity of lime to be added to a barrel
charge, there are several considerations to be taken into account.
A saturated solution of chlorine acts more rapidly than non-
saturated solutions. At the ordinary temperature and pressure,
water dissolves about 2^ volumes of the gas, and it is desirable
that the amount added to the barrel should be enough for this
saturated solution to be formed. Now, suppose the barrel to
have a capacity of 40 cubic feet, and that the charge consists
of say 125 gallons of water and 2,500 pounds of ore. The
volume of water added is 20 cubic feet, and if the particles of
ore have a mean density of 3'0, which is approximately true in
the case of many samples of roasted pyritic ore, the volume of
ore and water will be, together, 33^ cubic feet, leaving 6^ cubic
feet of air space inside the barrel. There must be, therefore, 46f
cubic feet of chlorine dissolved in the water to make a saturated
solution, and GJ cubic feet of chlorine in the air space, in order
to keep that amount in solution. The weight of 53 cubic feet of
chlorine is about 10'4 pounds, which could be supplied by the
decomposition of about 30 pounds of bleaching powder or 24
pounds per short ton of ore. This quantity may be taken as
one well adapted for the efficient chlorination of ordinary ores,
the chlorine acting rapidly at that pressure. Nevertheless, much
smaller quantities are in use with complete success on certain,
ores, and, as has been already stated, no general rule, applicable
to all cases, can be laid down. When much oxide of copper or
other metallic oxides capable of absorbing chlorine are present,
or if the ore is insufficiently roasted, the pressure rapidly falls
off, and a further addition of chemicals may become necessary,
as at the Pho3nix Mine. The total pressure exercised by the air
and chlorine, if the solution is saturated by the last-named gas,
is equal to two atmospheres i.e., one atmosphere in excess
of the normal and this may be taken as the greatest pressure
which it is advisable to maintain inside a barrel. If greater
pressures are used, either very expensive barrels are required,
or else the valves, manhole, <fec., soon begin to leak.
It must not be forgotton that, at higher temperatures, water
dissolves less chlorine than the amount mentioned in the last
paragraph, and, consequently, a pressure of chlorine equal to
CHLORINATION : THE BARREL PROCESS. 271
that of the atmosphere will be obtained with the use of smaller

quantities of chemicals. Thus, at 40 C., with the other condi-
tions identical with those given above, the amount of chlorine
required to give the same pressure will be only 6 -5 pounds
instead of 10'4 pounds, and, as the temperature rises above this,,
the quantity of chlorine necessary falls off rapidly.*
The relative amounts of bleaching powder and acid used is
varied with different ores. Theoretically, according to the equa-
tion given above, 7 parts of chloride of lime require 6 parts of
sulphuric acid for complete decomposition, but practically a
little more sulphuric acid is always added, because it is desirable
to maintain an excess of the acid in the charge. This is done
in order to prevent lead and lime from getting into solution as
chlorides, which would entail a loss of chlorine, and also to assist
in dissolving any oxide of copper that may be present, which
otherwise would also absorb chlorine. The proportions added
are usually 6 parts of bleaching powder to 7 or 8 parts of
sulphuric acid of 66 B.
The time occupied in chlorinating usually varies from three to
six hours. The continued presence of an excess of chlorine gas
should be tested from time to time by opening a small valve momen-
tarily. If the presence of the free gas cannot be detected, the
barrel must be opened and further supplies of chemicals added.
The amount of pressure in the barrel is not alone sufficient to>
prove the presence or absence of an excess of chlorine, as other
gases may be generated and exert considerable pressure. When,
the chlorination is finished, the excess of chlorine is discharged
by a hose-pipe outside the building, the barrel is filled up with
water, again revolved, and the liquid decanted on to large,
shallow filter-beds. The barrel is again filled up, revolved, and
decanted as before, and, finally, the whole charge is emptied out.
and another wash- water given to the charge on the filter. In
comparing this method of washing by decantation with that
of direct filtration, usually adopted, Mr. Thies found that, with
similar charges, the amount of water used in the latter case was
nearly double, and the time occupied much longer, while the
tailings contained 5J dwts. of gold per ton against about 1 dwt.
when washed by decantation. He
also found that there is little
difficulty in filtering through a bed of fine ore from 3 to 4J inches
thick, but if the thickness of the bed is greater, direct leaching
becomes very tedious and ineffective, and decantation is much,
better.
The leaching vats are usually constructed of wood, which is.
either lined with lead or coated with tar and pitch. At Bunker
*
For the amount of chlorine dissolved by water at different temperatures,
vide Schonfeld, Ann. Chem. Pharm. , vol. xciii., p. 25; vol. xcvi., p. 8. The
results are quoted in Roscoe & Schorlemtner's Treatise on Chemistry, vol. i.
p. 123, 1881.
Hill the filter tanks are rectangular, and measure 6 feet by 18
feet,and 18 inches deep. They are lined with lead, and incline
towards the drain hole, where the bottom is one inch lower than
it is at the other side of the tank. The filter-bed, as is usual in
California, is made of quartz-pebbles, gravel, and fine sand. In
other mills the leaching vats are usually round.
The Thies barrel process has been greatly altered and improved
during the last five years. The modern barrel chlorination pro-
cess, as practised in Dakota, differs from it in several essential
particulars ; it is described in Chapter xiv., pp. 291-304.
Mechanical Difficulties of Leaching. The difficulties of
leaching vary enormously with the character of the ore and the
treatment to which it has been subjected. Concentrates are
among the best leaching ores, as, even if they were originally
in a state of extremely fine division, the oxidising roasting in
many cases appears to cause an agglomeration of the particles
into porous granules, which do not pack down readily, and do
not resist the passage of liquids. A
small quantity of red oxide
of iron in a very fine state of division is often present in roasted
concentrates, and this is carried away by the water, passing into
and partly through the filter-bed and appearing in the precipi-
tating vat. This material, on settling to the bottom of the gold
solution, often appears to carry clown with it a large proportion
of the gold, forming a layer of slimes which are extremely rich,
and very difficult to treat except by smelting. In some cases,
these iron oxide slimes are present in roasted pyrites in such
quantities that leaching is greatly interfered with, and made very
tedious. Siliceous ores, if properly pulverised, usually present
no difficulty in leaching, but aluminous ores are exceedingly
troublesome. At Mount Morgan, Queensland, the ore consists
chiefly of h yd rated oxides of iron, which offer the greatest
possible resistance to leaching if treated raw, but, if roasted, the
loss of their water of hydration is found to be accompanied by
a remarkable agglomeration of the particles of ore, a-nd ordinary
gravitation leaching is thus rendered possible. The roasting
is in this case performed merely for the
purpose of facilitating
the leaching, as it is stated that in the raw ore there are no
constituents, except gold, which are readily acted on by chlorine.
In order to quicken the process of leaching various appliances
liave been suggested. Different forms of vacuum pumps have
been used, the amount of air in the space below the filter-bed
being reduced by them, and the liquid thus forced through by
atmospheric pressure. Such methods have not apparently been
attended with any conspicuous degree of success, as the increased
packing of the ore tends to neutralise the effects of the pressure
on the liquid. In 1889, at the Colorado Gold and Silver Extrac-
tion Company's Mill at Denver, the effect of the use of increased
pressure of air applied directly on the surface of the liquid was
THE PRECIPITATION OF GOLD. 27$
tried. A
special cast-iron vat was constructed capable of sustain-
ing an internal pressure of 100 pounds per square inch, and
furnished with valves, so that air and water could be simul-
taneously pumped into it. It was found that ores, which
entirely prevented the passage of water through them, even
after a vacuum of 20 inches of mercury had been established
beneath the filter-bed, could be leached with great speed under
a pressure of from 30 to 50 pounds per square inch. Moreover,
when the leaching was complete, the ore could be freed from the
water more completely by the passage of a current of air through
it than by gravitation alone. This method, which was suggested
by Mr. Dennes, the Company's engineer, has since been adopted
at the J vapid City chlorination mill, and also, in a modified form,
at the Golden Reward Mill, Dakota. The chief objection to it
seems to lie in the great additional expense incurred in the
construction of the leaching vats and in working the pumps.
Mr. Riotte has suggested* that the wash-water should be
thoroughly mixed with the ore by agitation, and then removed
as completely as possible by squeezing in a filter -press. To
effect this, the ore, togetherwith the necessary amount of water,
is passed successively through two revolving barrels, entering
and leaving them by means of hollow trunnions. The mixing is
accomplished inside the barrels by means of projecting inter-
nal ribs, and the charge passes continuously through, and is
received into large filter-presses. These are set to work as soon
as they are full, and squeeze out all the liquid, retaining the
tailings, the pressure used being from 25 to 30 pounds per
square inch. Mr. Riotte finds that the average amount of
moisture retained in the ore after being squeezed is about 6 per
cent. As this would cause the retention in the tailings of from
3 to 6 per cent, of the soluble gold, according to the amount of
wash-water used, the method is obviously inapplicable to rich
ores, which would require to be subjected to treatment twice.
Centrifugal leaching has also been proposed, and it is stated that
the Mount Morgan ore can be leached in this way without
previous roasting.
THE PRECIPITATION OP GOLD.

In adopting a system of precipitation suitable for application
on a large scale, there are several points to be taken into con-
sideration. (1) The precipitant should be capable of decomposing
gold chloride rapidly and completely, even when the latter is
present in extremely dilute solutions. (2) All other substances
likely to be present in the solution should be left unprecipitated.
(3) The excess of free chlorine in the solution should be removed
by the precipitant by being converted into hydrochloric acid or
into chlorides, so as to avoid running
any risk of the precipitated
*
Eng. and Mng. Journ., March 31, 1888.
18
gold being partially redissolved. (4) The precipitate of gold

must also be in such a form that it is easily separable from the
liquid. It is possible to obtain it in such a finely divided form
that it will not settle in water at all, and will pass through the
closest filter.
The precipitants which have been proposed or used may be
conveniently divided into two classes.
1. Soluble precipitants and gases, by which solid particles of
gold are formed, suspended in the liquid. These are either

allowed to subside and are separated by decantation, or they
are removed by nitration.
2. Insoluble solid precipitants, on which the gold forms as a
deposit, and from which it has to be separated by subsequent

operations.
1. SOLUBLE PRECIPITANTS.
Ferrous Sulphate. This is the best known precipitant,

having been used by Plattner, see p. 217, and being still em-
ployed more frequently than any other. Some account of its
use has already been given in the section on precipitation in the
description of the vat process, p. 259. It is made by dissolving
iron in sulphuric acid, with the aid of heat, and the crude solution
thus prepared, which is in most cases .added direct to the gold
solution, always contains some free sulphuric acid. Precipitation
takes place according to the equation
2AuCl s + 6FeS0 4 = Au 2 + Fe 2 Cl 6 + 2Fe 2 (S0 4 ) 3
From this it might be inferred that one part by weight of iron,

dissolved in sulphuric acid, would precipitate 1 J parts of gold,
but the oxidation of the ferrous salt is effected in other ways,
notably by the excess of free chlorine present in the solution, so
that much more sulphate of iron is required than is indicated by
the equation. The difficulty of collecting and saving the preci-
pitated gold has already been dwelt on. The gold settles better
if it is well stirred, and Aaron recommends an addition of more
sulphuric acid and vigorous stirring, two hours after the

precipitation is complete, as a means of assisting the settling.
It was proposed by Mr. Vautin to collect the precipitated gold
by the use of centrifugal force. For this purpose the liquid,
containing finely-divided gold, was placed in a circular vessel,
capable of being rotated at great speed. On rotating the
cylinder, the liquid in it was also gradually put in motion, and
the finely-divided gold then moved outwards by centrifugal force,
and was pressed against and adhered to the inner wall of the
vessel, while the clear liquid could then be siphoned off. Although
stated to be successful on a small scale, this device has not yet
been adopted in practice. Besides gold, the only other metals
precipitated by ferrous sulphate are those which form insoluble
SOLUBLE PRECIPITANTS. 275
sulphates viz., lead, calcium, strontium and barium. The last

two of these are rarely present, and the others are dealt with in
the manner already described above. Basic iron salts are not
precipitated if enough free sulphuric acid is present, and, when
precipitated, they may be removed from the gold by treatment
with acids, or by slagging them oft' in the furnace.
Other proto-salts of iron are equally efficacious, but are never
used in practice.
Organic Substances. Many organic substances, such as
oxalic acid, formic acid, ether, &c., also decompose chloride of
gold, but are unsuitable as precipitants in practice, owing to
their slowness of action and high cost, or to the
extremely fine
state of division in which the gold is thrown down, so that it
is made difficult or impossible to collect. Nevertheless, it is
probable that some of the more easily oxidisable organic com-
pounds may at some future time be found to be suitable for the
precipitation of gold on a large scale.
Sulphuretted Hydrogen. This was formerly made at Deloro
by heating paraffin and sulphur together, and the gas diluted
with air was forced through the solution by means of a small air-
pump. The use of the air was to keep the solution agitated, and
to expel part of the chlorine mechanically, and so economise the
sulphuretted hydrogen, which is decomposed by chlorine. The
*
equation for this decomposition is given by Mr. W. Langguth
as follows :
2SH + 4H 2 + 8C1 = H 2 S0 4 + 8HC1

but more likely that the greater part of the change is repre-
it is
sented by the well-known equation
H 2 S + C1 2 = 2HC1 + S
although a small amount of sulphuric acid may be formed at the
same time. The precipitate of sulphur thus formed before and
during the precipitation of the gold, which begins before the
whole of the chlorine has been destroyed, is to be avoided, and
Mr. Langguth has, therefore, suggested the use of sulphur
dioxide, generated by burning sulphur or pyrites, or by heating
sulphuric acid with charcoal, to destroy the free chlorine, the
reaction being as follows
C1 2 + S0 2 + 2H 2 = H 2 S0 4 + 2HC1
When almost the chlorine has been thus converted into
all
hydrochloric acid, the passage of SO 2 is stopped, and sulphuretted

hydrogen, now generated by the action of sulphuric acid on iron
matte, is forced into the solution, destroying the last traces of
chlorine and precipitating the gold. This system was intro-
duced at the Golden Reward Chlorination Works in 1891, and
has been subsequently adopted at other mills in Dakota with
conspicuous success.
*
Eng. and Mny. Journ., Feb. 14, 1891.
The gold is precipitated as a sulphide mixed with more or

less sulphur, the reaction being represented by the following
equation
2AuCl 3 + 3H 2 S = Au 2 S 3 + GHC1
It issaid to take less than an hour at the Golden Reward

Works to precipitate the gold from 5,000 gallons of solution
(resulting from the lixiviation of from 25 to 50 tons of ore).
The liquid is quite cold, but the precipitate is in a collected,
voluminous and flocculent form, that settles quickly. It is left
undisturbed for two hours, and the liquid is then drawn off to
within 4 inches of the bottom of the vat, and passed through a
Johnson filter -press, provided with, a set of heavy, canton-
flannel filter-cloths. The head of liquid used for filtering is
25 feet, and the filtration is said to occupy from three to four
hours, according to the amount of sulphides already contained
in the filter. When the latter is full, a small air-pump is con-
nected with it and a current of air passed through it for an
hour to dry the mass of sulphides into hard cakes, which are
easily handled and removed. The precipitate is then roasted in
a muffle furnace, the filter-cloths being burnt with it. It is then
melted down with a little borax and nitre, the total loss in
handling being very small. The bullion is about 900 to 950
fine in gold, the remainder consisting chiefly of silver, copper,
lead and arsenic. The bulk of the precipitate remains at the
bottom of the vat. It is allowed to accumulate for a fortnight,
and then treated, together with the material from the filter
presses.
Of course, all the lead, copper, and silver contained in the
liquids are precipitated with the gold. If much copper is pre-
sent, the bullion may be very base, and Langguth suggests the
removal of the copper from the precipitate by dilute nitric acid.
It might also be removed with less cost from the roasted preci-
pitate by dilute sulphuric acid; but, when much copper is present,
it would probably be better to use some other agent for
precipi-
tation than sulphuretted hydrogen.
Sulphurous Acid, which has been already mentioned as a
cheap agent for destroying free chlorine, precipitates gold very
completely from dilute solutions, but its action in the cold is
slow until the liquid is almost saturated with the gas (water
absorbs 39 volumes of the gas at 20), and the gold settles slowly.
At higher temperatures the action of the gas is much more rapid
and satisfactory, and very little is wasted in saturating the liquid.
The experiment was made at the Portland Mining Company's
Mill, Dead wood, Dakota, of using sulphurous acid gas instead of
sulphuretted hydrogen, but it was found that the precipitate
obtained was so finely divided that it was not retained in a
Johnson filter-press, much of it passing through all the filter-
SOLID PRECIPITANTS. 277
clothswhich were tried, and being lost. As a precipitant, the

gas does not appear to present any advantages over ferrous
sulphate.
2. SOLID PRECIPITANTS.
Charcoal. The first experiments on the reducing action of

charcoal on chloride of gold in solution were apparently made
by Percy. In his laboratory, some sticks of wood-charcoal were
immersed in water, and 32-50 grains of gold in the form of
chloride were added 011 August 7, 1869. 85 more grains of gold,
in the form of chloride, were added on November 3, 1869, and
the bottle was left to stand. This bottle is in the Percy
collection, at South Kensington, and the surface of the charcoal
is now coated over with metallic gold. In the same collection
is a model in gold of the surface of the end of a stick of charcoal
left immersed in a strong solution of chloride of gold. The
fibres and vessels of the stem are all shown on the surface of
the metal.
Vegetable charcoal was first employed as a precipitant for
solutions of gold chloride on the large scale by Mr. W. M. Davis,
in a chlorination mill in Carolina, in the year 1880. Its use was
discontinued, however, and nothing further was heard of it until
it was adopted by Messrs. Newbery <fc Yautin, at the Mount
Morgan Mine, Queensland, in 1887. The method adopted there

is as follows: The solution is heated to boiling, the free chlorine
being thus expelled ; the liquid is then made to run slowly
through large shallow tanks filled with pieces of charcoal, varying
from the size of a small hazel nut to less than that of a pea.
The tanks are about a foot deep, and the overflow from one is
allowed to run through others, until the solution is found, on
testing, to be free from gold, which is deposited on the surface ol
the charcoal. When the latter is coated sufficiently with the
precious metal, it is burnt in small furnaces, furnished with
large dust chambers or apparatus for condensing the fumes, and
the ashes melted with borax, the gold being thus obtained in a
remarkable state of purity. Animal charcoal cannot be used
owing to the difficulty of burning it afterwards.
The exact action of the charcoal has not been fully demon-
strated. It acts slowly on cold solutions, and its action is not
rapid even at boiling point. It is under the disadvantage that
it does not
destroy free chlorine, which must therefore be
expelled by boiling or by passing a current of air through the
liquid before the precipitation of the gold is begun. Mr. Davis
states that 240 parts of charcoal are required for the precipitation
of 19^ parts of gold. The prevailing opinion is that the hydro-
gen and hydrocarbons remaining in the charcoal are the active
agents in the precipitation, hydrochloric acid and free gold being
formed. Charcoal is not now used in Carolina, the only place

where it is still employed being the Mount Morgan Mine.
Insoluble Sulphides. The use of sulphide of copper as a
precipitant for gold chloride was covered by a British patent,
taken out many years ago, but the discovery was not applied in
practice, and had been forgotten when Mr. C. H. Aaron described
his method of application.'* Experiments were subsequently
conducted at the various mills engaged in working the Newbery-
Yautin process, with the object of finding if the sulphides of
copper or of other metals could be used in practice on a large
scale. It was found in Denver that the best method of preparing
the precipitated sulphide of copper is to add a boiling saturated
solution of sulphate of copper to a boiling saturated solution of a
mixture of the polysulphides of sodium,- prepared by adding
sulphur to boiling caustic-soda solution, and to stir the mixture
vigorously. The sulphide of copper is precipitated in a granular
form, which settles quickly in water, and allows liquids to filter
through it readily, but offers a large surface, from the porosity of
the granules. It is moreover easily decomposed, and so is
especially active in precipitating gold. The reaction that occurs
may be expressed thus :
3CuS + 2AuCl 3 = 3CuCl 2 + Au 2 S 3

Some metallic gold may possibly be formed at the same time,
according to the equation :
3CuS + 8AuCl 3 + 12H 2 = 8Au + 24HC1 + 3CuS0 4

The cupric chloride is removed in solution and the sulphide of
gold precipitated on the surface of the granules of copper
sulphide. The precipitant may conveniently be contained in two
or three small vessels, through which the gold-solution passes
successively, either flowing in at the top and out at the bottom,
or in the reverse direction. The sulphide of copper and the
precipitated gold are partly carried off in suspension in the
liquid, which must, therefore, be passed through a filter of flannel
or other material. In order to quicken both the precipitation
and the rate of filtration through the filter-cloths, the liquid may
be heated, while a head of 10 or 15 feet makes filtration very
rapid. The free chlorine need not be expelled from the liquid,
as it is destroyed at once by the sulphide.
Mr. Blomfield found that the subsulphide of copper, Cu 2 S,
prepared by fusing together sulphur and copper, is more easily
handled than the precipitated sulphide, besides being cheaper.!
The Cu 2 S is prepared for use by crushing it and retaining only
that part which remains on a sieve having 100 holes to the
linear inch, after having passed through a 60-mesh sieve.
The precipitation vessels which he recommends are stout glass
*
Eighth Report Cal. State Min., 1888, p. 839.
t Eng. and Mng. Journ., Jan. 24, 1891.
SOLID PRECIPITANTS. 279
cylinders, 6 inches in diameter by 12 inches long, closed at each

end by glass-lined cast-iron caps held together by two bolts.
Each vessel has a false bottom to support the filter-cloth, and
is charged with 4 to 8 inches of the precipitant. Mr. Blomfield
recommends downward filtration of cold solutions, thus saving
the expense of heating them. It is stated that this precipitant
is now being used on the large scale at a chlorination mill in
South America.
Whether precipitated or fused sulphide of copper is used, it is
advisable to defer the clean-up as long as possible, until the first
filter has had the greater part of its copper replaced by gold. It
is then cleared out, and the contents of the second filter trans-
ferred to the first, and so on, fresh sulphide being added to the
lowest filter. The mixture of gold and copper sulphides from,
the first filter is carefully dried, mixed with borax, and melted
in plumbago crucibles. If the bullion produced is too base, some
of the copper may be removed by roasting the sulphides and
treating them with dilute sulphuric acid before fusion.
It has been proved by laboratory experiments'* that both fused
and precipitated sulphide of iron are more rapid in their action
than the corresponding copper compounds. The only apparent
disadvantage in their use lies in the fact, that some of the copper
contained in the solution is precipitated with the gold. When
copper is not present in the ore this drawback would not be felt.
Metals. In recent experimentst the author has found that,
in the laboratory, iron turnings constitute the quickest and most
trustworthy precipitant for gold chloride, its superiority over
the insoluble sulphides being more marked, however, at GO or
80 C. than at ordinary temperatures. The separation of the
gold from the iron may be readily effected by methods similar
to those used in the separation from zinc, which has to be done
in the MacArthur- Forrest cyanide process, as described below,
p. 318.
The order of the rate of precipitation of gold from moderately
dilute solutions of its chloride (one part in 20,000) appears to be :
metallic iron (quickest), sulphide of iron, sulphide of copper,

charcoal, ferrous sulphate, metallic copper (slowest). This order
is that noted near the boiling point of water. The relative
rate of precipitation by sulphurous acid and sulphuretted
hydrogen was difficult to compare closely with that of the above
reagents, but they appeared to act more slowly than iron and
the sulphides. It does not necessarily follow from the results of
these experiments that metallic iron is the best precipitant on a
large scale, and it has not been tried in practice as yet. Pos-
sibly the order given above would not be preserved unchanged
at ordinary temperatures.
*
Mining Journal, March 11, 1893.
\~Loc. cit.
MODERN PATENT PROCESSES OF CHLORINATION.

The Newbery-Vautin Process. In this process an effort
is made to combine the advantages of the Mears and Thies pro-
cesses. Since a high pressure of chlorine increases the rapidity
with which gold is dissolved, it is desirable to use pressure, but
Thies had shown that the economy effected by reducing the
quantity of chlorine employed outweighed the advantages gained
by the high pressure. Messrs. Cosmo Newbery and C. Vautin
proposed to obtain the desired pressure by pumping air into the
barrel, keeping the amount of chlorine as low as possible. It
was stated by them that the chlorine would be kept in a liquid
state or, at any rate, that it would be entirely dissolved in the
water if the air in the barrel were at a pressure of 60 to 80
pounds to the square inch. This is contradicted by theoretical
considerations, and no attempt appears to have been made to
prove it in practice. The only effect of the increase of air
pressure would be an increase in the amount of air, but not in
that of chlorine, dissolved in the water. The latter can only
be affected by variation of chlorine - pressure. According to
experiments made by the author at Denver, using the Newbery-
Vautin barrels, air-pressure does not seem to exercise any
influence on the rate of solution of the gold in ores, or upon the
percentage of extraction. The other proposal made by Messrs.
Newbery & Yautin was to leach by means of a single-acting
vacuum pump. Vacuum leaching has already been discussed on
p. 272. The particular form of pump proposed by Messrs.
Newbery & Vautin does not seem to possess special advantages.
Neither of these methods appear likely to pass into permanent
use in gold chlorination.
It has often been asserted that the Newbery-Vautin process
is identical with that formerly in use, with conspicuous success,
at the Mount Morgan Mine, Queensland. Nevertheless, as lar
as the author is aware, neither air-pressure in the barrel, nor
vacuum-leaching by means of a single-acting pump ever formed
part of the method of procedure at Mount Morgan. Mills de-
signed to use the Newbery-Vautin patents were erected about
the year 1888 in the Transvaal, New Zealand, Colorado, and
Hungary, but it appears that none of these mills are now at work.
The Pollok Hydraulic-pressure Process. In this process
also, there seems to have been a confusion of thought in the
mind of the inventor, who does not clearly distinguish between
the pressure of chlorine gas, which increases its chemical activity,
and air- or water-pressure, neither of which has any effect on the
condition of the chlorine or the rate of solution of gold. Mr.
Pollok ensures the solution in water of the whole of the chlorine
present in the barrel by filling the latter completely with water.
MODERN PATENT PROCESSES OF CHLORINATION. 281
When the barrel is closed and all the air has been suffered to
escape, more water is pumped in, until the pressure inside the
barrel rises to at least 100 pounds per square inch. The chlorine
is generated inside the barrel
by the action of bisulphate of soda
on chloride of lime, but it is doubtful if there is any economy on
any gold-field in the use of bisulphate of soda in place of
sulphuric
acid. Pollok claims that by this high pressure the solution of
chlorine is forced rapidly into the pores of the ore. The process
in other respects presents no novel features, and it does not
seem likely to come into general use.
The Swedish Chlorination Process. This method was
devised by Mr. Munktell, who seems to have worked it out
without having visited either the vat or the barrel chlorination
works already established in other parts of the world. The
process enjoys the distinction, according to the published accounts,
of having been worked at a profit. It was first used at the
Fahlun Copper Works, Sweden, and has since been introduced
into Hungary, where it is in successful operation at Brade and at
Boitzas. The process as worked at Fahlun may be briefly
described as follows The ore is roasted at a low temperature
:
with the view of obtaining the copper in the form of sulphates.

If silver is present, salt is added in the roasting furnace. The
calcined ore is then placed in false-bottomed, wooden vats and
leached with hot water, followed, if necessary, by hot, dilute
sulphuric or hydrochloric acids, by which all the copper and
much of the iron are removed from the ore. The solutions thus
obtained are run through tanks containing scrap-iron, by which
the copper and any silver in solution are precipitated. The
solution of ferrous sulphate may be saved and used subsequently
to precipitate the gold. The residues in the tubs are now in a
condition to yield up their gold readily, and are accordingly
treated by a solution of 0'6 to 0'7 per cent, of chloride of lime
(bleaching powder) in water, mixed with an equal volume of
dilute hydrochloric acid of specific gravity 1-002 or 1*003, or of
dilute sulphuric acid. These solutions are mixed in troughs just
before they flow into the ore-vat. Chlorine is slowly generated
by the action of the acid on the hypochlorite of lime, and, being
partly present in the nascent state, attacks the gold vigorously
in spite of the extreme dilution of the solution, while there is
very little smell of the gas above the vats. The liquid is passed
through the ore until gold ceases to be dissolved, after which the
tailings are thrown away. The solution is heated to 1 60 F. by
steam, and precipitated by ferrous sulphate. The collection of
the gold is expedited and ensured by adding acetate of lead to
the solution by this device lead sulphate is precipitated with
;
the gold, and, in settling to the bottom, carries the particles of

precious metal with it. This process was in continuous operation
at the Fahlun Copper Works from 1885 to 1888. In that period,
1,500 tons of gold ore and the tailings from 29,000 tons of copper
ore were subjected to treatment. It is stated that in the year
1886, the tailings from 14,000 tons of copper were treated with
the following results :
AVERAGE AMOUNT OF GOLD IN TAILINGS.

Before treatment,
After
.
... . . 41-82 grains per ton.

4-04
COST PER TON.

Chloride of lime (6 Ibs.), . 5d.
Sulphuric acid (8-37 Ibs.), . . Id.
Fuel for steam,

Labour, ....
Plumbic acetate and other reagents, .
9d.
Total,
The low cost assigned to labour is very remarkable. These

tailings had, of course, been previously crushed and roasted
before being treated as above.
In the same year 960 tons of gold ore were treated at a cost of
12s. 2|d. per ton. the ore containing 523'62 grains of gold per ton
before treatment, and 6*02 grains per ton after treatment, so
that no less than 98 85 per cent, of the gold was extracted.
-
The following details concerning the practice at Brade,

Hungary, may be added.* In addition to gold, the ores contain
iron pyrites, barytes, zinc-blend^, antimonial minerals and
argentiferous galena, and also in soine cases calcite and carbonate
of magnesia. The concentrates, which are subjected to treatment
by the Munktell process, contain 0'86 oz. of gold and 4 ozs. of
silver per ton, and 36 to 40 per cent, of sulphur. It is obvious
that such concentrates could not be treated advantageously by
the modern barrel process as described on pp. 291-304.
The oxidising roasting takes twenty-eight hours, after which
5 per cent, of salt is thrown upon the ore and mixed thoroughly
with it, and four hours later the charge is drawn from the
furnace and allowed to cool slowly in a covered-in brick pit.
The loss by volatilisation is, however, considerable, and it is
proposed to build seven-story furnaces. The leaching vats are
made of wood, lined with lead, and are 3 metres broad, 5 metres
long, and 0-75 metre deep, and hold 10 tons, the charge being
0*5 metre deep. There is a false bottom and filter-bed of quartz
as usual. The ore is leached with five different solutions in
succession, viz. :
(1) Warm
water, 2,500 gallons being required
for the charge :Chlorides of copper and zinc and about 25 per
cent, of the silver are removed in solution. (2) solution A
containing 2 per cent, of hyposulphite of soda, 1,600 gallons
*
Proc. Inst. Civil Eng., Session 1891-92.
MODERN PATENT PROCESSES OP CHLORINATION. 283
being used ; this dissolves the rest of the silver. (3) Dilute
sulphuric acid, 2,700 gallons being used to remove the oxides of
iron. (4) Weak solutions of bleaching powder and sulphuric
acid to dissolve the gold ; and (5) Cold water to wash the ore.
The solutions are all preserved separately. The leaching takes
in all nine or ten days, the chlorination alone occupying three
days. The gold and silver are precipitated from their solution
by sodium sulphide, and the precipitate is dried, pressed, roasted,
and melted down.
The expenditure per ton of concentrates is as follows :
s. d.
Salt (110 Iba.), . . 17
Sulphuric acid (102 Ibs.), 6
Fuel, ....
Bleaching powder (25 Ibs. ),
Hyposulphite of soda (14'7 Ibs.),
Labour,
3
8
3
1
1
8
7
7
10
Amortisation,
Total, . . . . 26 3
The percentage of extraction is not given.
CassePs Process. In this process a solution of salt in
contact with the ore is decomposed electrolytically, and the
chlorine thus set free attacks and dissolves the gold, which is
deposited in a hollow iron shaft in the centre of the vessel. The
apparatus is complicated and ill-adapted for its purpose, and the
process has not been a practical success. It has now been
completely abandoned, but is interesting as being the first
attempt to generate chlorine by electrolysis for attacking gold.
The Julian Process. This was devised by Mr. Julian, of
Johannesburg, South Africa. After chlorination of the ore in a
barrel, mercury is added to the charge and the barrel again
revolved. The result of this is to amalgamate the coarse gold in
the ore and to reduce the gold chloride, chloride of mercury and
metallic gold being formed, the latter amalgamating with the
excess of mercury. The charge is then emptied out and passed
over a set of amalgamated copper-plate tables, by which the gold
amalgam is partly retained. The tailings are then passed
"
through a series of electrolytic cells," each of which has a bath
of mercury lying at the bottom, while an electric current is
passed through the water. In these cells, the compounds of gold
and silver, soluble in the water, and all floured mercury and
amalgam are supposed to be collected. The process has not yet
been proved to be a success in practice. The idea of decomposing
gold chloride by mercury is not new, and previous experience has
shown the action to be very slow and partial. The complicated
nature of the whole process is against it.
Green-wood Process. In this process gold is dissolved in
barrels by chlorine produced by the electrolytic decomposition
of a solution ofcommon salt. The problem of producing chlorine

economically by electrolysis is one of enormous importance, as
one of the chief causes of expense in chlorination lies in the cost
of the chemicals required. If power is alone required, then
wherever water power is available, the whole process may be
much cheapened. In the Greenwood patents, the chlorine so
obtained is not well applied, and the machinery need not be
described here.*
CHAPTER XIV.
CHLORINATION: PRACTICE IN PARTICULAR MILLS.
THE following description of the practice actually employed at

the works named will serve to show how far the general descrip-
tions given are applicable to particular cases :
THE VAT PROCESS.

1. Butters' Ore Milling Works, Kennel, California.!
The practice varies with the nature of the ore. The Idaho
sulphides from the Grass Valley constitute a case of extreme
difficulty. "They contain considerable quantities of lime and
magnesia, and some manganese, zinc-blende and chalcopyrite,
Avith 4 ozs. of gold and 8 to 12 ozs. of silver per ton. Roasting
with salt is prohibited by the heavy volatilisation losses thereby
incurred. The roasting is performed at a black heat, the charge
withdrawn, allowed to grow quite cold, and then wetted down
and screened immediately, in order to prevent the agglomeration
of the ore, as the presence of the anhydrous sulphates makes the
mass a natural cement. The ore is then leached in a vat for
from 48 to 72 hours with cold water, and the soluble sulphates
of copper, iron and zinc thus removed. The solution thus
obtained contains small quantities of silver and gold, perhaps
dissolved by the agency of the metallic salts. These are
recovered in -the copper-precipitating tank. The alkaline bases
and carbonates are next neutralised with sulphuric acid (60 to
100 Ibs. of 66 Baume acid being added per ton of ore), and
20 Ibs. (diluted with water) of the same acid per ton are then
added and circulated for 24 hours to dissolve the oxide of copper,
and the charge washed for 48 hours with cold water, shovelled
*
Adetailed description of the Greenwood process is given in Eissler's
etalluryy of Gold, pp. 351-309.
t Eng. and Mng. Journ., Dec. 20, 1890.
THE VAT PROCESS. 285
out, partially dried, screened back into the tank and impregnated
with chlorine. After leaching, the tailings contain from $4 to
$0 in gold and all the silver. Five per cent, of salt are added,
and the charge dried and roasted at as high a temperature and as
fast as possible, when little loss by volatilisation is experienced.
The silver is then removed by leaching with hyposulphite
solution."
It seems difficult to believe that this complicated system
(which, it will be observed, bears a strong resemblance to

Mr. Mimktell's methods in several important particulars), can
really be the most economical one that can be devised for this
ore.
2. Plymouth Consolidated Gold Mining Company,
Amador County, California.* The ore contains 11 dwts. of
gold, mostly in the free state. It is crushed in a stamp battery,
no. 8 screens (40-mesh) being used, passed over a length of
20 feet of amalgamated plates (the upper one of which is copper
and the rest silver-plated) and then concentrated on Frue
vanners. The concentrates amount to from 1J to H
per cent,
of the weight of the ore, and contain from 5 to 10 ozs. of gold
per ton. They are treated at the rate of 100 tons per month,
being kept damp until charged into the roasting furnace, " to
prevent the formation of lumps. A " Fortschaufelungsoferi is
used for roasting, 80 feet long by 12 feet wide, its hearth
consisting of a long continuous plane, holding three charges at
one time, which are kept separate. The three stages are called
"
the " drying," " burning," and " cooking stages. In the
(middle) burning stage, the bed of ore is kept thin and occupies
double the space of each of the other charges. The furnace is
worked by three eight-hour shifts of one man. The charges
weigh 2,400 Ibs., including 10 per cent, of moisture, and on an
average contain 20 per cent, of sulphur. Just before the sulphur
ceases to flame, f per cent. i.e., 18 Ibs. of salt are added.
The chloridising vat is 9 feet in diameter and 3 feet deep it :
holds 4 tons of ore. The filter-bed is 6 inches deep, and con-

sists of, (a) at the bottom, wooden strips, f inch wide, placed
1 inch apart ; (b) above this, 6-inch boards placed 1 inch
apart and laid across the strips ; (c) coarse lumps of quartz,
diminishing upwards to fine stuff; and (d) at the top, a cover of
6-inch boards placed similarly to the lower layer, but crosswise
to them. Chlorine is introduced on both sides of the vat, which
is left luted-up for two days, and then leached for four or five
hours, the tank being kept full of water during the operation.
A gunny-sack protects the surface of the ore from the direct
impact of the water from the hose. The ore in the impregnation
vat contains about 6 per cent, of water (crumbling after it has
*
For a more complete description, see that given in the Eighth Report
Ccd. Stat. Min., 1888, of which the account appended is an abstract.
been sieved), and is sieved into the vat through a screen of

t^-inch mesh.
The gold solution is acidulated with sulphuric acid to pre-
cipitate the lead, and ferrous sulphate is then added to
it in
another tank, after which it is allowed to settle for two days

before the supernatant liquor is siphoned off. The gold is left
to accumulate for fourteen days and the wet gold is then filtered
and fused in graphite pots. The average extraction is from
95 to 96 per cent, of the gold, and 7 dwts. per ton are left
in the tailings. All wood is protected from the action of the
acids by paraffin- paint. The cost of milling is said to be
39 cents per ton of ore, and the cost of concentration, roasting,
and chlorination, $13-40 per ton of concentrates.
3. The Alaska Treadwell Mine.* These works turned out
more gold than any other chlorination mill in the United States,
until the Golden Reward and other Dakota mills were started.
The old-fashioned vat process in use has been adopted after
thousands of dollars have been spent in testing newer methods.
The works have been in active operation since the year 1884.
The material treated consists of the sulphides collected by Frue
vanners from a stamp battery, and contains 40 per cent, of
sulphur, mostly as iron pyrites, although the percentage of copper
is increasing. The gangue is quartzose, containing from 2 to 5
per cent, of calcite, which necessitates the addition of salt in the
roasting furnace.
Roasting. This was first effected in Bruckner cylinders, which
were abandoned owing to the large amount of fuel consumed and
the enormous losses by dusting and volatilisation (which are said
to have amounted to 30 per cent, of the gold). The automatic
Spence furnace was then tried and proved to be useless until it
was used as a reverberatory. Six were erected, and the cost
of roasting was reduced by one-half, but the capacity was small
(10 tons per day in the six furnaces) and the consumption of
fuel great, and a reverberatory furnace being erected, was found
to be more satisfactory. The Spence furnaces were accordingly
discarded, and 20 tons of concentrates are now roasted per day
in four reverberatory furnaces of 13 by 65 feet, inside measure-
ment, at a low cost. The ore is not roasted quite dead, Mr.
Burfeind, the superintendent, having decided that when salt is
added the best results, with least volatilisation, are obtained by
leaving some sulphates undecomposed. He gives it as his
opinion that improper roasting is responsible for all losses of gold
in the tailings and by volatilisation.
The roasted ore is spread on the cooling floor, wetted down
and sifted carefully into vats, each of which holds 4| tons, and
is then
impregnated with the gas. This operation occupies four
hours, after which the vat is left untouched for thirty hours,
*
Eng. & Mng. Journ., April 11, .1891.
THE BARREL PROCESS. 287
fresh gas being forced in a few hours before leaching. The

leaching usually requires twelve hours. The tailings are sampled
and assayed, and, if found sufficiently poor, are sluiced into
the sea.
The solution is run into collecting tanks and thence to the
precipitating vats, which already contain the necessary amount

of ferrous sulphate in solution. The precipitation is complete
when all the solution has been run in, or when the vat is full.
It is then stirred briskly for a few minutes and left to settle for
from eighteen to twenty -four hours, when the supernatant liquor
is siphoned off and passed through a large filter. The super-
natant liquor usually contains from 23 to 25 cents of gold per
ton of material treated, and this is all saved in the filter.
The clean-up is made twice a month, the drying and melting
of about 600 ozs. of gold being done by one man in one day. The
gold is washed into a small tub, allowed to settle, and the
supernatant liquor returned to the precipitating vat. The gold
is dried in an iron pan without filtering, and melted with a
little borax.
Each one of the chlorination vats, holding 4J tons of ore, cost
1, and lasts for three years without any repairs. The filter in
it costs only the price of a few gunny-sacks, and lasts for six
months without any attention. The other vats are expected to>
last a lifetime. The ferrous sulphate is prepared on the works
from sulphuric acid and scrap-iron. During the six months,
ending November 31, 1892, 120,002 tons of ore were mined and
milled, the total cost being at the rate of $1-32 per ton. The
concentration of the ore yielded 2,703 tons of sulphides, which
were chlorinated at a cost of $8-42 per ton. This is very low
for the Plattner process, although perhaps high when compared
with barrel chlorination. In 1890, the cost was at the rate of
about $10 per ton, the yield in gold being over $40 per ton.
Later results are as follows:
YEAR.
taken from the account given by the manager, Mr. John G.

Eothwell, M.E. The ore is hand-sorted, crushed in rock-
breakers and Cornish rolls so as to pass a 16-mesh screen, then
sized and concentrated by jigs. The concentrates are roasted in
two White-Howell cylinders, being passed through one, where
nearly all the sulphides and arsenides are burnt, and then into
the other, where dead-roasting is performed the ore falls thence
;
direct on to the cooling floor, from which it is loaded into cars

and sent to the barrels. The first cylinder is 30 feet long and
5 feet in diameter, and has eight brick shelves to assist in
stirring the ore. The second cylinder is 20 feet long and 4 feet
in diameter, with six shelves. The air supplied to the first
cylinder is heated by the escaping gases of the second. Both
cylinders are jacketted by air-spaces and a covering of mineral-
wool and paper. Ten tons of concentrates are roasted in
twenty-four hours, yielding over 4 tons of arsenious oxide,
which is condensed in brick chambers. These cylinders are
specially arranged so as to receive an unusually large supply of
air,by apertures in the fire-bridge and around the fire-box; they
are said to treat more ore with less loss by dusting than ordinary
White-Howell furnaces.
The chlorinators are cylinders of
J-inch boiler-iron, having
on; each head has a long
cast-iron heads ri vetted or bolted
trunnion turned true and cored-out, with a shafting-box and
gland on it. The cylinders are lined with 20-lb. (to the
square foot) sheet-lead, the joints being burned. The barrels
are of two sizes. The smaller size is 40 inches in diameter and
44 inches long (inside measurement) ; it has a capacity of
32J cubic feet, and holds 1 ton of ore. The larger size is
60 inches in diameter arid 72 inches long, has a capacity of
118 cubic feet, and takes a charge of 3 tons of ore.
A manhole in the centre of the shell enables a man to get
inside the barrel for repairs ; in the centre of the manhole cover
is an aperture for charging and discharging. Inside the cylinder,
and extending from the charging aperture in the direction in
which the cylinder turns, is a pocket covered and lined with
lead, which is burnt on to the lining of the cylinder. The pocket
in the 1-ton barrel holds 100 Ibs. of sulphuric acid, in the
3-ton barrel, 300 Ibs. There are several lead-covered shelves
bolted to the inside of the cylinder, their use being to stir the
pulp thoroughly. Through the trunnion passes the "gooseneck,"
(viz.,
an iron pipe covered and lined by lead), which is turned in
the direction in which the barrel revolves. The gooseneck is
held stationary in its upright position by clamps.
About 120 gallons (i.e., 60 per cent.) of water are added to the
ton (of 2,000 Ibs.) of ore. The proportions of chemicals used are
6 parts of bleaching powder to 7 parts of sulphuric acid, the
absolute amounts varying with the quality of the ore and the
. THE BARREL PROCESS. 289
degree of perfection of the roasting. The usual amounts are from

36 to 54 Ibs. of bleaching powder, and from 42 to 63 Ibs. of acid
per ton. The lime is mixed with the ore and water, and the acid
poured into the pocket. The covering plate is then screwed on and
the barrel is set in motion, when the acid is immediately emptied
out of the pocket and is mixed gradually with the rest of the
charge. The cylinder revolves for from one and a-half to three
hours, after which the excess of gas is drawn off and the barrel
is partly exhausted through the gooseneck. Then the barrel is
opened and dumped into the filtering vat, which has a loosely
fitting cover placed on it, and an exhaust-pump connected with
the space below the filter-bed to quicken the leaching. Another
method adopted is to decant the liquid contained in the barrel
and conduct it to settling tanks; then to add more water, revolve
for a few minutes and decant again. In this way a charge of
three tons can be washed in two or three hours, while by
exhaust-leaching, 1 ton takes from six to eight hours.
The clear solution is precipitated by sulphuretted hydrogen
made from paraffin and sulphur, thus The generator is a small
cylinder of heavy boiler-iron with cast-iron heads rivetted in, and
the seams caulked. In the head is a hole big enough to admit a
man's hand for the purpose of cleaning out the vessel. In the
shell is a charging hole and an outlet for the gas, consisting of a
2-inch gas-pipe. The generator is built in a small brick fire-place
with a charging hole or pipe on the top, and a hand-hole outside.
A small fire is used, and the charge consists of one part of single-
pressed paraflin to two parts of common brimstone.
The precipitating tanks are fitted with tight covers (in which
are well-secured manholes), and the only outlet from them is a
wide exhaust-pipe leading out of the building. In front of these
tanks are lines of steam- and gas-pipes connected separately with
each one. The gas, after being generated, passes into a receiver to
condense the oils, &c., and an air-pump forces it from this through
a perforated lead pipe, nearly to the bottom of the precipitating
vats. An air-hole left in the receiver prevents the formation of a
vacuum there, and enables the pump to pass a mixture of sul-
phuretted hydrogen and air through the liquid, an arrangement
which economises the sulphuretted hydrogen. When precipita-
tion is complete, a short time elapses and the liquid is then
allowed to flow through several small filters.* Three of these
filters will empty a 1,500-gallon tank in from three or four hours.
The precipitate is then pressed into cakes, dried, roasted, and
melted with borax.
The ore is an arsenical sulphuret of iron in a gangue of quartz
and calcspar. The "mineral" contains about 42 per cent, of
arsenic, 20 per cent, of sulphur, and 38 per cent, of iron. The
* and Mng. Journ., Mar.
For a deocription of these filters, see Eng. 25,
19
roasted concentrates average $78-67 per ton in gold, and the

tailings about $3, the extraction being at the rate of 96-16 per
cent, of gold.
2. The Thies Process at the Phoenix Mine, Concord,
North Carolina, and the Haile Gold Mine, South Carolina.*
- The barrel
process is used at these mines, the methods of pro-
cedure being similar at the two places. Chlorine was formerly
forced in through a gooseneck, passing through the trunnion
(Hears* process), but it was found that the gooseneck wore out
quickly, and that its inner lead lining collapsed when the gas
was exhausted. It was, therefore, discarded after two years'
experience, and the chlorine is made inside the barrel by means
of sulphuric acid and bleaching powder. The ore is roasted in a
double-bedded reverberatory furnace, having a capacity of 3
tons of concentrates per day. The chlorinator is 42 inches in
diameter and 60 inches long, the capacity being 48 cubic feet, and
the charging-hole is 6 inches in diameter. The lining consists
of 10-lb. to 12-lb. sheet-lead. The weight of the charge is
from 2,000 to 2,500 Ibs., to which from 100 to 125 gallons of
water is added i.e., enough to make an easily flowing pulp. At
the Phcenix Mine, where the ore contains copper, one-half of the
acid and bleaching lime is added first and the barrel turned at
the rate of 15 revolutions per minute for 3 to 4 hours. For the
Phoanix ores, this half charge of chemicals consists of 20 Ibs.
of lime and 25 Ibs. of acid. After this time has elapsed the
barrel is opened, and the other half of the chemicals added, after
which the barrel is rotated for from two to three hours longer.
The charge is then let fall on to a filter-bed, 6 feet long and 8 feet
wide, the depth of the pulp on it being 4 inches, while the
filter-bed is 5 inches thick. The first solution is allowed to run
through until the surface of the ore begins to be exposed, when
it is again covered with water to the depth of 3 to 4 inches, and
when the ore surface appears again, the whole space above it,
11 inches deep, is filled, which is enough for the Phcenix ores.
In all, 2 tons of water are used to leach 1 ton of ore.
The bottom of the filter-bed consists of perforated glazed tiles
or the substance known as mineraline, and stones, gravel and
sand are successively piled on these as usual. The tanks for
precipitation are made less than 3 feet deep. The precipitation
is effected by ferrous sulphate, and care is taken that no soluble
chloride of calcium shall remain in the solution, or a bulky
precipitate of sulphate of lime will come down on the addition
of the precipitant. It is for this reason that an excess of
sulphuric acid is added to the charge, and more is added to the
solutions if they do not already contain some free acid. The
large amount of lime and acid added to this ore is necessitated
by the presence of copper. The solutions are allowed to settle
*
Eighth Report Col. Stale Min., 188S, p. 836.
for three days and then allowed to run through tanks filled with
scrap-iron, by which the copper is precipitated.
At the Haile Mine, where iron pyrites, free from copper,
is treated, only 10 Ibs. of lime and 15 Ibs. of acid per ton of
ore are used. Ten tons of roasted ore are treated in the large
barrels per day of ten hours, and 94 per cent, of the gold is
extracted.
The cost at the Haile Mine is $4*65 per ton, without reckoning
the expenses of superintendence and assaying, the cost of roasting
being $2 -70, and of chlorination, &c., $1'95. The following
details are given on the authority of Mr. Adolph Thies, the
inventor of the process used and the manager of the mill :
Fire-assay value of ore, delivered to the stamps, $4-50, of

which $1-45, in free gold, is caught on the plates.
Average assay value of raw concentrates, . . .
$30 per ton.
Sulphur in raw concentrates, . 40 to 45 per cent.
...
. .
Value of roasted concentrates, . . .

$50 per ton.
Average value of tailings from chlorination, . . . $2 per ton.
Two
Cord of wood at $1-40,
labourers at $1 -00,
......
COST OF ROASTING.
.
*
. . . .200
$0'70
Per ton of roasted ore, . . . .

$2 '70
10
15
chloride of lime at 3 cents,
Ibs. .....
COST OF CHLORINATING.
....
Ibs. commercial sulphuric acid at 2 cents,
Two labourers, J day at 90 cents,
. . . .'30
$0*30
Power, ....
Chloriuator, \ day at $2 '00,
Sulphuric acid for ferrous sulphate,

Wear and
'45
50
10
tear, . .
Superintendence, 05
Per ton of roasted ore, . . . . $T95

TOTAL COST.
Per ton of roasted ore, . . . . . . .
$4*65
Per ton of raw concentrates, . . . . . 3 '50
A which had been in use for five years at

lead-lined barrel
this mill showed no signs of wear, the abrasion of the lining by
the ore being evidently almost nil.
The Modern Barrel Process at the Golden Reward
3.
Chlorination Works, Deadwood, Dakota. Mr. John E.
Roth well, in exchanging his position at Deloro for that of
manager of these works, adopted a system differing considerably
from the above, as may be seen from the following description
which is based on his writings,* and on accounts furnished
*
Eng. and Mntj. Journ., Feb. 7, 1891, and Mineral Industry for 189$,
p. 233.
privately to the author by a late foreman of the mill and by

others:
The ore is crushed in a large Gates crusher and two sets of

Krom rolls, roasted partly in four Bruckner furnaces of 3 tons
capacity each, partly in a large White-Howell furnace, and
chlorinated in barrels of 3 and 4 tons capacity. After being
passed through the rock-breaker, which delivers no pieces larger
than 1 inch in diameter, the ore is dried by passing through
a revolving cylinder, 18 feet long and 5 feet in diameter, the
capacity of which is somewhat increased by being divided into
four longitudinal compartments, by the same number of parti-
tions, made of plate-iron, which terminate at a distance of 2 feet
from each end of the cylinder. The fire-place is arranged to
heat a large quantity of air, which is not entirely admitted
through the grate, but partly through channels by its side.
The revolving screens are hexagonal in shape, and are made
double with a coarse mesh inside the fine mesh, the object being
to protect the latter from undue wear. The mesh-frames are
slid into grooves in the screen-frame, and are interchangeable
and readily replaced when worn. The partially oxidised ores
("red ores"), coming from workings near the surface, are roasted
in the White-Howell furnace, while the undecomposed pyritic
ores (" blue ores "), derived from deeper levels, and containing
much more sulphur than the first-named class, are roasted in
Bruckner furnaces. The ore from the roasting furnaces is spread
out in a thin, furrowed layer on the cooling floors, and, when
partially cooled, it is sprinkled with water and raked over, the
water drying out.
The actual capacity of the barrels used at the Golden Reward
Mill is 3J tons of roasted ore per charge. The following dimen-
sions apply to a 5-ton barrel. This consists of a ^-inch steel
shell, 9 feet long and 5 feet in diameter inside, lined with (J-inch)
lead of 24 Ibs. to the square foot. The cast-iron heads are 2J
inches thick and heavily ribbed, and are inserted inside the shell
and bolted to it through a flange. A 3-inch iron rod, covered
with lead, passes through the trunnions and the entire length of
the barrel. The trunnions are 12 inches in diameter, and the
charging-holes are oval, 11 by 16 inches, the cast-iron covers swing-
ing off on levers when required. There are two charging-holes
placed in the cover-plate of the manhole. The chlorination
barrel is also the washing and leaching vessel; this is arranged
by placing a supporting diaphragm, for a filtering medium, to
form the chord of an arc of the circle of the barrel.
The filter or diaphragm, as it is called, is made of asbestos
cloth, resting on a framework consisting of oaken planks, each
11 inches wide and as long as the barrel, and 2 inches thick.
The area of the filter is nearly 30 square feet. In the planks
are grooves running longitudinally, each groove being J inch
wide and deep, and the same distance apart. Transverse grooves,
a little deeper than the longitudinal ones, are also cut at
intervals of from 4 to 6 inches, and j-inch auger holes are bored
through the planks from the bottom of the transverse grooves at
short intervals. These planks are supported on wooden cross-
pieces placed transversely to the barrel, which rest on longi-
tudinal strips bolted to the shell. The liquid passing through
the asbestos filter-cloth collects in the grooves, and is drained off
by the filter-holes. This oaken filter-plate, which was devised
by Mr. D. Dennes, costs little, and is very effective and durable.
Before his suggestion was adopted, an iron grating covered with
lead, which was burnt-on, had been used to support the filter-
cloth. In the grating on one barrel there were no less than
28,000 holes, through each of which the lead coating had to be
carried and burnt-on separately. The grating thus made with
great labour only lasted for a short time, as a faulty joint in the
lead in any one of these holes allowed the iron to be corroded, and
caused the grating to go to pieces. Boards with auger holes
were tried, but the area of the filter was thus restricted to the
sum of the areas of the apertures, and the speed of filtration
greatly reduced. With the oaken-plates, however, the filter-
cloth rests on the sharp ridges between the grooves, the surface
being almost entirely available for filtration. Of course, when
the pressure is applied, the cloth sags down into the grooves,
but this only increases the area available for filtration. On the
top of the corrugated plates is placed the filtering medium, an
open-woven asbestos cloth. It is nearly as coarse as the ordinary
gunny-sack, but the warp and woof are of much heavier thread.
Over this is placed an open wooden grating, and the whole is
held in place by cross-pieces, the ends of which rest under straps
bolted to the inside of the shell although the filter, when made
:
in this way, is rigidly held in place, it can be very easily and

quickly removed when the changing of the asbestos cloth becomes
necessary. The time occupied in changing the cloth is about
one hour and a-half, and, under ordinary conditions, it will last
for from 15 to 18 charges or from 2| to 3 days. The wear is due
to the scouring action of the pulp, from which the filter-cloth is
very imperfectly protected by the wooden grating. This is made
of 1^-inch wooden pieces, so that its apertures are 1^ inches
square.* The woodwork of the supporting plates and gratings
lasts much longer than the filter-cloths, but requires replacing at
somewhat frequent intervals. One carpenter is continually em-
ployed in making the woodwork for the filters of the 4 barrels,
and in fixing the new frames in position when necessary. The
lead lining of the barrels, in which several thousand charges have
*
At the Independence Mill, which stands near the Golden Reward Mill,
the asbestos cloth is protected by a secret method, and lasts over three
months.
294 THB METALLURGY OF GOLD.
been treated, show little signs of wear yet. Two valves on each
side of the shell of the barrel, above and below the filter, are for
the inlet and outlet of the wash-water and solution respectively.
The barrel is charged by first filling the space under the filter
with water, which at the same time is allowed to pass through
the filtering medium, and wash it then the required quantity of
;
water is put in above the filter. The sulphuric acid is then

poured into the water, through which it sinks in a mass to the
bottom, without mixing with it; the ore is then charged in, as
follows :The hoppers are furnished with 2 shoots, one for each
charging-hole. The ore is let fall through these shoots alter-
nately, the hole through which ore is not being passed serving
as an air-vent. Meanwhile the bleaching powder has been
weighed-out and placed in two small kegs. When the ore has
all been introduced into the barrel, a workman, stationed at each
charging-hole, hollows out a space in the surface of the dry ore

with his hands, and, emptying one of the kegs into the barrel,
closes up the charging-hole as quickly as possible. If all these
operations have been conducted rapidly without a hitch, there is
no immediate evolution of chlorine, but, if some time is suffered
to elapse after charging-in the ore, the acid liquid, thoroughly
stirred-up and mixed by the fall of the ore into it, gradually
rises through and wets the charge, and the bleaching powder,
fallingon ore which has been wetted with acid, gives off copious
fumes of chlorine before the cover-plate can be screwed on.
After the chlorination is complete viz., in about one and a-half
to two hours the barrel is stopped, so that the filter assumes a
horizontal position ;
the hose is attached to one of the outlet
pipes and conducts the solution to the settling tanks. A hose is
also attached to the inlet pipe, and water is pumped in under a
pressure seldom exceeding 40 Ibs. per square inch, and the
leaching commences. The air in the top part of the barrel is
compressed and forms an elastic cushion, which gives the wash-
water perfect freedom to circulate evenly over the whole surface
of the charge, and wash every portion of it thoroughly. By
washing in this manner, no chlorine is allowed to escape into
the building, as it is all absorbed by the water. At intervals
the leaching is suspended, and the barrel is again revolved for
a few minutes, so that its contents are thoroughly mixed-up
together again. In this way the formation of channels in the
ore are prevented, and perfect leaching is ensured. The wash-
water coming from the barrel is tested for gold with sulphu-
retted hydrogen.
The length of time required for the leaching varies with the
leaching quality of the ore treated charges having been leached
in forty minutes with a pressure of from 30 to 40 Ibs. per square
inch. With higher pressures the time can be materially shortened.
In order to facilitate the leaching of charges carrying an excess
THS BARREL PROCESS. 295
of dust or slime, the following method, now not in use, is

stated by Mr. Rothwell to have been formerly employed at
this mill. A
valve placed in the casting of the head, on a
level with the surface of the pulp, is opened just after the barrel
is stopped, and the dust and slimes which remain in suspension
are run into an outside washing filter-press, or to the slime-filter
where it can be treated separately, and the charge washed in the
usual way.*
The tailings are discharged into a car which will hold the
whole charge of ore and water, and then run out of the building ;
or, if water is abundant, they are discharged into a sluice, and
washed away. The filter-cloth is washed clean by a jet of water
under pressure directed successively to all parts of it. This water
is discharged by revolving the barrel.
" For
leaching purposes, the amount of water necessary to
wash a charge varies very little with the richness of the ore,
which goes to show the perfect leaching condition of the ore in
the barrel. The amount required is about 120 gallons per ton
over and above the quantity used in the barrel for chlorination,
which is about 100 gallons per ton. The water from the final
washings, which only contain small quantities of gold, is forced
up into an overhead tank, and used for the succeeding charge in
the barrel ; the quantity of solution to be precipitated is thus
reduced to about 120 gallons per ton of ore treated." f
The solution coming from the barrel, passes through a 2-inch
acid-proof hose-pipe into a large settling tank, one of a consider-
able number. These are round, wooden vats, lined with lead,
and are about 10 feet in diameter and 7 feet deep. When the
first tank is full, it overflows into the next one, which is placed
at a somewhat lower level, and this arrangement is repeated
through the whole series, the lowest vat overflowing into a
sump, whence the liquid, now nearly free ft-om suspended
particles of ore which have settled to the bottom of the vats,
is pumped up into the precipitating vats, situated above the
cooling floor near the top of the building. The force-pump

used for this purpose is perhaps the weakest point in the whole
plant as it is rapidly corroded by the acid liquid, and requires
frequent renewal and repairs.
At the Independence Mill, the slimes are separated by being
passed through slime-presses, similar in shape to a Johnson
filter-press. | Asbestos cloth is used for the partitions, with the
addition of a layer of Canton-flannel if the slimes are very
*
This device was first used by the late Mr. J. T. Blomfield, at the
Newbery-Vautin Chlorination Testing Works, in London, in 1889, and was
afterwards adopted by Mr. Hothwell in 1890.
t Eny. and Mng. Journ., Feb. 7, 1891, p. 166.
For description of Johnson's filter press, vide Stetefeldt's Lixiviation
of Silver Ores, p. 126.
finely divided. The same pressure which affects the leaching in

the barrel forces the solution through the slime-presses. When
these are full of slimes, they are washed out by water forced in
the opposite direction to that in which the solution had moved.
Slime-niters were formerly in use at the Golden Reward Mill
also ; they consisted of lead-lined cast-iron cylinders of 18 inches
deep and 30 inches in diameter, in which were arranged filter-
beds of sand and asbestos cloth. The chief difficulty encountered
in the use of these slime-filters is stated to have been that the
solution was squirted out in all directions in small jets, and part
of the valuable liquid thus lost. At the Independence Mill, the
filter-press is placed above a large, wide, shallow, lead tank, while
it is completely surrounded by a leaden hood, against which all
the jets of liquid strike and then run down into the tank.
At the Golden Reward Mill there are two precipitating tanks
for each barrel. They are 6J feet in diameter and 10 feet deep,
are fitted with covers, and are constructed of wood and lined
with lead of 8 Ibs. to the square foot. The filter press, which is
situated on the ground-floor, over 25 feet below the precipitating
vats, has twelve chambers of 19 inches square, and has a filtering
area of 57 square feet. Two lead pipes connect the precipitating
tanks with the press, one leading from the bottom of the tank
and the other from a point in the side about 4 inches above the
bottom. The precipitating gas generators are cast-iron cylinders
of the same size as the slime-filters ; only the sulphuretted hy-
drogen generator is lined with lead.
Before the precipitating tank is full of solution, the sulphur-
ous acid gas generator is started, and the gas forced through a
2-inch leaden pipe leading to the bottom of the tank. The gas is
made from sulphur burned in the generator, with the aid of a
current of air forced in at the bottom and deflected over the sur-
face of the burning mass. The mixture of sulphur dioxide and
air, which is present in excess, is passed into the solution, where
the former is absorbed, while the air, bubbling through, serves to
stir the liquid. When the free chlorine has nearly all been con-
verted to hydrochloric acid, sulphuretted hydrogen, generated
from sulphide of iron and dilute sulphuric acid, is forced in,
together with a large excess of air, by which the liquid is kept
in a state of agitation, and the free chlorine expelled, thus
effecting a saving of SH 2
. The precipitate is floccnlent, and
settles quickly.
After the precipitation is complete the "tank is allowed to
stand for two or three hours, when it has settled enough to
draw the supernatant liquor off through a filter-press by the
pipes in the side of the tank. There is little danger of precipi-
tating arsenic or antimony that may be in the solution when it
is worked cold, as they do not commence to come down till some
time after the gold has precipitated and collected. Of course,-
any copper or lead in solution will be precipitated with the gold,

but small quantities only are present, they can be removed
if
subsequently. The loss in gold is considerably less if the preci-

in the tanks, and a clean-up
pitate is allowed to accumulate
made after six to ten precipitations, than if it were filtered
through a press and collected after every precipitation. Continu-
ous filtration causes the filters to be coated and clogged with the
sulphides, and retards the operation unless high pressure is used,
which is sure to increase the loss, while a similar result attends
the handling of a large number of filter-cloths."* Roth well
has stated more recen ly that gold can be fractionally precipi-
tated from solutions containing copper if free acid is present.
The sulphide cakes in the presses are compressed by blowing air
through them, and are turned out in the form of lumps, which
break up into powder when they are touched ; they are then
dried, roasted, and fused with borax, nitre, carbonate of soda
and sand.
The amount of precipitants required to precipitate a tank of
2,500 gallons of solution are sulphur, 2 Ibs. ; iron sulphide,
4 to 5 Ibs.; sulphuric acid, 16 Ibs.; water, 9 gallons. The
chemicals required for chlorination are bleaching powder, 8
Ibs., and sulphuric acid, 15 Ibs. per ton of roasted ore. The
power is furnished by an engine of 125 H.P., consuming six or
seven cords of wood, or about 4 tons of bituminous coal. Steam
is also required for the air-compressor, for steam-pumps, and for
heating the building. No salt is used in roasting, as the ore

treated contains little copper, &c., and no attempt is made to
extract the silver ; the fuel used (chiefly pine-wood) amounts to
a little more than that needed for power. The men employed
on the mill number about forty. " One man of ordinary intelli-
gence and a helper are able to take care of three barrels that
is, to look after the charging, leaching, and discharging. If the
tailings are sluiced out they can also attend to that ; but where
they have to be trammed out, one more man is necessary."!
There are four barrels, each of which treats 6 charges in 24
hours, or a total of over 90 tons per day. The cost of treatment
is given below. When the ore is crushed through an 8-mesh
screen, the tailings usually contain about 3J dwts. of gold per
ton, or 20 per cent, of the total contents ; but it is stated that 92
per cent, was being extracted in 1894. J
The disad \ antages of the process used are due to the construc-
tion, but do not seriously interfere with the successful working of
the barrel; they are, among others, the amount of space taken
up by the filter and the portion of the barrel under the filter,
and the fact that whenever the barrel is charged and running, it
is not in a These disadvantages
perfectly balanced condition.
could be minimised, according to Mr. Rothwell, by using a filter
*
Luc. cit. t Hid. $ Eng. and Mng. Journ., Jan. 27, 1894.
placed close to the shell, and separated from it by a space only

just enough to allow of free circulation, and reaching to the same
height on the sides as the horizontal filter; then, by using
compressed air to displace the solution and wash-water, an
equally good result could be obtained. The mechanical diffi-
culties in constructing a curved filter-bed, however, prevent
its introduction.
According to Mr. Demies, it appears that one of the disadvan-

tages in the use of the diaphragm is that it is difficult to keep it
in place. The ore and water wash to and fro through the filter-
cloth, and there is a strong tendency for the latter to become
displaced. If one plank in the somewhat complex framework,
by which it is held in place, becomes loosened, the whole soon
gives way as the barrel revolves, and the mishap is not detected
until leaching begins, when ore and water come out together.
The asbestos cloth is expensive, and only lasts for about three
days, and all repairs must be done by a man getting inside the
barrel, owing to its construction. It has been more recently
suggested that the asbestos cloth could be replaced without
difficulty by a sand filter.
The cost of barrel chlorination in 1891, at the Golden Reward
*
Mill, is given by Mr. John E. Rothwell, as follows :
July. August. September.

Amount of ore treated, 1,430 tons. 1,195 tons. 1,513 tons.
treated per day, 46 '1 38 '5 50 '4
Cost per ton
Milling, . $1-49 $1-40 $1-44
Roasting, 1-53 1-35 1-37
Chlorination, 1-76 167 1-65
Office, salaries, 0-40 0-47 0-37
Construction and
repairs, . 0*36 093 0'21
$5-54 $5-82 $5-04
The total cost given above includes all the working expenses,
but is exclusive of interest on capital, taxes, insurance and
amortisation (i.e., extinction of capital by a sinking fund, which
covers the depreciation of plant, <fec.)
The roasting was at that time done in Bruckner cylinders, and
its cost was afterwards reduced to about 70 cents per ton by the
adoption of a modified White-Howell furnace, and in January,

1894, the total cost of treatment per ton was stated to be only
$3-77.
The amount of chemicals used in the chlorination and precipi-
tation departments, in 1891, is given on the next page.f
*
Eng. and Mng. Journ., Mar. 25, 1893, p. 269. t Loc. cit.
of manipulation; the men's eyes only were slightly affected.

From 6 to 13 Ibs. of bromine were added to the charge of 4 tons,
the amount being adjusted so that a very slight excess of free
bromine remained at the end of the operation. Chlorine
generated from sulphuric acid, and bleaching powder was used
for several months before the introduction of bromine, but the
cost was greater and the tailings only dwt. poorer in the former
case. Moreover, the filter cloths lasted longer, the health of
the workmen suffered less injury, and the gold precipitate
was purer when bromine was used. The barrel was revolved for
Fig. 55.
from thirty minutes to one and a-half hours, and then discharged
into the leaching vat below, shown at Fig. 55. This is 7J
feet in diameter and 3 feet deep, made of cast-iron lined with
lead, and has a strongly ribbed cover, so that it is capable of
withstanding an internal pressure of 100 Ibs. to the square inch.
Internally, it tapers slightly upwards. The sides of the vat do
not rest on the bottom, but are supported by columns direct from
the floor. The bottom is not connected with the rest, but is
supported by a hydraulic ram, shown in the figure. On this
true bottom there is a filter-bed and false bottom. The filter-
cloth consisted of gunny-sacking. The vat was worked as
follows: When a charge was about to be introduced, the bottom
was raised by the ram and pressed tightly against the sides of
the vat, a good joint being made by a thick, rubber ring. The
hydraulic pressure was more than enough to overcome the air
pressure subsequently applied. The charge was then introduced,
the cover replaced and fastened down tightly by screw-bolts, and
air was pumped into the vat above the surface of the charge.
"Water was also introduced by means of a lead pipe extending
round the vat at about the surface of the charge. This pipe was
pierced with a number of small holes, by means of which jets of
water were thrown upon all parts of the ore surface. The air
pressure, which never exceeded 60 Ibs. to the square inch, forced
the wash-water through the charge very quickly. The solution
was of a strong ruby-red colour, due to the presence of free
bromine, and of the bromide of gold, the colour of which in
solution is very intense. There was no need to test the issuing
liquid, as its colour was found to be a perfectguideto an experienced
eye. When it had become colourless, the water supply was shut
off and the charge dried by pumping air through it. This was
necessary, as the company was not allowed to sluice the tailings
into the river. When no more water could be driven out, the
air-pressure was let off, and the bottom of the vat lowered until
the top of the filter-bed was just clear of the sides of the vat.
The bottom was then drawn sideways from below the vat by
means of a second hydraulic ram, and the charge fell into a large
ore-car below, having a capacity of 4 tons. This ore-car then
ran by gravitation to the dump, and was emptied and drawn
back by a wire rope winding on a drum actuated by steam
power. It is stated that the leaching occupied only twenty
minutes, and the vat was used to leach the charges from the two
barrels, which discharged into it alternately. The vat was so
rapid that it was always waiting for the barrels, and 100 tons
per day could be leached by it. It was designed by Mr. D.
Dennes, and constructed by Messrs. Eraser & Chalmers, and,
though of high initial cost, was found to be of such great
efficiency that its introduction was thought to be amply justified.
The liquid was forced into large, lead-lined wooden precipita-
tion vats, each 20 feet x 10 feet x 6i feet, which were placed
outside the mill, buried under 2 feet of manure to keep them
warm in winter. The precipitation was effected by the successive
use of sulphur dioxide and sulphuretted hydrogen, applied in
exactly the same manner as at the Golden Reward Mill. The
collection of the sulphides was also effected in a similar manner.
When dried, the sulphides were put, together with the necessary
amount of borax, into a small barrel, revolving by machinery,
and were mixed thoroughly. The mixture was then transferred
by a scoop to a red-hot clay crucible (size No. 100) in the furnace,
additions being made at intervals until the crucible was full
of bullion and molten slag. The latter was very rich and.
fullof shots of metal. It was stored and eventually sold to the

smelters, but owing to the difficulty in sampling it, it would have
been better to fuse it with lead and to sell the latter.
The air-compressor used was 12 inches in diameter, and had
an 18-inch stroke ; its maximum velocity was eighty revolutions
per minute. A
large air receiver, 16 feet long and 4 feet in
diameter, was used between the pump and the leaching vat.
The maximum air pressure in the receiver was 80 Ibs. per square
inch. The dimensions of the bullion smelting furnace were as
follows Depth 3 feet 3 inches. Inside diameter at bottom 2
:
feet 8 inches, at top 2 feet 2 inches. The flue measured 8 inches

by 12 inches, and was situated 10 inches below the top of the
fire-box.
From 60 to 75 per cent, of the gold in the ore was extracted,
the tailings usually containing over 5 dwts. of gold. The reason
for this unsatisfactory result is said by Dr. L. D. Godshall (Eng.
and Mng. Journ., Jan. 6, 1894) to lie in the coarseness of the
crushing, a 10-mesh sieve being used. When a 30-mesh sieve
was used, only about 2J dwts. were left in the tailings, and the
product was easily leached. It seems possible that inefficient
leaching, due to the formation of channels in the ore, may have
been another cause of loss.
Cost Of Treatment. The cost of treatment of the ore by chlorination
at the Rapid City Mill for 40 tons per day, everything running double
shift, was as follows :
Crushing Mill.
'2 men (1 each shift), running ore from bins to Blake, at $2 -00 tf day = $4 '00
2 ,, ,, attending to the Blake, 2 '00 400
2 Rolls, 3-00 6-00
2 ,, ,, oiling in the mill, 2-50 500
Coal used by rotary dryer, 4 ton, at . 5-75 2-87
Oil used by the mill, 2 galls., at 40 '80
f of the whole power used (see Power Account), 19-65
Consumption of roll-shells and Blake jaws, 4-80
$47-12
or $1-177 per ton.
RoastAng.
2 men (1 each
attending to White-Howells, at
shift), $3 '00 $ day = $6 '00
2 ,, helping at 2-00 ,, 4-UO
2 ,, ,, cooling ore from ,, at 2-00 4-CO
4 ,, (2 each shift), wheeling ore to barrels, at 2-00 8-00
2 ,, (1 each shift), getting wood for roasters, at 2-00 400
Cordwood consumed per 24 hours by 2 roasters = 6 j cords,
at $3-50 per cord, 2275
Oil for this department, including lighting, and broken
T*t
lamp chimneys,
of the whole engine and boiler power, ... . ,,
1 '05
l'9o
|51'76
or $1-294 per toa.
Chlorinating and Leaching.

2 men (1 each shift), running the barrels, at . .
$3 '09 $ day = $6 -03
2 leaching apparatus, at 3 '00 6 '00
2 ,, ,, ,, tailings out on to
dump, at 2-00 ,, 4'00
Gunny-sacking for filtering table of leaching apparatus,
of the whole boiler and engine power,
Chloride of liine, 14 Ibs. per too, or 560 Ibs. per day (at 24
... .
.
,,
,,7*86
"10
cents), . = 14-00
Sulphuric acid, 24 Ibs. per ton, or 960 Ibs. per day (at 3
cents),
011 for this department, at 40 cents per day, ... =
=
33-60
'40
$71-96
or $1-780 per ton.
Precipitating and Recovering.

12 men
(1 shift), precipitating and handling solu-
each
=
tion, and drying aud emptying filter presses, at
Sulphur for generating S0 2 and H 2 S, 20 Ibs.,
$3'00 $ day
Flannel cloth for filter presses, $10 '00 per clean-up, or $2*00
,, ... . $6'00
'40
per day (clean-up say every 5 days),

TV of the whole boiler and engine power,
Borax glass, caroonate of soda, crucibles, and coke,
...
..
.
. .
.
,,
,,1'96
,,
2 '00
4-50
$14-86
or 37 '1 cents per ton.
General Expenses Power.

2 Engineers, each $3 -25 = $6 -50
2 Firemen, 2-50 = 500
Coal, 7 tons at 2'75 = 19-25
Oil, = -70
$31-45
or 78 '5 cents per ton.
(This is included in the other items.)
Offices, <kc,
1Assayer, at $5'00
1 Assistant, at 2'00
1 Clerk, at 2'00
1 Oflice boy, at . 1 -00
$10-00
1 Blacksmith, . . . . . . . . . $3-50
1 Machinist, 3*50
1 Helper, ., 200
Blacksmith's coal and iron per day, average,
Oil, waste, and general stores consuaied per day, .... 175
2'00
$12-75
= $22-75 or 56 '9 cents per ton.
'301 THE METALLURGY OF GOLD.
Cost of bromination in place of chlorination as follows 1'5 to 3'25 Ibs.

:
of bromine at 17 cents per Ib. for each ton of ore = '25'5 to 55 '25 cents
per ton, labour and other things being equal.
Total cost of chlorination as above
Crushing, $1 '177 per ton.
Roasting,
Chlorinating and leaching,
Precipitating, &c.,
.... 1"294
1'780
'371
.,
,,
General expenses, '569 ,,
$5-191
Superintendence and amortisation are not included in this total.

It will be noticed that a saving of at least 60 cents per ton is effected by
substituting bromine for chlorine.
The Future of Chlorination. At the present day chlori-

nation is the most favoured method of treating auriferous
concentrates obtained from the tailings of stamp batteries,
except in districts where smelting works, producing mattes or
base bullion, exist. It is probable that, in general, such con-
centrates may "be more economically dealt with by the cyanide
process, but it seems doubtful if the high percentage of extrac-
tion attained by the use of chlorine can be equalled by the
newer method without roasting. During the last few years
the chief extension given to chlorination has been due to the
perfecting of the barrel process ; by its use certain virgin ores,
especially those of Mount Morgan, of Carolina, and of South
Dakota, have been successfully treated. The latest methods
introduced in Dakota, especially at the Golden Reward Mill,
certainly offer great advant;ges over the original practice in
Carolina and Queensland, and improvements in detail will pro-
bably continue to be made. For example, the recent increase
in size of the barrel in Western America from one containing
3 tons of ore to one holding 10 tons is of considerable import-
ance. Nevertheless, a doubt may be expressed whether the
complicated and costly machinery necessitated by these methods
will not stand in the way of their adoption in other gold-fields.
It is at least possible that American pioneers of chlorination
are working in the wrong direction. Dr. Munktell appears to
have shown that a weak solution of chlorine in water, acting on
ores not subjected to agitation, is, if sufficient time is allowed,
almost as efficacious in dissolving gold as a strong solution com-
bined with agitation, or as chlorine gas acting on moistened ore.
If this should prove to be the case with most auriferous pyrites,
a plan of operations in chlorination, much simpler than those
hitherto in vogue, can be pursued. The roasted ore might be
charged into large round tanks of wood or masonry holding
hundreds of tons of material, such as those now in use with the
cyanide process in the Transvaal. It would then be sufficient
THE BARREL PROCESS. -305
to leach with a weak solution of chlorine, allowing it to run

through by gravity, and, as it issues at the bottom, continually
raising it by means of a pump (preferably of the injector type).
The solution couTd then be poured on the surface of the charge
again, strengthened by the addition of chemicals if necessary,
and the whole operation repeated until the dissolution of the
gold is complete. There can be no doubt as to the cheapness
of the plant required in this case, and it is to be remembered
that similar methods have been selected as most advantageous
in the old hyposulphite, the Russell and the cyanide processes,
in all of which the progress that has been made is in the direc-
tion of greater simplicity. In consideration of these facts, it
seems likely that, in spite of the great technical success of the
operations in Dakota, it is in the results of the Munktell pro-
cess in Central Europe that indications may be found of a
possible extension of chlorination to the treatment of large
quantities of low-grade refractory ores.
As an instance of the direction in which the industry is pro-
gressing, the operations in Queensland may be quoted.* Here
a few years ago barrels were almost exclusively used in chlorin-
ation. At Mount Morgan, the largest chlorination mill in the
world, wooden barrels were used in which chlorine was generated
by bleaching powder and acid. However, both at Mount
Morgan and generally throughout Queensland, a return to the
vat process has been made, with a view to reduce the working
costs. The vats are, however, made larger, those at Mount
Morgan holding 25 tons of ore. The chlorine is now applied
there as chlorine water, and it is claimed that the consumption
of the solvent when used in this form can be better controlled
and kept within the absolute requirements for the extraction
of the gold. The gas is generated in stills by means of man-
ganese dioxide, salt and sulphuric acid, and the use of chloride
of lime has been completely abandoned. The chlorine is ab-
sorbed by water in scrubbing towers. The ore contains 1-66 oz.
of gold per ton, and 95 per cent, of this is extracted at a cost
of about 1 2s. per ton, the amount treated being about 200 tons
of ore per day. It will be observed that all the American
methods have now been abandoned at this mill, and the simple
method of procedure now adopted seems to have much to recom-
mend it.
*
See article by E. A. Weinberg in Report for 1894 on Queensland Mines.
20
CHAPTER XV.
THE CYANIDE PROCESS.

The Cyanide Process, like most others, gradually grew up
from small beginnings. The power possessed by potassic cyanide
solutions of dissolving metallic gold and silver has long been
known, as is pointed out, p. 333, but it was formerly believed
that the use of an electric current in conjunction was needed to
quicken the action, which was otherwise too slow to be of any
practical value. The use of cyanide of potassium in the stamp
battery has already been referred to, p. 124. It is probable
that its effect in dissolving gold was entirely overlooked when it
was being used in this way, the effect which it was desired to
produce being the removal or prevention of formation of the
crusts of carbonate of copper on the plates. In California, prior
to the year 1870, gold ores were sometimes digested in tubs
with water, to which a lump of cyanide had been added ; after
digestion, the ores were amalgamated by stirring with mercury,
to which a little sodium-amalgam was sometimes added. It is
obvious that, when no sodium was present, only the undissolved
gold was amalgamated, the dissolved gold being lost, not being
precipitated by mercury. These operations, however, can hardly
be said to foreshadow the cyanide leaching process, as they were
conducted in ignorance of the principles on which it is based,
and partly in defiance of them.
The first attempt at the direct extraction of gold by the use of
cyanide was made by J. H. Rae, who took out a patent in the
United States in 1867 (No. 61,866) for a process dependent on
the removal of gold and silver from their ores by the combined
action of a "current of electricity and of suitable solvents or
chemicals," such, for instance, as cyanide of potassium, the gold
being simultaneously precipitated on copper plates by the electric
current. This patent has now lapsed, and, although other
United States patents claiming the use of cyanide in the treat-
ment of gold ores were issued in 1880 and 1881, no real progress
seems to have been made.
In 1885, J. \V. Simpson, of New Jersey, obtained a patent in
the United States (No. 323,222) which was of greater interest
He proposed to treat ores containing gold, silver and copper by
a solution containing 3*0 per cent, of potassium cyanide and
THE CYANIDE PROCESS. 307
0-19 per cent, of ammonium carbonate. Copper was to be dis-

solved at the same time as the gold ; if silver was present also,
an addition of common salt was made. The inventor appears
to have believed that, by using carbonate of ammonia, the
necessity was obviated of employing an electric current, in con-
junction with cyanide of potassium, in order to dissolve the gold.
Of course the carbonate of ammonia would have no effect on the
amount of gold dissolved, whilst its action in inducing the
hydrolytic decomposition of cyanide, described in the sequel, is
a decided disadvantage. The precipitation of gold was effected
"
by a piece or plate of zinc." The process does not appear to
have been tried on a large scale, nor did any results follow from
the later experiments made on the subject by Endlich and
Muhlenberg,* and by Louis Janin, Jr.,f in America.
In 1886, however, a series of experiments on wet processes
of treating gold ores were begun by J. S. Mac Arthur and R. W.
and Wm. Forrest in Glasgow, and it was entirely owing to their
energy and skill that cyanide of potassium was successfully
applied in practice to the treatment of gold ores. Patents were
taken out by them in Great Britain in 1887 (No. 14,174), in the
United States in 1889 (No. 403,202), and in other parts of the
world. Their process consists essentially in attacking gold and
silver ores by dilute solutions containing less than one per cent,
of KCy, caustic soda or lime being added to ores rendered acid
by the oxidation of pyrites, and then in precipitating the pre-
cious metals by zinc shavings. This process has now passed
into use in all parts of the world. Its success is complete on
many ores, and its extension will probably be very great, partly
at the expense of the chlorination process. The chief advantage
of the cyanide process over the chlorination process is that
roasting is not necessary, even if sulphides are present; this
is a most important point in the treatment of low-grade ores,
especially in places where fuel and labour are costly. More-

over, silver as well as gold is extracted by cyanide solutions.
Messrs. MacArthur and Forrest, in the course of their investiga-
tions on wet methods, became dissatisfied with chlorine as a
solvent, owing to its energetic and preferential action on sul-
phides of base metals and other bodies, and its inapplicability
to ores containing silver. They desired to find a solvent which
would exercise a selective action in favour of the precious
metals, instead of other substances. With this object in view,
they experimented with a number of solvents (such as ferric
bromide, ferric chloride, <kc.), and finally decided that potassium
cyanide possessed advantages over all other substances. These
chemists lay especial stress on the advantages to be obtained by
using very dilute solutions, as they believe that, in this way,
*
Enrj. and Mng. Jorn., Jan. 17, 1891, p. 86.
t Ibid., Dec. 29, 1890.
308 THE METALLURGY OP OOLD.
the waste of chemicals, caused by reactions with the base

minerals present, is almost entirely avoided, while the dissolu-
tion of gold and silver is not retarded.
The MacArthur-Forrest Process. The following general
description of the working of the MacArthur-Forrest cyanide
process is based partly on accounts furnished by Mr. J. S.
MacArthur, to whom the author is much indebted for a large
amount of information, and partly on the actual work being
done in South Africa, in America, and elsewhere. The
mechanical treatment of the ore viz., crushing, charging into
vats, leaching, and discharging the tailings is based on the
same principles, and could be effected by the same general

methods as are adopted in chlorination or in any other leaching
process. Many details will, therefore, be omitted as having
been already covered in the course of the descriptions given in
Chapters xi. to xiv.
The process comprises four distinct operations, viz. :
(1) Pre-
(2) Solution of the gold
in
paration of the ore for treatment.
potassium cyanide. (3) Precipitation of the dissolved gold by
means of zinc ; and (4) Conversion of the precipitated gold into
bullion by fusion.
1. Preparation of the Ore. The ore is crushed to a suitable
degree of fineness, which depends partly on the condition of the
gold and partly on the nature of the gangue. In the Transvaal,
where tailings are treated, the ore is crushed in stamp batteries
previous to amalgamation, the screens being usually about
25
or 30-inesh. In this case, sizing is performed to some extent
during the operation of catching the tailings, the slimes in part
passing off suspended in the water used in the stamp-mill.
Nevertheless, masses of slimes occasionally accumulate and
resist the passage of the cyanide solution or retain the dissolved
gold. If mixed with sand, the slimes are less harmful, but
they tend to separate from the sand again when the leaching
begins.
At the Mercur Gold Mine, Fairfield, Utah, the ore is a silicious
limestone,* carrying magnetic oxide of iron, and in places con-
taining much silt, one of the proofs of its aqueous origin. Any
attempt at fine crushing results in the sliming of the greater
portion, and renders leaching impossible.
The course of pro-
cedure adopted is consequently as follows : The ore is passed
through a Dodge rock-breaker, set to break the ore to a
maximum size of 1 inch ; it is then reduced to J inch by a
pair of 20-inch corrugated rolls ; thence it is passed over
a
i-inch grizzly to a similar pair of 12-inch rolls, which reduce it
to a maximum size of ^ inch. After this treatment, 61 per cent,
remains on a No. 12 screen, while 26 per cent, passes through a
No. 30 screen, this part consisting almost entirely of impalpable
*
Eng. and Mny. Journ., Nov. 5, 1892.
powder. This material can be leached, although somewhat more

slowly than is desirable.
As a rule, ores are crushed so as to pass through screens
varying from sixteen to thirty holes to the linear inch ; in the
product thus obtained there is usually from 90 to 95 per cent,
of the gold which can be reached by the solution, and, if rolls
have been used for crushing, it is not difficult to leach. Stete-
feldt gives an instance* in which leaching by hyposulphite
solutions was successfully performed on the Bremen tailings, at
Silver City, New Mexico, when the average fineness was such
that 87 '8 per cent, of the ore would pass through a screen with
150 holes to the linear inch. The rate of leaching in this case
was about inch per hour, with a vacuum of 14 inches of
mercury, a degree of slowness which would militate strongly
against the success of any works. For the separation of slimes
from crushed ore, see p. 311.
2. Solution of the Gold. This is effected in wooden, iron,
or cement vats, with false bottoms similar in construction to
those used in the Plattner process (see p. 250), or to those
employed in the Russell process described by Stetefeldt.f The
vats should be round, and may be discharged by sluicing through
a door in the side or bottom, or by shovelling out. Those in
use in Utah have sheet-iron sides with a wooden bottom, and
have a capacity of 14 tons of ore. In South Africa the vats are
often of 2,000 cubic feet capacity, and hold 75 tons, but the
Langlaao;te tanks contain 450 tons each, and the Simmer and
Jack vats 600 tons each.
The use of large vats in preference to small ones is to be
strongly recommended, as the attention and superintendence
needed by a 10-ton vat costs as much as for a 600-ton vat.
Stetefeldt considers j that constructive difficulties limit the
diameter of wooden vats to 16 feet, but the Transvaal vats of
40 feet diameter give no trouble. The inside depth, Stetefeldt
continues, should not be more than 5 feet if the tailings are
shovelled out, and 7 feet if they are removed by sluicing. The
depth has usually little influence on the rate of leaching, as the
" head" of of
greater liquid tends to compensate for the increase
compactness at the bottom of the vat. At Cusi, in treating
silver ores by the Russell process, it was found that the rate of
leaching was greater for a charge 6 feet deep than for one
which
was only 2 feet deep. In South Africa the vats are often 10 feet
deep.
In calculating the capacity of a leaching vat, the volume of a
ton of raw ore may be taken at from 22 to 28 cubic feet when
dry, and from 20 to 26 cubic feet when wetted down.
The false bottom is usually a wooden framework, constructed
* t Ibid., pp. 114-119.
Lixiviotion of Silver Ores, p. 149.
$lbid., p. 114. Ibid., p. 149.
of boards pierced with numerous auger holes, or of wooden slats

crossing each other and covered by canvas or cocoa-nut matting,
which are not i-apidly destroyed by the solution, as they are
in the chlorination process. Below this a layer of coarse sand
and pebbles, through which the solution percolates, is sometimes
placed.
An excellent false bottom recommended by the Cassel Gold
Extracting Company, the owners of the MacArthur- Forest
patents, consists of angle-irons with the angles uppermost, and
the free edges nearly touching. The irons are supported on
wooden slats 1 inch thick, which give a space ot that depth
below the filter-bed, in which the liquid may accumulate. The
triangular channels between the irons are tilled with sand.
When the vat is being cleared out, the workmen's shovels slide
along the top of the angles and leave the filter-bed undisturbed.
Triangular pieces of wood may be used instead of angle-iron.
Thick canvas duck, resting on matting, forms a trustworthy
filter-cloth.
As in the case of chlorination, a layer of perforated glazed
porcelain tiles to support the filter-cloth is better than wood, as
no absorption of the solution takes place when the former is
used. If wood is used it is covered by a coating of paraffin
paint, or by a mixture of asphaltum and coal-tar. It was found
by Mr. Julian of the Salisbury Cyanide Works, South Africa,
that pine wood, lying 34 hours in a solution containing 0'3 per
cent, of cyanide reduced the strength to 0-005 per cent., while
glass reduced it to 0-2 per cent, and cement to 0'24 per cent.
Cement tanks have been used at the Langlaagte Estate since
May, 1892.
An iron pipe communicates with the space below the false
bottom, and conveys the liquid to the pumps or to the zinc
boxes. The solution does not attack wood or iron brass and
:
bronzes are attacked and corroded rapidly.

The crushed ore is conveyed in cars pushed by hand, running
on an overhead railway, and emptied into the vats by tipping.
The supports of the railway must not touch the sides or any
part of the vats or their supports, as the jarring caused by the
running of the cars makes the ore settle down, and so renders
the leaching more difficult and tedious.
At the Langlaagte Works the tailings from the stamp batteries
are hauled for 600 yards up an incline of 1 in 15 to the tramway
above the vats, at a cost of 2d. per ton. At the Salisbury
Cyanide plant, a tailings wheel lifts the discharge from the plates
into a flume, which carries it to a hydraulic separator, where the
slimes are separated from the coarse tailings. The latter are
then run directly into the cyanide vats, thus saving handling;
this method of direct filling is now adopted in several works on
the Rand, but the separation of the sliines from the sand is not
complete, and this has a prejudicial effect on the percentage

of gold extracted. The separation of the slimes is beneficial in
expediting the filtration in the vats.
A widely-adc pte 1 form of slime separator is that known as
Butters & Meiu's collecting vat. It consists of a large round
vat, 20 to 30 feet in diameter and 20 feet deep. The tailings
from the stamp battery are run direct into a revolving distri-
butor, which spreads the pulp evenly over the surface of the
water with which the vat is filled. The sand settles to the
bottom and accumulates there, while the slimes overflow at the
top. When the accumulation of sand in the vat reaches to
within 5 feet of the top, the water is drained off and the ore
discharged by shovelling through an aperture in the bottom into
ore-cars below. This sand is now in a suitable condition for
leaching, containing only small quantities of slimes. At the
Jumpers mine the pulp collected in this way is treated directly
without removal. At the May Consolidated and the Croesus
mills, where the method of "double treatment" of the tailings
is employed, the first solution of cyanide is directly applied to
these vats, and, after draining, the pulp, wetted with cyanide
solution, is transferred to the second treatment vats (Hatch and
Chalmers). As already stated, the usual method is to remove
the collected sands to the cyanide vats for treatment.
The vats are filled to within a few inches of the top, and the
charge is levelled by means of hoes. The amount of ore charged-
in is such that, after the solutions have been applied, the surface
of the charge may stand at about 12 inches below the rim. In
levelling the ore, the labourer must not step into the vat or
forcibly press down the ore, as irregular filtering is produced by
this cause. The shrinkage of the charge on the addition of
liquid is from 10 to 18 per cent. The ore should be charged-in
as dry as possible, but a few per cent, of moisture makes very
little difference to the subsequent leaching.
The charge is now ready for lixiviation. If soluble salts are
present in it, a preliminary washing with water is required,
while, if decomposed pyrites are present, treatment with an
alkaline solution is also necessary, for reasons to be explained
when the chemistry of the process is considered. In each case,
from 7 to 9 cubic feet of liquid per ton of ore are added and
allowed to remain on the charge for some hours before being
displaced. The liquids are run on to the top of the charge
through pipes of 2 to 4 inches diameter, and are
distributed
a box with a bottom, by which the force of
by large perforated
the discharge is broken. The orifice below the filter-bed is
closed before beginning to add the liquids, and not opened until
the solution stands about 3 inches above the surface of the
ore, when the leaching is begun. After the caustic alkali (if it
in contact with the ore for
is
used) has been allowed to remain
a sufficient length of time, it is displaced by wash-water, which

is added above, while the alkaline solution is allowed to run out
at the bottom. A dilute solution of cyanide is then run in, the
stopcock below the filter-bed being closed as soon as all the
wash-water has been displaced. In leaching, the charge is
always kept completely submerged, and the surface of the liquid
is never allowed to sink below the surface of the ope for reasons
which will be given subsequently.
The cyanide solution is often allowed to stand in contact with
the ore for from 12 to 24 hours undisturbed before the stopcock
below the filter-bed is opened and the solution drawn off. It is
usually conveyed at once to the zinc boxes, but it was formerly
preferred either to raise it and pass it through the charge again
(circulation method), or to transfer it to a second or even a third
charged vat before precipitating the gold. The advantage of these
" circulation" and "transference" methods is that the solutions
become much richer in gold than if they are only allowed to per-
colate through a single charge of ore, and consequently they give
a cleaner deposit on the zinc with much less consumption of
cyanide, the volume of solution passing through the precipitation
boxes being less. At the Mercur Mine the circulation is kept
up for from 24 to 240 hours, according to the speed of leaching,
the usual time being about sixty hours. At the Robinson Mine
20 tons of solution cover the ore in a 75-ton vat, and are con-
tinually pumped backed into the same vat for thirty-six hours,
and then passed to the zinc boxes.
The " Simmer and Jack " works, South Africa, are shown in
Fig. 55a.* Here it was necessary to raise the tailings, and this
was done by hauling them up inclined platforms carried on
trestles. A double gangway runs over the vats for filling.
The leaching vats are the largest on the Rand gold-fields, each
holding 600 tons, being 40 feet in diameter with 14-foot staves,
the staves, as well as the bottom, being of 9-inch by 3-inch
material. There are eight discharge doors, 16 inches in diameter,
in the bottom of each vat. The three solution vats are 26 feet
in diameter, with 15-foot staves, and the liquid is returned to
them from the sumps by three 2-inch centrifugal pumps.
It has been thought desirable to stir the solution and ore
together, in certain cases, but the loss by decomposition of the
cyanide thus caused, and the cost of the power, are usually
more important than the additional amount of gold recovered.
The Cassel Gold Extracting Company recommend the following
method of procedure when agitation is used The solution-vat
:
or agitator (A, Fig. 56) is 4 feet in diameter and 4J feet deep.

This is filled to a depth of from 18 to 20 inches with the solution,
from the vat C, and the mixer-shaft set going at about sixty
*
Reproduced by permission from the paper by Butters and Smart in
the Proceedings of the Inst. of Civil Engineers, vol. cxx., pt. ii., 1895.
revolutions per minute. The ore (li tons, or double the weight
of the liquid) is then added gradually from a hopper, B, overhead,
in which it has been stored, and agitation is continued until the
gold is dissolved. The pulp is then discharged through a 2-inch
iron pipe into D, the large filtering vat, which is 12 feet in
diameter and 4 feet deep, having a capacity of about 20 tons ;
here the leaching is performed as usual. This treatment is
recommended for some highly pyritic ores, and in all cases where
Fi-. 53.
leaching is slow owing to the sliming of the ore, which is

especially liable to occur when much galena is present.
The power required to agitate the pulp is said to be at least
one horse-power per ton, and it is also stated that the ore cannot
be completely freed from the solution, which does not drain readily
from the ore after the agitation is finished, although the super-
natant liquid, forming about half of the solution, can of course be
removed by decantation. In Fig. 56, E is the vacuum boiler,
F the zinc box, and G the sump.
The rate of leaching may be increased by passing the liquid
upwards and downwards alternately, or by creating a vacuum

below the filter-bed. The latter may be done in several
ways,
as lias already been pointed out in
Chapter xiii., p. 272. The
use of a large boiler, in which a vacuum is created
by a Westing-
house or other pump, is perhaps the most
advantageous course;
as soon as the pressure in the boiler falls to about half an
atmosphere, it is connected with the aperture of the vat below
the filter-bed. Korting's injector is recommended by Stetefeldt
for the production of a vacuum. The rate of leaching is often
doubled by the diminution of the pressure, below the filter-bed,
to half an atmosphere, and in some cases it is increased from
J or 1 inch to 7 or 8 inches of liquid (in the leaching vat) per
hour. In some cases, as already remarked (p. 272), the creation
of a high vacuum is rather a cause of delay than otherwise, the
ore packing down and making leaching quite impossible. The
author has seen an attempt to leach a roasted ore, containing
much slimes, by means of a vacuum of 20 inches of mercury,
result in the production of a compact mass which could
scarcely
be penetrated by a pick-axe, while water stood above and could
not be drawn through in any way.
Strength of tJie Solution. In the Transvaal it is customary
to use two different strengths, viz., a "strong" solution, con-
taining from 0-25 to 0'35 per cent, of cyanide of potassium, and
a "weak" solution, containing from 0-05 to 015 per cent, of
cyanide. The strong solution is allowed to filter through during
a period of from eight to forty-eight hours, and the weak one is
then run through for twenty-four or forty-eight hours. The
amount of each solution, used is, in general, about J ton per ton
of ore, but, as before stated, only about 5i cwts. of solution per
c
ton of ore is used at the Robinson Mine with the " circulation
system." At the Mercur
.
Mine the strength of the solution in
use is 0-25 per cent.
When leaching is complete, so that no more gold is being dis-
solved, a point which can be determined by testing the escaping
solutioto with bright zinc shavings, the addition of fresh solution
is stopped, and the surface of the ore is laid bare by the sinking of
the liquid. There still remains in the vat, however, about 6 or 8
cwts. of solution per ton of ore. This might be reduced to 3 or 4
cwts. by draining for some time, but the charge would then shrink
and crack and separate from the side walls, so that the succeeding
operations would be less effectual. To displace the solution,
therefore, water is run on to the ore surface as soon as the latter
emerges, and is kept running until the quantity added is equal
to that of the solution originally retained by the charge. Up to
this point the discharge from below the filter-bed is allowed to
run into the "stock" solution, although somewhat more dilute
than the main mass, owing to slight admixture with the water.
The operation is now stopped and the tailings may be drained
and discharged, but, as they still contain a little cyanide, a slight
saving is effected by displacing the first wash-water by a second,

hi this case the first wash-water is run into a separate vat, and
used again to displace the stock solution, which is thus reduced
in strength less than if fresh water were used. More cyanide is
left in the tailings if the solution be drained off as
completely
as possible before being displaced, as the wash-water finds its
way down the cracks, leaving the unbroken masses of ore less
efficiently freed from cyanide.
this practice the volume of the solution is kept constant,
By
but its strength falls off considerably, owing to decomposition in
the ore and in the zinc boxes. It is tested in the " sump" to which
itflows from the zinc boxes, and enough cyanide added to raise
itto the required strength. The cyanide is usually dissolved in
a little water before being added to the stock solution, as the
amount of KCy present is more easily determined in a strong
solution than in any other form. At the Robinson Mine lumps
of cyanide are added to the liquid which is circulating through
the ore, as the insoluble impurities ("carbides of iron") are in
this way left on the surface of the charge, and are dumped with
the tailings instead of accumulating in the sumps.
Messrs. MacArthur & Forrest recommend the concentrated
solution of cyanide to be made and kept in a small vat standing
at a higher level than the large vats used for the storage of the
stock solution. Cyanide of potassium is supplied in zinc-lined
boxes holding 190 to 195 Ibs. of cyanide, containing about 78 per
cent, of available KCy. The contents of these boxes are broken
into lumps and placed in wire-gauze trays, .which are immersed
in water contained in the small vat. The soluble salts are soon
dissolved, potassium cyanide in particular being very rapidly
removed, and the tray is then lifted out, still containing the
insoluble material, which is thrown away. The strength of the
concentrated solution, which is kept to raise the stock solution
to its normal strength, may be from 10 to 25 per cent. It
should not be prepared until it is wanted, as it undergoes some-
what rapid decomposition.
Disposal of the Tailings. These are sometimes sluiced out of
the tanks by water under pressure. The sluice gates may be of
about 1 or 2 square feet in area, and are placed so that their lower
edges are level with the surface of the filter. Where the supply
of water is not large, or the fall of the ground is insufficient
for sluicing out the tailings, they are removed by shovelling out,
the cost being about 6d. per ton with European labour. Before
the vats are emptied, samples are taken, usually by means of a
long iron semicircular probe, shaped like a cheese-taster, which
is thrust to the bottom of the vat, then revolved by means of
the handle, and withdrawn with the tailings adhering to it. If
the results are high, samples from various parts of the vat should
be assayed separately, so as to find out if the percolation has
been uneven. At the Roodepoort Works, round vats have been

constructed, 40 feet in diameter and 8 feet deep, with a capacity
of 360 tons. They are placed on firm concrete foundations,
raised 6 feet from the ground, and the tailings are discharged
through four openings in the bottom communicating with tunnels.
It is said that these vats show no signs of leakage, and that they
can be discharged in six hours by twenty-eight natives at a cost
of l|d. per ton.*
At the Langlaagte Estate, the five large 450- ton vats, whose
upper edges are level with the surface of the ground, are dis-
charged by a steam crane by which truck loads of tailings are
elevated and placed on a tram line. The trucks are lowered
into the vats and loaded up by Kaffirs, and the total cost of dis-
charge is said to be 2d. per ton, including maintenance. The
vats and sumps are excavated in the solid ground, the walls
being built of brick and lined with cement.
3. Precipitation of the Gold. The precipitation of the
gold is effected by shavings of zinc freshly turned on a lathe.
In South Africa, the zinc linings to the boxes in which the
cyanide is imported are worked up for the purpose, a light
spongy mass easily traversed by the solution and presenting a
large surface for precipitation being formed. The shavings are
" zinc-boxes "
placed in wooden troughs, the ; they are divided
into compartments by partitions, which cause the solution to
flow alternately upward and downward. Each alternate com-
partment is empty, in order that, in passing through the shav-
" It is
ings, the solution may invariably flow upwards. arranged
so, because, if the solution were to flow down, the gold slimes
would collect on the upper surface of the zinc, and impede fur-
ther flow, but by upward flowing the gold slimes are precipitated
on the under surface of the zinc, from which they continually
drop off, and permit free passage of the solution."
At the Robinson Mine, the shavings are supported on wire-
screen trays, so that the finely-divided, precipitated gold falls
through to the bottom of the trough, while the unaltered zinc
remains on the sieve. At this mine the boxes are 20 feet long
and 2 feet square in cross-section ; there are ten compartments
in each, and about 40 Ibs. of shavings in each compartment,
except those at the head and foot, which are left empty in order
to allow the slimes to settle. About 30 tons of solution pass
through each box in a day of nine hours, so that every part of
the solution is in contact with the zinc for about twenty-five
minutes. The solution contains from 1 to 3 ozs. of gold per ton,
and, after passing through the boxes, contains from ^ to 2 dwts.
the "strong"
per ton. There are two sets of boxes, one for
" weak." In the latter, less zinc is
solution, the other for the
consumed, but the gold-slimes are poorer, less gold being con-
*
Eng. and Mng. Journ., Mar. 25, 1893, p. 273.
318 THE MKTALLURGY OF GOLD.
tained in the "weak" solution. The total amount of zinc

consumed is 100 Ibs.per day, the gold extracted being about
120 ozs. At the Mercur Mine, the boxes are 40 feet long and
12 inches square in cross-section, with compartments 3 feet long,
and contain zinc enough to maintain the contents of the last
compartment quite bright. The consumption here is about 1 Ib.
of zinc for 1 oz. of gold recovered.
When the solution comes in contact with bright zinc, the
latter turns black at once, owing to the deposition on it of finely
divided gold. The zinc is gradually dissolved, and the shavings
fall to pieces, those in the first
compartment being consumed
most rapidly. As the precipitation proceeds, the zinc is trans-
ferred from the lower compartments to the upper ones, and fresh
zinc is added at the foot of the box.
Clean-up. At the Robinson Mine, the clean-up takes place
once or twice a month. The screens containing the undissolved
shavings are lifted out of the zinc boxes, and the zinc-gold slimes,
remaining at the bottom of the boxes, are allowed some time to
settle. The clear liquor is then siphoned off and the slimes
allowed to run out by withdrawing a plug in the bottom of the
box, and drained through a 40-mesh screen, which retains a
small part only. After the residue has been rubbed on the
screen with sticks tipped with indiarubber, and well washed, it
is put back again into the first compartment of the box, on the
top of fresh shavings, as it consists mainly of small pieces of

unconsumed zinc. The subsequent fineness of the bullion
greatly depends on the care with which these operations are
conducted. By careful manipulation most of the undissolved
zinc can be separated in this way. The slimes proper are now-
transferred to enamelled iron pans and carefully and slowly
dried. The richness of the dried precipitate depends on the
strength of the cyanide solution, and on the time of contact as
well as on the quantities of base metals which are present in
the solution. By the prolonged action of the cyanide, the zinc
shavings become partly corroded and disintegrated, so that the
precipitated gold is mixed with zinc debris. Ordinary commer-
cial zinc contains a considerable proportion, generally over 1 per
cent., of lead, and a small quantity of carbon, besides other im-
purities, such as arsenic and antimony. All these accumulate
and are collected with the gold. The precipitated gold and the
carbon from the zinc are invariably in a state of fine division,
and, by using a fine sieve, the gold can be very closely separated
from the zinc, but, as the moderately fine particles of zinc still
retain a considerable proportion of gold entangled, it is not
always desirable to use a very fine sieve. If the cyanide solu-
tion contains copper, it is found with the gold.
According to Messrs. Butters and Clennel,* the pans in use
*
Eng. and Mnrj. Journ., Oct. 15, 1892, p. 365.
:at the Robinson Works contain about 5 or 6

gallons of dried
precipitate (often called "gold slimes"), which may contain as
much as 150 ozs., or as little as 20 ozs., of gold. A
little silver is
also contained in it, the remainder being
chiefly zinc and lead,
with smaller quantities of tin, antimony, organic matter, <fec.
The average in South Africa is said to be approximately :
Gold and silver, 20 to 50 per cent.
Zinc, . 30 to 60
Lead, about 10
Carbon, 10
A carefully prepared sample of gold slimes which was obtained

at one of the African works contains :
Gold, 60 per cent.

Silver, 10
Base metals and carbon, 30
This is by no means average, but is

exceptionally good, and
must not be taken as showing what is usual. The Hanauer
Smelting Company found the precipitate from the Mercur Mine
to contain the following substances :
*
Zinc, 39*1 per cent.

Carbonate of lime, 36-7
Gold, 4.4
Cyanogen, 3-5
Sulphur, . 2-6
Iron, 2-4
Residue, . 6-0
94-7
The carbonate of lime present is probably deposited in the boxes

chiefly from suspension in the solution, the gangue of the ore
mainly consisting of limestone.
4. Production of Bullion from the Precipitate. In the
United States, most mines will, no doubt, find it convenient to
follow the example of the Mercur Mine and sell the dried
precipitate to smelting companies, which pay about $'20 per oz.
for the gold contents. In South Africa this course cannot bo
pursued, and various methods have been suggested for effecting
the elimination of the zinc and other base metals. The usual
course is to melt the mass in graphite or clay crucibles, using as
fluxes sand, borax and bicarbonate of soda. The slag, which con-
sists of silicates of zinc, soda, <fec., corrodes the pots rapidly. Large
quantities of zinc oxide are given off as fumes, forming thick

crusts in the flues, and doubtless taking some gold with them, and
evil-smelling products of decomposition of the cyanides are also
evolved. The bullion produced varies in colour from a pale
yellow to a brownish linnet-green, and is about 650 fine, but
*
Eny. and Atng. Journ., Nov. 5, 1892, p. 440.
cannot be obtained uniform in composition, so that accurate

assays are difficult to obtain.
The results of analyses made on three ingots of bullion pro-
duced by the Mac Arthur-Forrest process in South Africa, and
shipped to London, are appended :
removed in solution. The residue is then spread in a thin layer

on iron trays, and dried and heated to a barely-perceptible red
heat in the flue of a furnace. The carbon, zinc, arsenic, &c.,
ignite readily, being in a very fine state of division, and
roasting proceeds regularly through the mass to the bottom
without any stirring, leaving a completely oxidised porous mass,
aggregated more or less into granules. This residue contains no
carbon and little arsenic, and consists chiefly of oxides of zinc
and lead, with metallic gold and silver. It is rapidly acted on
when heated with dilute sulphuric acid, which dissolves the zinc,
and forms insoluble sulphate of lead. There is no tendency for
the mixture to boil over, as is the case when cyanides are present,
and the residue is easy to filter, the granular sulphate of lead
offering little resistance. When the residue is washed and dried,
it is easy to fuse, not requiring large amounts of fluxes, and
forming liquid slags in which the greater part of the remaining

impurities are contained. The bullion thus prepared is often
worth 3 10s. per oz., the average value being over 3 per oz.
At the Robinson Mine, over three-fifths of the foreign metals
present in the gold bullion consists of silver, the remainder being
chiefly zinc. However, the presence of the last-named metal
does not diminish the price given for the bullion, and, if the
latter is more than 500 fine, no abatement is made in the
payments for the gold.
Motions Method of Precipitation. Instead of using zinc,
Molloy proposes to pass the gold solution through some kind of
amalgamator, such as the hydrogen amalgamator described at
p. 164. A "weak" electric current is sent through the solution,
the mercury being used as the negative pole. The solution is
decomposed and metallic potassium is released on the surface of
the mercury, bub instantly exchanges itself for the gold in
the solution, gold amalgam being formed and potassium cyanide
regenerated. In another modification, sodium is liberated by the
decomposition of sodium carbonate and replaces the gold in solu-
tion, an amalgam of the latter being formed. The sodium cyanide
in the solution is as efficacious as potassium cyanide for the pur-
pose of dissolving more gold, and it is said that there is much less
decomposition of the cyanide than if zinc is used. Moreover the
gold amalgam is much more easily handled than the zinc slimes,
and considerable losses are thus avoided, while a saving of skilled
labour is effected. These methods have never got beyond the
experimental stage, and it is not likely that they will prove of
practical value. It is certain, from the results obtained by
Dr. Gore's experiments* that mercury itself will not precipitate
gold from the cyanide solution, and therefore the action depends
entirely on the electric current.
Electro-deposition of the gold is used in the
Siemens-Halske
a***
process (see p. 329) and the conditions necessary for success

have been worked out and stated by the inventors ot that
method. One of the most important provisions is that the
layers of liquid between the anodes and cathodes must be very
thin in order that the ions may not be compelled to travel far
be tore they reach the poles. Hence, when great quantities of
liquid have to be treated, it is necessary for the electrodes to
have a very large surface. Moreover they should stand in a
vertical position so that they may not become coated with slime
settling from the liquids, which are usually more or less muddy.
It follows that mercury cannot be used in practice for the
cathodes. Von Gernet has calculated that 80 tons of quick-
silver would be required for the treatment of 100 tons of
solution per day, and the initial cost of the mercury would,
therefore, be some 20,000. Besides this, about 3,000 ounces
of gold, worth nearly 13,000, would be retained by the bath
after the whole had been subjected to nitration. Thus the
produce of, say, the first four months' work would be perma-
nently locked up, until the mercury was re-distilled.
Messrs. MacArthur & Forrest experimented on the action of
sodium amalgam, using a small tower filled with pieces of sodium
amalgam, through which the solution of gold in potassium cyanide
was allowed to trickle slowly. They came to the conclusion
that sodium was not more efficacious than zinc, but was much
more expensive. The saving of expense in the production of
the bullion from the precipitate was outweighed by the cost of
the sodium required to form the precipitate.
Plant Required. For the treatment of 2,000 tons of tailings
per month, Messrs. MacArthur & Forrest consider that six cir-
cular vats each 19 feet in diameter and 4 feet deep are sufficient.
Each vat holds about 40 tons of ore, and must be charged,
leached, and emptied in three days. The other requirements of
this plant are two large vats, each of 2,000 cubic feet capacity,
to hold the strong and weak cyanide solutions, and a similar vat
to hold the soda solution. If lime in the solid state is mixed with
the ore before it is charged into the leaching vat, this last tank may
be dispensed with. Zinc boxes, sumps, pumps, pipes, and launders
are also required, and furnaces to fuse the zinc residues. The
amount of water required for this plant is about 2,000 to 3,000
gallons per day, or 30 to 45 gallons per ton of ore. The labour
needed consists of eight men for a shift of twelve hours. A
zinc
box 14 feet long, 2 feet deep, and 1 foot wide, divided into six or
eight compartments, and holding 100 Ibs. of zinc, should suffice
to treat in about nine hours the gold solution obtained each
day.
A plant capable of treating 20 tons of pyritic ore by means of
agitation, consists of six agitator-vats (each holding a charge of
1 j tons of ore), three large filtering vats, and two vacuum boilers
to assist in the filtering. The discharge from the agitators to the

filters is by means of pipes which are at least 2 inches in diameter.
Pumps are required to raise the liquid from the sump to the stock
solution vat, and to create a vacuum in the boiler. The power
required for the agitators and pumps is about 6 H.P. The con-
sumption of water is from 600 to 1,000 gallons per day, or from
30 to 50 gallons per ton of ore. Labour consists of three men
for a shift of eight or twelve hours.
Treatment of Ore Slimes. Ore slimes, consisting of fine
particles of clay or of ferric hydrates, sometimes exhibit a curious
action in withdrawing gold from almost any of its solutions.
This has been noticed both in chlorination and cyanide mills.
As the slimes settle, so that the muddy solution becomes clear,
the gold appears to go to the bottom with them, so that they
often contain many ounces of gold to the ton after treatment,
although not especially rich at first. This fact must be remem-
bered in the treatment of ores which slime badly. It is possible
that the precipitated gelatinous ferric hydrate formed in the vats
by the action of alkalies on oxidised salts of iron may have some
similar effect. This substance should be collected and assayed
occasionally.
On the Rand, there have been accumulated hundreds of
thousands of tons of slimes which cannot be made to yield any
of its gold to cyanide on the large scale, owing to its imperme-
ability, although the gold contained in it is readily soluble in
the laboratory. The average value of the Robinson slimes is
between 7 and 8 dwts. of gold per ton, and the fineness is such
that it would pass through a 225-mesh screen. It has been
proposed by Bettel to treat this material in Johnson's filter-
presses. The slimes are mixed with a very dilute (0-01 per cent.)
cyanide solution and thoroughly agitated, then filtered, and
water forced through under a pressure of 100 Ibs. to the square
inch. This pressure leaching is similar to that suggested by
E. N. Pviotte in connection with the hyposulphite lixiviation
process (see p. 273). It is stated that about 98 per cent, of the
assay value of the slimes can be extracted in this way, but the
method has not yet been applied on the large scale.
Filter press separation was in use at the Crown Mines, New
Zealand, as early as the year 1889. Separation of the liquid by
decantation has also been used with even greater economical
success.
Testing of Ores. Experiments with small quantities of
material in the laboratory will usually determine the maximum
extraction that can be looked for. The weight of ore taken may
be from 100 to 200 grammes. It should be digested in a beaker,
with sufficient solution to form a thin mud, with occasional
stirring, or better still, in a funnel or lamp glass, through
which
the solution slowly passes. The solution is then separated,
and the residue thoroughly washed by filtration, and assayed

to find how muchof the precious metals has been removed.
The consumption of cyanide may be determined by titrating
the solution before and after use. In this way, the maximum
extraction of gold and silver from an ore which has been crushed
to different degrees of fineness, with solutions at different
strengths, acting at various temperatures and for different
lengths of time, may be determined. Such data will be of great
value in determining the degree of suitability of the process to
any given ore, but the results obtained cannot usually be
repeated on the large scale. Mechanical difficulties preclude
fine crushing beyond a certain point : the relative amount of
solution used in practice must be less than the large quan-
tities giving the best results on a small scale. Moreover, the
washing and filtering is more perfectly performed in the labora-
tory than in the mill, but, on the other hand, the consumption
of cyanide may be less in the latter case than in the former.
Preliminary tests as to the amount of alkali which it is neces-
sary to add, and the amount of cyanide decomposed are made as
follows :
(a) Alkali. 200 grammes of ore are made into a paste
with water in a porcelain basin, and tested with litmus paper.
If an acid reaction is observed, a titrated solution of caustic soda
is run in, little by little, until the mixture is neutral. It is
convenient to make the alkaline solution of such a strength that
each c.c. corresponds to one pound of caustic soda per ton of the
ore which is being tested. (b) Decomposition of Cyanide.
100 grammes of the ore is shaken in an 8-oz. stoppered bottle
with 50 to 100 c.c. of a solution of KCy for fifteen minutes, then
leftundisturbed for twelve hours, and finally shaken again for
five minutes. The available cyanide in the separated solution is
then estimated and compared with that in the original solution.
By merely shaking for a few minutes, most of the decomposition
of cyanide is made to take place, but not the whole. The
strength of solution used may be from 0*1 to 0'5 per cent, of
KCy according to the strength required for dissolving the gold.
In general, the stronger the solution the greater the amount of
cyanide decomposed. If soluble salts are present in the ore,
they should be removed by washing with water before applying
the cyanide, and the effect of a soda solution in saving cyanide
should also be tested.
The Cassel Gold Extracting Company supply a small testing
plant capable of treating 3 cwts. of ore at one time, under
conditions closely resembling those obtaining in a good mill.
The plant consists of a small revolving barrel for solution, a
filtering vat and vacuum boiler, and a zinc box for precipitation.
The time of treatment in the barrel varies from six hours
upwards. Such a barrel is at the Royal College of Science.
Direct Treatment of Band Ore. Several attempts have
been made to treat auriferous banket by the cyanide process

alone, without first amalgamating it. In general, however, these
have been unsuccessful. The coarser particles of gold, usually
removed on the copper plates, require far too much time for
their complete dissolution to admit of a high percentage of
extraction being attained. The pyritic ores have not yet been
proved to be amenable to this method, but with certain oxidised
ores from near the surface of the ground, all the gold is in a
very finely-divided state, and 75 to 80 per cent, is extracted by
cyanide in two or three days. At the George and May Mining
Company's works, the oxidised banket is very friable and easily
disintegrated, and this was crushed dry, and treated direct with
complete success in 1895. Here nearly half the ore was con-
verted into intractable slimes if it was crushed in the battery,
and in the last few months of 1894 only 55 per cent, of the gold
had been extracted by amalgamation, followed by treatment
with cyanide. In the method now adopted the ore is crushed
dry by a Gates crusher, and dumped at once into the vats, where
it is treated as follows
(Hatch & Chalmers) The ore is saturated
:
with a solution containing O06 per cent, of KCy, and this is

displaced by a "strong solution" containing 0*28 percent.; a
" "
weak solution containing (HO per cent, of cyanide follows,
and is washed out with one containing 0'05 per cent. "About
sixty hours are required for the whole treatment. Some lime
is mixed with the ore, and on an
average 0'89 Ibs. of cyanide of
potassium are used per ton of ore treated." The ore is so friable
that it is reduced to barren pebbles and loose sand by merely
passing through the rock breaker. About 75 per cent, of the
gold is extracted, the ore containing about 5 dwts. per ton.
The total cost of mining and treatment is about 12s. per ton.
Use of Cyanide in the Stamp Battery. Efforts have been
made in South Africa and the United States to treat the ore
direct from the battery instead of first passing it over amalga-
mated plates. Ore from the May Consolidated Mine, Johannes-
burg, was crushed by the African Gold Recovery Company in
the latter part of the year 1892,* with cyanide solution instead
of water, and led at once into the filtering tanks. The results
are stated to have shown that the coarse gold resisted the attack
of the cyanide for so long a time as to render the process
uneconomical. No doubt this system will meet with great
success in cases where all the gold is in a finely divided
condition.
At the Stewart Mine, Bingham, Utah, a combination of the
amalgamation and cyanide processes is said to have been applied
successfully to ore containing a mixture of coarse free gold and
fine rebellious particles. f The ore is crushed and amalgamated
*
Eny. and Afnff. Journ., Get 8, 18.2, p. 342.
., April 15, 1893, p. 339.
in Huntington and Crawford mills successively, the water used

being a solution of cyanide of potassium. After leaving the
mills the solution and pulp are run into settling tanks and the
liquid drawn off. It would appear likely that the decomposition
of the cyanide in the mills would be excessive, but no details of
the working have been published.
Method of Treatment at the Robinson Mine. At the
Robinson Mine* there is a 60-stamp battery, pulverising 280
tons per day through a 40-mesh sieve, the ore containing nearly
2 ozs. gold per ton. Of this amount 71*04 per cent, is caught on
the amalgamated copper plates, 5'13 per cent, is extracted from the
concentrates at the chlorination works, and 18 '55 per cent, from
the tailings at the cyanide works, while 5-28 per cent, is lost.
The pulp from the battery is concentrated, and the tailings caught
in pits and shovelled into cars and conveyed to the cyanide
works, while the concentrates go to the chlorination works.
There the concentrates are placed in heaps, assayed for sulphur,
and mixed so as to make a uniform product for roasting. They
are roasted in three furnaces, each 60 feet long, treating 600 to
800 tons per month. When roasted, the material is charged
into ten circular chlorination vats with bottom discharge and
luted rings and lids. The chlorination occupies four days, 6 Ibs.
of chlorine being used for each ton of concentrates. The gold is
precipitated by a solution of ferrous sulphate, the mass being kept
agitated by means of a jet of compressed air. The consumption
of ferrous sulphate is from 10 to 15 Ibs. per ton of concentrates.
The gold is allowed to settle for some time in a series of large
vats, and a complete clean-up takes place only four times a year.
At the cyanide works there are twelve vats, each holding 75
tons. The treatment here also occupies four days, the output
being 225 tons per day. The whole of the tailings are subjected
to treatment, the coarse gold having been removed on the plates,
and the pyrites by concentration, so that almost clean sand,
containing from 8 to 10 dwts. of finely divided gold, is left for the
action of the cyanide. Of this from 1 \ to 2 dwts. are left in the
tailings, the extraction being over 70 per cent.
The Cyanide Process at the Sylvia G-old and Silver
Mining Company, Thames Valley, New Zealand.! The
ore from the deeper levels of the mine contains a high percentage
of complex minerals, consisting of galena, zinc-blende, copper,
and iron pyrites, and does not, as a rule, show any visible gold.
It is crushed in a battery and passed over amalgamated Muntz-
metal plates, by which the coarse gold is extracted, the amount
thus saved being about 5 dwts. per .ton. The ore is then
*B. H. Brough, Journ. of the Soc. of Arts, vol. xli., 1893, p. 173.
tlhis description is taken chiefly from the reports of the manager,
Dr. A. Scheidel.
carefully sized and concentrated by a complicated system of jigs,

rotary tables, buddies, and sizers. The concentrates obtained are
classed as follows jigger concentrates, first-class slime concen-
:
trates, second-class slime concentrates, and buddle concentrates.

The jigger concentrates contain 4 ozs. 5 dwts. of gold and 20 ozs.
of silver per ton, and consist chiefly of iron and copper pyrites
and zinc blende, with a little galena and quartz. The galena in
the ore carries greater amounts of gold and silver than the other
minerals. It forms slimes for the most part, and is found in
the slmie concentrates. Of these the first-class contain about
20 per cent, of galena, and an average of 10 ozs. 6 dwts. of gold
and 44 ozs. of silver per ton the second-class concentrates contain
:
very little lead, and assay 4 ozs. 5 dwts. of gold and 20 ozs. of
silver per ton. The buddle concentrates are midway in richness
between the two slime-concentrates. About 80 per cent, of the
total values in the ore are saved on the tables and in the con-
centrates, and the latter can all be treated by the cyanide
process.
It was originally intended to dispose of the concentrates by
sale, but the prices realised, after deducting the expenses of
bagging, carting, shipping, insurance and treatment, were so
small as to render treatment on the spot an imperative necessity.
The system ultimately adopted was the MacArthur-Forrest
process, after exhaustive trials, and a plant was erected capable
of treating 20 tons per day. " It consists of three
large agitators,
three vacuum niters, a grinding-pan, cyanide solution tank,
tanks for gold solution, vacuum and other pumps, and some
minor appliances. The whole plant is of local manufacture."
The filtration of the concentrates after agitation is difficult to
accomplish, and necessitated the introduction of a vacuum filter
" The time of
patented by Dr. Scheidel. agitation, and the
strength of solution applied, vary in accordance with the
quality of the material. The quantity of cyanide used for
the highest grade of ore amounted to less than 1 per cent.,
and for low-grade material to considerably under 0'5 per cent.
The time of agitation varied between five and twenty-four
hours."
The Sylvia Company acquired the right of using the
MacArthur-Forrest process on payment of a royalty of 7J per
cent, on the bullion extracted. The patentees did not interfere
with the construction of the plant, which was left in Dr.
Scheidel's hands. He states that " the results of extraction
have varied in accordance with the quality of the material, the
slimes generally giving better results than the other products,
and the (richer) first-class slimes returned a higher percentage of
gold and silver than the lower-grade materials." The extraction
was as follows :
cent, of cyanide is then run on, and replaced

by a stronger
solution, which is allowed to stand for twelve hours, and then
drained off and the sand removed to the leachino-
ordinary
tanks. Here a solution containing 0'2 per cent, of cyanide is
added, and allowed to stand for twelve hours it is then re-
;
placed by weaker solutions, 0-08 and O04 per cent, being used
successively, and finally water, rendered alkaline by milk of
lime. An average of 78 per cent, of the gold is extracted,
while, before the "double treatment" was introduced, only 55
per cent, was obtained. The capacity of the plant is 17,000
tons per month, and the working expenses average 4s. Id. per
ton. The consumption of chemicals per ton of ore is as
follows :
cyanide, 0'98 lb., zinc, 0-22 lb., and caustic soda,
0-22 lb. The bullion recovered is worth 3, 5s. per oz.
Siemens-Halske Process. In this process the gold is de-
posited from solution by the passage through the liquid of a
current of electricity. Moreover, as the precipitation is equally
complete and as readily obtained in extremely dilute cyanide
solutions as in those containing 0*1 per cent, or over, the
strength of the solutions used in dissolving the gold from the
ores is made less when electrical precipitation is employed than
if zinc is the precipitant. Hence the whole method forms an
interesting variation of the ordinary Mac Arthur-Forrest process,
and bids fair to assume great importance in the future. Gold
can be extracted from its ores as completely by a solution
containing 0'03 per cent, of cyanide as by one containing 0-3
per cent., the only difference being that the time required is
considerably longer. On the other hand, the advantage in
using the more dilute solution is that the selective action in
favour of the gold is increased, and the amount of cyanide
decomposed by "cyanicides" in the ore is diminished. In
addition to this, some cyanide solution is invariably left in the
ore, and if the "weak" solution used to finish the dissolution
of the gold contains only O'Ol per cent, of cyanide, instead
of 0-1 per cent., the amount of cyanide lost in this way by
mechanical means is also reduced.
The process was adopted on a large scale at the Worcester
first
mill in the Transvaal. Here the vats are 20 feet in diameter
and have their sides formed of staves 10 feet long; the five vats,
hold 135 tons each. The battery pulp, after passing over Frue
vanners, is classified into four products by hydraulic classifiers.
The first series, which consists of spitzluten, removes the coarse
sand and pyrites, amounting to 15 per cent, of the pulp and
containing 15 dwts. of gold. This product is treated for nine
after being
days with solutions of 0'08 per cent, of cyanide, and,
washed with O'Ol per cent, solutions, gives residues assaying
cent, of the gold is,
1J to 2 dwts., so that from 87 to 90 per
extracted. The second product yielded by the hydraulic
classifierscomprises f.O per cent, of the pulp, contains 6 dwts.

of gold per ton, and, after five days' treatment with solutions
ranging from 0'05 per cent, downwards, yields residues con-
taining from 1 to 1*25 dwts., showing an extraction of 80 to 84
per cent. The finest sand is separated from the slimes by
pointed boxes ; the slimes amount to 25 per cent, of the whole
pulp, and are not treated, being unleachable, though they assay
4J dwts. The fine sands, constituting 10 per cent, of the pulp,
contain 4|- dwts. of gold, and after treatment yield residues
assaying 1 dwt. per ton. The consumption of cyanide averages
J Ib. per ton of the tailings, of which 3,000 tons are treated per
month.
The precipitation plant consists of four boxes, each 18 feet
long, 7 feet wide, and 4 feet deep. Copper wires are fixed along
the tops of the sides of the boxes and convey the electric current
from the dynamo to the electrodes. The anodes are made of
iron, and the cathodes, on which the gold is deposited, of lead.
These metals appear at present to be the most suitable for the
purpose, but it is possible that other substances may be found
equally good. It is necessary to have a very large surface on
which to deposit the gold, so that a bath of mercury is out of
the question, von Gernet calculating that it would require 80
tons of the metal to give the 10,000 square feet of surface
necessary to deal with 100 tons of solution in twenty-four hours.
Amalgamated copper plates were tried but abandoned, as the
mercury penetrated the copper under the influence of the
current, and a dry amalgam resulted which did not adhere to
the plate. Lead answers all the requirements of the cathode
laid down by von Gernet, which are :
(1) that the precipitated
gold must adhere to it, (2) that it must be capable of being
rolled out into very thin sheets to avoid unnecessary expense,
(3) that itmust be easy to recover the gold from it, and (4) that
it must not be electro-positive to the anode, in order to pre-
vent return currents being generated when the depositing
current is stopped. A fifth requirement might be added, that
the gold should be separable from the cathode without de-
stroying the latter. This requirement is not fulfilled by lead.
At the Worcester mill the anodes are iron plates, 7 feet long,
3 feet wide, and J inch thick they are supported in a vertical
;
position by wooden strips nailed to the box, and are covered

with canvas to retain the small quantity of Prussian-blue pro-
duced. The sheet-lead cathodes are stretched on wires fixed
in light wooden frames which are suspended between the iron
plates. There are in all 3,000 square feet of cathode surface.
The solution is made to circulate between the plates passing
alternately over and under the edges of the anodes. The boxes
are kept locked except when cleaning up is necessary, when the
cathodes are removed and replaced with fresh sheets of lead.
The lead, which contains from 2 to 12 per cent, of gold, is then

melted into bars and cupelled. The consumption of lead is
750 Ibs. per month, its cost being equal to l^d. per ton of
tailings. The gold is comparatively free from base metals,
The total cost of treatment at the Worcester mill is under 3s.
per ton of tailings. The electrodes are placed inches apart,
and a current of 4 volts is employed.
Results of Process. The success of the process in South
Africa, where a new industry has been created, has been com-
plete. The rapid progress which has been made there may be
judged from the production of gold by cyanide in each year since
its introduction in 1890 :
Year.
generally amounts to about Is. per ton of the tailings treated.

The amount of tailings now being treated is about 8,000 tons
per day, and the average yield is about 4*8 dwts., which is equal
to about 70 per cent, of the gold contained in the material
treated. It is evident from these results that the minimum
value of tailings capable of being treated at a profit in the
Transvaal ranges from about 1 J to 4 dwts. of gold per ton. The
cost of the treatment of concentrates is usually over 20s. per ton,
while the amount extracted is about 90 per cent.
The importance of the process to the mining industry is not
entirely represented by the output given above, as few of the
mines could work at a profit if they depended merely on the
gold extracted on the plates. The profit of say 15s. per ton
derived from the tailings converts the mine from a losing
into a paying concern in a large number of cases, and this profit
could not be earned without the cyanide process.
In other parts of the world, the gold-mining industry is less
dependent on the cyanide process than in South Africa, but
nevertheless it is gradually making its way on every quartz-
mining gold-field of any importance. In Australia many mines
have adopted it, and a large customs mill is at work in Queens-
land where over 14,000 ozs. of gold bullion were produced in
1894. In New Zealand the process is being worked at the
Waihi Mine, the Crown Mine, and a large number of others.
The progress in this colony may be judged from the fact that
in 1893, 17,271 ozs. of bullion of the value of 28,657 were
produced by cyanide, while in 1895 the amount had increased
to 133,162 ozs. of gold. The process has only lately been intro-
duced into India, where in 1895, 3,614 ozs. of gold were pro-
duced, but several works are in course of erection, and the
output may be expected to be greatly increased in the near
future. In the United States the progress of the Mac Arthur-
Forrest process was for some years somewhat slow when com-
pared with that in South Africa and New Zealand. It was
adopted at the Mercur Mine, Utah, as early as the year 1891,
and was a great success from the start. It is now at work in
many of the States, the production in Colorado from its use
being considerable. In 1895 the production of gold in the
whole of the United States by cyanide was valued at over
300,000. The process will certainly continue to grow in im-
portance for some time to come in all parts of the world, and
in 1895 the total output of gold by its use could not have fallea
far short of 3,000,000.
CHEMISTRY OF THE CYANIDE PROCESS. 333
CHAPTER XVI.
CHEMISTRY OF THE CYANIDE PROCESS.
Action of Potassium Cyanide on Gold and other Metals.
It has long been known that metallic gold is soluble in potassium
cyanide. Elkington, in 1840, in a patent specification, speaks
of dissolving finely-divided metallic gold in this solvent, and
Bagration, in 1843,* studied the action of cyanide on plates of
gold, and announced that they were slowly dissolved. Faraday,
in 1857,t pointed out that gold-leaf is dissolved by a dilute
solution of the salt, and also showed that if the gold floats on the
surface of the liquid, so that one side of the leaf is in contact
with the air, while the other is bathed by the solvent, the action
is much more rapid than if the metal is completely submerged.
Eisner had previously proved J that the presence of oxygen is
required for the solution of the gold. On evaporating the solu-
tion, colourless octahedral crystals of auro-potassium cyanide,
KAuCy2 are formed, which may be viewed as being a double
,
cyanide, produced as follows :
4Au + SKCy + O2 + 2H 2 = 4KAuCy 2 + 4KOH

This equation is exothermic. In calories, it may be expressed
4( + 82-3 + x + y - 64'7) = 4(17'6 + x + y)
Here, +82-3 is evolved by the formation of potash from potas-

sium, oxygen, and water, x is evolved by the union of Au and
Cy, and is probably positive (see below, p. 335), y is evolved by
the union of AuCy and KCy, and is also a positive quantity,
and - 64-7 is absorbed by the decomposition of potassium cyanide.
The equation for the solution of silver is
2Ag + 4KCy + O + H 2
= 2KAgCy 2 + 2KOH
This may be expressed in stages thus
2Ag + 2KCy + + H 2 = 2AgCy + 2KOH
AgCy + KCy = KAgCy 2
in the
Theoretically, therefore, 130 parts by weight of KCy
the solution of 196*8
presence of eight parts of oxygen suffice for
parts of gold. This has been recently proved by Mr. J. S.
Maclaurin to be the case in all carefully conducted experi-
*
Bull, de I'Acad. des Sciences de St. Petersburg (1843), vol. ii., p. 136.
t/toy. Inst. Proc., vol. ii., p. 308.
+ Erdm. Journ. Prak. Chem., vol. xxxvii. (1846), pp. 441-446.
Journ. Chem. 6oc. (1893), voi Ixiii., p. 724.
ments. The amount of oxygen dissolved in liquids not specially-

prepared, to say nothing of that contained in a porous mass of
pulverised ore, is consequently enough for the solution of great
quantities of gold.
The voltaic order of the metals in different solutions of cyanide
of potassium is given in the following table :
1 part KCy in 8 parts

water.*
The heat disengaged in the union of equivalent quantities of

the metals in the solid state (except Hg) with cyanogen is given.
Messrs. MacArthur and Forrest have stated that no increase
in the rate of solubility of gold and silver takes place on
increasing the concentration of the cyanide solution. As excep-
tion has been taken to this statement by other authorities, the
author conducted the following experiments in the laboratory
at the Royal Mint, in order to increase the available data on
the subject as far as metallic gold is concerned.
Gold cornets, which had all been subjected to the same
treatment, being cupelled, boiled, and annealed together, were
submerged in 30 c.c. of KCy solution. Cornets offer a large
surface for action, as they are spongy in texture. After some
time the cornets were taken out of the solution, well washed,
dried, ignited, and weighed.
The results obtained were as follows :
from an ore by a cold solution of cyanide, was re-precipitated on

the ore by heating towards the boiling point."
Mr. J. S. Maclaurin of Auckland University, New Zealand,
has recently demonstrated * that the rate of dissolution of pure
gold, in the form of plates, in potassium cyanide solutions passes
through a maximum when proceeding from dilute to concentrated
solutions. The maximum is reached when the solution contains
0*25 per cent, of KCy. The solubility of gold is very slight in
solutions containing less than O005 per cent., but increases
rapidly as the strength rises to O'Ol per cent., when the rate
of dissolution is ten times as great as in the 0-005 per cent,
solution, and about half as great as that in the 0'25 per cent.
The rate increases slowly as the strength rises to 0-25 per cent.,
and thereafter decreases much more slowly, until in 15 per cent,
solutions the rate of dissolution is about equal to that in O'Ol per
cent, solutions. Higher strengths show a gradual diminution in
the rate of dissolution up to saturation point. Silver is also dis-
solved at a maximum rate in solutions containing 0'25 per cent,
of cyanide, and the changes in the rate are similar to those
noted above in the case of gold, the rates for silver being always
about two-thirds of the corresponding rate for gold, or, roughly,
in the same ratio as the atomic weights of the two metals. In
both cases there is hardly any change in the rate of solubility as
the strength rises from O'l per cent, to 0'25 per cent. From
these results, it is seen that the most active solutions are now
used in the ordinary practice of the MacArthur-Forrest process
in South Africa, and that the solutions used in the Siemens-
Halske practice (O'Ol per cent, to 0-1 per cent.) are little inferior
to them in activity. It is remarkable that the solubility of
oxygen in cyanide solutions undergoes similar changes as the
concentration increases, and it is to this fact that Maclaurin is
disposed to attribute the variations in the rate of dissolution of
the gold. The oxygen which unites with the cyanide, converting
it into cyanate, is thereby made inactive, as the presence of
cyanate of potassium has no effect on the rate of dissolution

of the gold.
The table on p. 338 gives the results of some experiments
made by Mr. Louis Janin, jun.,f on the solubility of metallic
silver in potassium cyanide. In this case it appears that a
maximum rate of dissolution is reached when the strength of
the solution is only about 1 per cent. ; the rate then diminishes
gradually as the strength increases from 1 to 3 per cent., and
continues to diminish, although much more slowly, until the
strength reaches 15 per cent., after which an increase
in con-
centration has no effect on the rate of dissolution. The influence
of the volume of solution seems to be small, and that of time not
very great, after the first few hours.
* and vol. Ixvii. (1895), p. 199.
Journ. Chem. Soc., vol. Ixiii. (1893), p. 731 ;
.and Mng. Journ., Dec. 29, 1888.

22
Weight of
Cement
Silver.
Mgs.
proceed slowly even in the cold, especially if free alkali is

present. It is obvious, from the observed facts of the decom-
position of cyanide solutions mentioned above, that (1) the
solution must be kept from contact with the air as far as
possible, by having closely fitting lids to the storage and leaching
vats, the zinc boxes, &c., and by transferring the solution from
one to the other by me.ins of iron pipes (not open launders); and
(2) the solution must be kept free from acids, and this can be
effected by the addition of a little alkali.
Decomposition in the Zinc Boxes. Pure zinc has only a.
very slow action on solutions of potassium cyanide, but the action
of the gold-zinc couple, formed by the black deposit of gold
(which may be really a compound of gold and zinc), and the
unaltered zinc, is much more vigorous. This gold-zinc couple
probably develops enough electromotive force to decompose
water thus
Zn + 2H 2 = Zn(OH) 2 + H 2 (41'8-34-5= + 7'3)
The positive element, zinc, is thus oxidised, and subsequently
dissolved by the cyanide solution. As a matter of fact, Messrs.
Butters & Clennel* observed a vigorous evolution of small
bubbles, which proved to be mainly hydrogen, in the zinc boxes
at the Robinson Mine. The hydrogen thus evolved doubtless
carries off mechanically some hydrocyanic acid, and so the odour
of the latter, noticeable above the zinc boxes, may be accounted
for. The nascent hydrogen also unites directly with hydro-
cyanic acid, forming methylamine, thus
HCN + 2H 2 = CH 3 .NH 2
The presence of the methylamine may in part account for the
ammoniacal odour sometimes occurring above the zinc boxes,
which may, however, be also caused by the hydrolytic decomposi-
tion. The hydrate of zinc, formed as shown above, dissolves at
once in excess of the cyanide
Zn(OH) a + 4KCy = K 2 ZuCy 4 + 2KOH,
and the increase in alkalinity of the solution is thus explained.
It is observed that when strong solutions of caustic soda have
been used for neutralising the acid salts of the ore, a white deposit
is formed on the zinc, and Messrs. Butters & Olennel suggest
that this may be cyanide of zinc, ZnCy 2 formed by the action of
,
the sodic zincate, on the double cyanide of zinc and

Zn(ONa) 2 ,
potassium, the equations being

Zn + 2NaOH = Zn(ONa) 2 + H 2
Zn(ONa) 2 + K 2 ZnCy 4 + 2H 2 = 2ZnCy 2 + 2NaOH + 2KOH
In this way the solution would continually become more alkaline.
*
Emj. and Mng. Journ., 1892, p. 410.
By the reactions in the zinc boxes, not only is the potassium

cyanide solution weakened by decomposition, as above, but a
large amount of zinc is dissolved. The gold is precipitated
thus
2KAuCy 2 + Zn = K 2 ZnCy 4 + 2Au,
the silver in the ore being dissolved and precipitated by means
of precisely similar reactions to those given for gold. The
double cyanide of zinc and potassium is not available for the
solution of the precious metals, and as long as an excess of zinc
is present, no gold will be dissolved by a solution of potassium
cyanide flowing through the boxes. The presence of large

quantities of the double cyanide of zinc and potassium in
the solutions is not prejudicial to the solvent action of the
simple cyanides. At the Mercur Mine, the stock solution was
apparently as efficacious after nine months' use as at the start,
although it must have contained large quantities of zinc cyanide.
Feldtmann showed that gold in ores can be dissolved by zinc-
potassium cyanide, but J. S. C. Wells points out that the double
cyanide remains undecomposed by gold so long as any simple
cyanide is present. An accumulation of base heavy metals in
the solution can be got rid of by the addition of soluble
sulphides.
On the influence of temperature on precipitation, Mr. J. S.
MacArthur writes : "Wehave hardly any data regarding the
influence of temperature on the precipitation by zinc. The few
experiments which have been done did not indicate any
appreciable difference in precipitation by a rise of temperature,
and, on the other hand, they have clearly shown an enormous
waste of cyanide by the formation of urea, which manifested
itself by its strong unpleasant odour."
Action of Potassium Cyanide on Metallic Salts and
Minerals Occurring in Ores. The ordinary gangue of most
ores (viz., silica and silicates of the alkalies and alkaline earths)
exercises no direct influence on the cyanide solution. The
carbonates of the alkaline earths are also probably without
influence. The decomposing effects of sulphides of the heavy
metals vary with the physical state of the sulphides.
In 1892, Kedzie* made a large number of experiments, using
different samples of pyritic ores and concentrates, and solutions
containing from 0'5 to 1'5 per cent, of pure potassium cyanide,
and found that the consumption (decomposition) of the cyanide
varied from 3 to 50 Ibs. per ton of ore. He found that the
consumption increased with the time, the volume of the solution,
and the degree of concentration, the effect of the latter being
especially marked. An increase of concentration from 0'5 per
cent, to 0-9 per cent, caused the consumption to be doubled,
and a further increase to 1'5 per cent., doubled the consumption
*
Lng. and Mng. Journ., p. 606, 1892.
again. Sulphide of iron, he found, acts very little on potassium

cyanide, but sulphides of copper and zinc are rapidly dissolved.
Messrs. Mac Arthur and Forrest find that dilute cyanide solu-
tions exercise a " selective action in dissolving gold and silver
"
in whatever form they may be present, in preference to sul-

phides or other salts of the base metals. There are exceptions
to this rule, some of which are noted in the sequel. "For
instance, cyanide of potassium solution has a strong tendency
to dissolve precipitated sulphide of zinc, but its action on
the natural sulphide of zinc, blende, is almost nil. The same
holds for compounds of iron, and thus we prove selective action
by the average result of a series of experiments on ores. Let us
suppose a pyritic ore containing about 7 per cent, of iron and
8 per cent, of sulphur with about 1 oz. of gold to the ton. After
grinding, this ore is treated with a solution containing about
1
per cent, of cyanide of potassium. The most of the gold will
be dissolved and the rest of the ore left practically untouched.
It is obvious that the amount of cyanogen contained in the
solution is insufficient to combine with the iron present in tho
ore, yet, notwithstanding the much greater mass of iron sulphide
present and open to attack, it is the gold that is selected for
action by the cyanide solution. Taking the average result of
our work we find that a higher percentage of gold than of silver
is extracted, which justifies us in concluding that the selective
action is greater on the former than on the latter. One of the
ores on which our early investigations were done was composed
as under :
Copper, '15 per cent.

Arsenic, 15-09
Antimony, Traces.
Sulphur, 14-65 per cent.
Iron, . 18-77
Silica, 36-20
Lead,. 2-66
Zinc, . 4-00
Alumina, 4-20
Gold, per ton 2 ozs. 2 dwts. 16 grs.
Silver, ,,
2 ozs. 13 dwts. 8 grs.
In this ore we had an extraction of gold 85 per cent., silver

50 per cent., fora consumption of cyanide of about 0-45 percent.,
and investigations showed that the action was directed in tho
order, gold, silver, iron, zinc, copper. From the amount
of cyanide
consumed it is obvious that the amount of base metals dissolved
must have been very slight.
"The consumption of cyanide on fresh concentrates varies
In many
naturally with the composition of the
concentrates.
cases it is less than 0-2 per cent, of their weight. When the
concentrates contain marcasite there is a greater consumption oi
the ordinary yellow
cyanide than when the pyrites is entirely of
cubical description. The presence of compounds of copper,

physically soft, also tends to increase the consumption."
It has been laid down as a general rule that oxides, hydrates,
carbonates, sulphates and sulphides of those metals which are
electro-positive to gold in cyanide solutions are dissolved more
rapidly than the last named metal, whether it is present in the
metallic form or contained in its commonly occurring salts.
This rule certainly applies to the precipitated salts commonly
occurring in the laboratory, but J. S. MacArthur has shown that
the case may be quite different when the naturally occurring
minerals are concerned. Thus, not only is precipitated sulphide
of copper rapidly dissolved, but also a sooty form of the same
substance occasionally met with as a mineral occurring in ores.
On the other hand, fused copper matte is scarcely acted on at all,
and in the majority of cases the same may be said of the hard
dense sulphides of copper usually found in nature. Sulphide of
"
zinc exhibits the same differences of behaviour the "black-jack
:
concentrates of the Ravenswood Mine, Queensland, can be treated

with good results, little zinc being dissolved. Again, oxide of
copper, if freshly precipitated, is strongly acted on by the cyanide,
but if it is heated to dull redness in a muffle it becomes insoluble,
and a large excess of this material added to a gold ore makes no
difference in the percentage of extraction, while the consumption
of cyanide is not increased by its presence.
The action of cyanide solutions on sulphide of silver is similarly
dependent almost entirely on its physical state. Experiments
conducted by Fresenius showed that if a weak solution of silver
nitrate is precipitated by a weak solution of sulphide of ammonium,
the resulting sulphide of silver is soluble in a weak solution of
cyanide of potassium. On the other hand, if strong solutions of
silver nitrate and ammonium sulphide are mixed together, the pre-
cipitated silver sulphide can only be dissolved by concentrated
solutions of potassic cyanide, and if this solution is subsequently
diluted with water the sulphide of silver is reprecipitated. These
results tend to show that sulphide of silver is not decomposed by
cyanide of potassium, but is held in solution by it as a hydrate.
Similar peculiarities in the behaviour of sulphide of silver are
observed in practice when ores are being treated, and in this
case an increase in the strength of the solution quickens the.
action of the potassium cyanide even though dilute solutions
may be eventually efficacious if enough time is allowed.
A number of experiments made by Louis Janin, Jun.,* on
various salts of silver point to the following conclusions :
Silver chloride is readily soluble in cyanide, and the arseniate

is also
rapidly dissolved. Silver sulphide and antimonide are
less easily acted on, but are not so refractory as metallic (cement)
-silver, for the solubility of which see the table on p. 338. The
*
Enj and Mny. Jown.> Dee. 29, 1883.
presence of copper salts appears to exercise a detrimental action

on the solubility of silver sulphide.
Action of Potassium Cyanide on Oxidised Pyrites.
When the pyrites occur in tailings which have been subjected
to the action of the weather for some time before treatment, com-
pounds are formed which are more prejudicial to the solution than
the sulphides. Sulphide of iron, FeS 2 is oxidised by air and water, ,
ferrous sulphate and free sulphuric acid being formed, thus

FeS 2 + H 2 O + 70 = FeS0 4 + H 2 S0 4
The protosulphate suffers further oxidation, and normal ferric
sulphate (Fe 2 3SO 4 ) is produced, which eventually loses acid and
.
becomes a soluble basic sulphate, Fe 2 O 3 2SO 3 Other basic salts . .
of complex and unknown composition appear to be formed also.

In the presence of such oxidised copper and iron pyrites, the
following reactions take place :
(1) The free sulphuric acid liberates hydrocyanic acid.

H S0 2 4 + 2KCy = K 2 S0 4 + 2HCy.
(2) Ferrous sulphate reacts on the cyanide, forming ferrous
cyanide, which dissolves in the excess of potassium cyanide, so
that it does not appear in the free state.
FeS0 4 + 2KCy = FeCy 2 + K,S0 4
FeCy 2 + 4KCy = K 4 FeCy6 .
The potassic ferrocyanide, if sufficient acid be present, reacts

with fresh ferrous sulphate forming a bluish-white precipitate.
FeSO 4 + K 4 FeCy 6 = K 2 Fe 2 Cy 6 + K 2 S0 4
This precipitate oxidises in the air to Prussian blue if free acid
is
*
present
4K 2 FeoCy 6 + 2 +2H 2 S0 4 = 3FeCy 2 2Fe 2 Cy 6 (Prussian blue) +K 4 FeCy 6
.
+ 2K 2 SO 4 + 2H 2 Q
Both these precipitates are decomposed by potash or soda and
cannot therefore be formed in their presence. The reactions
may be represented as follows :
K 2 Fe 2 Cy 6 -f 2KOH = K 4 FeCy 6 + Fe(OH) 2

3FeCy 2 2Fe 2 Cy G + 12NaOH = 3Na 4 FeCy 6 + 2Fe 2 (OH) 6
.
Consequently, if free acid is not present Prussian blue is hardly

formed at as the solution soon becomes alkaline, and the
all,
is decomposed as fast as it is formed.
precipitate
It follows from these reactions that if the blue colour of
Prussian blue is visible in the vats or on the surface of the
tailings heaps,an enormous waste of cyanide must have taken
place,and the matter should be at once investigated.
(3) Ferric sulphates are decomposed by potassium cyanide,
hydrocyanic acid being evolved and ferric hydrate precipitated.
*
Valentine's Chemical Analysis, p< 42.
(4) A
mixture of ferrous and ferric sulphates produce Prussian
blue by reacting with potassium cyanide, ferrocyanide of potas-
sium being formed at first as above ; the equation is
3K 4 FeCy 6 + 2Fe 2 (S0 4 ) 3 = 3FeCy 2 . 2Fe 2 Cy 6 + 6K 2 S0 4
(5) Sulphate of copper, CuSO 4 acts differently from FeSO 4 ,
,
cuprous cyanide, Cu 2 Cy 2 being formed, soluble in excess ot KCy

,
to K 2
Cu 2 Cy 4 a compound very prone to decomposition. Copper
,
sulphate also gives a precipitate with potassic ferrocyanide,

thus
K 4 FeCy 6 + CuS0 4 = K 2 CuFeCy 6 + K 2 SO. t
(6) Ferrous hydrate, when formed as above, is instantly

dissolved in KCy, thus
Fe(OH) 26KCy = K 4 FeCy + 2KOH( + 17o-6)
+
Ferric hydrate, however formed, does not act on potassium
cyanide, its only action is mechanical, as it collects in a gela-
tinous mass on the filters and checks the flow of liquid.
Copper and zinc in the condition of hydrates or carbonates are
quickly dissolved in preference to the precious metals. If
sulphates of these metals are formed in an ore containing lime-
stone or clay, double decomposition occurs with the production
of sulphate of lime or alumina, and oxides or carbonates of the
heavy metals, which are dissolved by the cyanide, thus
ZnS0 4 + CaC0 3 = ZnC0 3 + CaS0 4
ZnC0 3 + 2KCy = ZnCy 2 + K 2 C03
The Soda Solution. Since acidity of the ore causes decom-
position of the cyanide, an obvious method of reducing the loss is
to add alkali in some form. Before doing this, the free sulphuric
acid and soluble salts may be removed by leaching with water,
and then a solution of caustic soda or lime is run on to the ore,
and after standing for some time is drained off and followed by
the cyanide solution. The insoluble basic salts are thus con-
verted into ferric hydrate and soluble sulphates
Fe 2 3 .2S0 3 + 4NaOH + OH 2 = Fe 2 (OH) + 2Na 2 S0 4 fi
2Fe 2 3 .S0 3 + 4NaOH + 4.0H 2 = 2Fe 2 (OH) c + 2Xa 2 S0 4
Leaching with water then removes the excess of alkali but, as ;
this cannot be done completely, except with considerable expendi-

ture of time, it is usual at the Robinson Mine to use lime instead
of soda. Although the action on the iron salts is slower, an
excess of lime is less detrimental than soda to the cyanide
solution, and does not attack the zinc. It is found that, even
after treatment of the oxidised pyrites by alkalies, the loss of
cyanide is much greater than in the case of free milling ores.
The reason for this may, in part at least, be attributed to the
action of soda on the protosalts (such as sulphates or carbonates)
of copper, zinc, &c., by which these metals are precipitated as

hydrates, readily soluble in KCy. The preliminary washing
with water must always be carefully performed, until no colora-
tion is obtained with ammonium sulphide, so as to remove the
soluble salts as far as possible, but some always remain and are
converted into hydrates by the alkali.
It is now usually regarded as more advantageous to add lime
as a dry powder to the ore before it is charged into the vats,
instead of an alkaline solution. The necessary amount is added
to each truck load of tailings, and is intimately mixed with it by
the time it is charged into the vat.
The amount of alkali to be mixed with a charge of ore is
determined by laboratory experiments, adding little by little an
alkaline solution of known strength to a given weight of the ore,
until the whole is neutral to litmus paper.
Re-Precipitation of Gold and Silver in the Leaching
Vats. If the solution is acid there is a
precipitation of gold
previously dissolved, insoluble aurous cyanide being thrown
down, according to the equation
KAuCy 2 + HC1 = KC1 + HCy + AuCy
This, however, need not be feared as long as there is an excess
of KCy, which must all be destroyed by the acid before the
aurous cyanide can be precipitated. There is danger in trans-
ferring a solution containing gold to a vat containing pyritic
material. If the latter should contain any soluble salts of the
heavy metals, insoluble salts are thrown down, e.g. :
2KAgCy 2 + ZnS0 4 = K 2 S0 4 + ZnAg 2 Cy 4

The salt ZnAg 2 Cy4 probably a true double salt, but the
is
opinion has been expressed that it is merely a mixture of simple

cyanides.
Testing the Strength of the Solution. The ordinary
method of estimating the amount of potassic cyanide present in
a liquid is by titratioii with a standard solution of silver nitrate.
Silver cyanide is formed, and re-dissolves in the excess of potassic
cyanide, until one-half of the latter has been decomposed. The-
equations are as follows :
AgN0 3 + KCy - AgCy + KN0 3

AgCy + KCy = KAgCy 2
When one-half of the KCy present has been converted to AgCy,,
an additional drop of AgNO 3 solution causes the formation of a
permanent white precipitate of AgCy. The amount of silver
solution added is then read off, and the percentage of cyanide
calculated. The equation of the end reaction is
KAgCy, + AgN0 3 - 2AgCy + KN0 3
34 G THE METALLURGY OF GOLD.
A few drops of a solution of potassium iodide are often added

to make the end reaction sharper.
This method is difficult to apply when solutions containing
soluble cyanides of zinc and other metals require to be titrated.
" A white flocculent precipitate occurs at a certain stage, pro-
bably consisting of simple (insoluble) cyanide of zinc, formed by
decomposition of the soluble double cyanide
K 2 ZnCy 4 + AgN0 3 = KAgCy 2 + ZnCy 2 + KX0 3
This precipitation occurs long before the whole amount of
potassium has been converted into the soluble double salt of
silver (KAgCy 2 ), for the solution, after the appearance of the
flocculent precipitate, still gives the Prussian blue precipitate
with acidulated ferrous sulphate." *
Mr. Bettel has devise^ the following methods! of testing solu-
tions for free cyanide, hydrocyanic acid, and double cyanides
respectively.
(1) Free Cyanide. 50 c.c. of solution is taken and titrated
with silver nitrate to faint opalescence, or first indication of a
flocculent precipitate. This will indicate (if sufficient ferro-
cyanicle be present to form a flocculent precipitate of zinc ferro-
cyanide) the free cyanide, together with cyanide equal to 7 '9
per cent, of the potassic zinc cyanide present.
(2) Hydrocyanic Acid. To 50 c.c. of the solution add a solu-
tion of bicarbonate of potash or soda, free from carbonate or
excess of carbonic acid. Titrate as for free cyanide. Deduct
the first from the second result, and the percentage of free
hydrocyanic acid is obtained.
(3) Double Cyanides. Add excess of caustic normal soda to
50 c.c. of solution, and a few drops of a 10 per cent, solution of
KI, and titrate to opalescence with AgNO 3 The zinc potassic
.
cyanide is decomposed by the silver nitrate, and the ZnCy 2 thus

formed instead of being precipitated is acted on by the soda,
sodium zincate being formed and some of the double cyanide
regenerated. This method gives the whole of the combined
CN present, whence the separate results can now be calculated.
According to Watts J and Fresenius, the total amount of
cyanogen in a solution, whether present as simple or double
cyanides, may be estimated by boiling with an excess of oxide
of mercury and water, when all the cyanogen is obtained as
cyanide of mercury and the metals pass into oxides. The
cyanide of mercury is then precipitated by nitrate of silver, with
the precautions recommended by H. Rose and Finkener,||
*
Encj. and Mng. Journ., 1892, p. 417.
t Chem. News, 1895, vol. Ixxii., p. 287.
4: Diet, of Chem., 1864, vol. i., p. 202.
Quant. Chem. Anal, 7th edition, 1876, vol. i., p. 376.

II
Loc. cit.
The amount of gold in the solution is usually determined by

evaporating a known bulk to dryness with litharge, reducing the
lead by fusion in a crucible with charcoal, and cupelling the lead
button. The amount of zinc and other heavy metals present
may be determined by concentrating the solution by evaporation,
" almost
adding an excess of sulphuric acid, and heating until
all the sulphuric acid has been expelled. The residual mass is
then free from cyanogen. It is dissolved in water, if necessary
with the addition of hydrochloric acid, and the oxides deter-
mined by the usual methods. This way is not adapted for
cyanide of mercury, as a little of the metal would escape with
the fumes of the sulphuric acid."
Strength of Solution Required. Below is a table, given
by Mr. J. S. Mac Arthur, showing the relative effect of weak
solutions up to 1 per cent., from which it appears that on
certain ores an extremely weak solution does practically the
same work in gold extraction as one eight times as strong.
While there is a slight tendency to raise the extraction of
silverby the increased strength of solution, the greater ten-
dency of the stronger solutions to attack the base metals is
shown by the fact that, where one of the stronger solutions is
used, the amount of cyanide consumed is equal to the whole
amount present in the weaker solutions.
Butters, to increase enormously the decomposition of the cyanide,

but Mr. MacArthur declares that this increase is not more than
O05 per cent, of the ore, or little more than 1 Ib. per ton.
Methods of increasing the Speed of Action of Potassium
Cyanide on Gold. These methods are described here for
convenience, as being more intelligible after the chemistry of
the process has been discussed. The necessity of the presence
of oxygen has already been dwelt on above. It has, however,
been frequently pointed out that in the interior of a mass of ore
undergoing treatment the conditions are not favourable for the
maintenance of a sufficient quantity of oxygen in a free state.
Both the cyanides and the pyrites of the ore tend to unite with
it, and further absorption of free oxygen from the air is extremely
slow. Hence the time required for the treatment of a charge is
many hours, or even days, although under favourable conditions
tl3 gold could be dissolved in a few minutes, or at most in
two or three hours. To supply the oxygen, various oxidising
substances have been tried. For example, Crosse passed a
current of air through the solution, and the addition of potas-
sium ferrocyanide, of bleaching powder, of hydrogen peroxide,
of manganese dioxide, and of other substances has been made.
Most of these oxidisers were tried by Dixon in 1877* in his
unsuccessful attempt to find a process for treating refractory
ores. These substances hasten the solvent action of cyanides
on metallic gold, but are not used in practice, as they act as
"
cyanicides," destroying large quantities of the solvent by direct
or indirect oxidising effects.
Dr. N. S. Keith suggests f that the action of oxygen is due to
its strong electro-negative relation to gold in cyanide solutions,
and has experimented with various materials which are electro-
negative to gold in solutions of cyanide of potassium. In the
list of such materials, given by Dr. Gore, are carbon, iron, lead,
and mercury (see p. 334). Keith tried the effect of finely-
powdered carbon mixed with the ore, and continually agitated
with it in a cyanide solution. He found that the gold was
more rapidly dissolved than if no carbon had been present, and
supposed that some of the particles of gold came into contact
with the carbon and formed galvanic couples, in which the
electro-positive element, gold, was quickly acted on. Such
contact, however, could not be attained in practice except to a
small extent, and the method is therefore useless, whilst the use
of lead or iron in this way is a fortiori impossible. On the
other hand, mercury can be more readily subdivided and distri-
buted through the ore, but, as it amalgamates with the gold, the
conditions are changed, and, as a matter of fact, the gold in pasty
amalgam is only slightly more rapidly dissolved by cyanide than
is pure gold.
*
Chemical Xew.*, December 20, 1878, p. 293.
t Engineering, vol. lix. (1895), p. 379.
From the fact that mercury is electro-negative to gold in

* concluded that
cyanide solutions, Skey in 1876 metallic gold
in contact with a solution of mercury cyanide would rapidly
dissolve and mercury be reduced. He found this to be the case
alike with gold and silver, which dissolved with almost equal
readiness. In 1895, Keith proposed f to add a small quantity
(2 ozs. to 12 ozs. per ton of liquid) of potassium mercuric cyanide,
HgCy.2 2KCy, to ordinary cyanide solutions to quicken their
.
action by enabling the presence of free oxygen to be dispensed

with. The gold displaces the mercury from solution, and so is
dissolved, whilst the mercury is precipitated on the surface of
the particles of gold and amalgamates them. Keith supposed
that the whole of the gold would thus be rapidly dissolved, and
the precipitated mercury would then be redissolved in the
cyanide, and thus be ready to react as before. This view seems
to be incorrect. According to the author's experiments, the
gold is at first rapidly dissolved, and the mercury precipitated.
As the action proceeds, however, the dissolution of the gold
becomes slower and slower, the mercury appearing to protect
it more and more as the percentage of gold in the amalgam is
diminished. If the particles of gold are only moderately fine
(e.y., gold precipitated
from the solution of the chloride by
sulphurous acid) the action becomes extremely slow after about
85 per cent, of the gold has been dissolved, the amalgam then
consisting of about three parts of mercury to one of gold. If,
on the other hand, very finely-divided gold is used, such as gold
leaf, the action is fairly rapid until about 95 per cent, of the
gold is dissolved, and in one case 98 per cent, of such gold was
dissolved in four days by a solution containing 1*5 per cent, of
HgCy 2 2KCy. In the case of gold leaf, however, which contains
.
both silver and copper, the rate of dissolution is higher than it

would be for pure gold in a similar state of subdivision, as the
presence in the alloy of either silver or copper favours the dis-
solution. The retarding
effect exercised by metallic mercury
when amalgamated with the gold is exemplified by the results
of some experiments in which the solutions contained 1'5 per
cent, of HgCy 2 2KCy, the time of treatment was thirty-six
.
hours, and the weights and state of aggregation of the metal

treated were approximately the same. Under these conditions,
about 86 per cent, by weight of some samples of pure gold were
dissolved, and only 14 per cen-t. of the gold contained in amalgams
consisting of two parts of mercury and one of gold. Dissolution of
gold by solutions containing mercury cyanide is greatly expedited
by heat. In the author's experiments on ores, the quickening
effect of mercury cyanide on solutions of cyanide was very slight.
The Hood Process. I Careful consideration of the known facts
* New Zealand Institute,
Transactions and Proceedings of the. vol. viii.,
p. 334. t Enginefrinfj, loc. ciL

J This section is inserted by permission of Dr. Hood.
regarding the dissolution of gold by cyanide solutions, led Dr.

J. J. Hood, A.R.S.M., to the conclusion that the
generally
accepted theories on the subject are only partly true. He
suggests that gold can only be dissolved by cyanide when some
other metal is present in the solution, which is displaceable by
gold. Thus when gold is digested with solutions containing
an alkaline cyanide, together with certain compounds of
mercury or lead, it is dissolved, and an equivalent quantity of
mercury or lead is precipitated.
If, for example, gold is digested with an aqueous solution of
potassium cyanide and one or other of the chlorides of mercury,

the action is represented, so far as weights are concerned, by
the following equations :
2Au + HgCl 2 = Hg + 2AuCl

2Au + Hg 2 Cl 2 = Hg 2 + 2AuCl.
These equations do not, of course, represent the whole of the
interchanges, The presence of the alkaline cyanide doubtless
plays some part in the dissolution of the gold, and it would seem
to be obvious that the cyanide is instrumental in keeping it in
solution, although there are no published experiments proving
that the dissolved gold exists either wholly or in part in the
form of the double cyanide, KAuCy 2 By using an excess of
.
gold, the whole of the mercury can be removed from solution,

and equivalents of gold dissolved as represented by the equations
given above, 2 x 196-8 parts of gold being dissolved when 200
parts of mercury are added as mercuric chloride or 2 x 200 parts
as mercurous chloride. In some experiments made by the
author employing 0*5 gramme mercuric chloride and 1*0 gramme
potassium cyanide, the amount of gold dissolved by the hot,
strong (5 per cent.) solution in thirty minutes was 0-724
gramme, whilst theory requires 0'726 gramme. Under similar
conditions, the amount of gold dissolved by potassium cyanide
alone was only O'OIO gramme.
Dr. Hood maintains that pure alkaline cyanides could not
dissolve gold, and that they act by means of impurities contained
in them. Suppose, for example, a trace of a metal were present
which could be displaced by gold ; the impurity would be pre-
cipitated like copper by zinc, or reduced like ferric chloride
dissolving tin or zinc, and an equivalent quantity of gold
dissolved. The precipitated metal could not of course be re-
dissolved by cyanide while there remained undissolved any
portion of the particle of gold on which it had been precipitated.
If, however, the impurity were oxidised by contact with air or
other means, it might be redissolved, and would then be avail-
able for the dissolution of a further quantity of gold. The
reduced compound might be similarly regenerated. In this
way a small amount of impurity might suffice for the dis-
solution of a comparatively large amount of gold. It is to
such indirect action that Dr. Hood attributes the efficacy of

oxygen in promoting the solvent action of potassium cyanide.
No doubt, he suggests, ores frequently contain soluble sub-
stances which would increase the solvent action of cyanide on
gold for example, a trace of an oxidised compound of lead
present in the ore but nevertheless to depend on such
fortuitous circumstances is unwise, when a solution efficient
in itself at the start can be used.
One of the solutions used by Dr. Hood in the treatment of
gold ores is obtained by dissolving in water 0*03 per cent, of
mercuric chloride and O06 per cent, of alkaline cyanide. He
finds that the mixture in this ratio is so far stable that the
decomposition of cyanide due to the presence of acid sulphates in
the ore is much less than if potassium cyanide alone were used.
He also adds caustic soda or carbonate of soda to the solution.
In the Hood process the presence of oxygen is quite unneces-
sary, and the treatment of concentrates is thus stated to be
more rapid and cheaper than it is by the Mac Arthur -Forrest
process. The recovery of the gold from solution is effected by
precipitation by means of the copper-zinc couple, which was
described by Gladstone and Tribe. This couple on a small scale
appears to be far more rapid in its action than zinc alone. The
excess of mercury left in solution is precipitated with the gold,
and can be recovered.
Many successful trials on gold ores have been made, and the
process is about to be tried on a working scale in Australia.
Experiments made upon some of the Australian gold bearing
iron oxides and pyritic ores ranging from a few dwts. to 3 ozs.,
as well as upon some Mexican auriferous sands carrying 700 ozs.
of gold to the ton are said to have given a very high percentage
of extraction. According, however, to some experiments made
by the author, it would appear that the process is inapplicable
to ores containing coarse particles of gold.
It is proposed to prepare the double compounds of mercury
from the solutions of the alkaline cyanides obtained in the
manufacture of cyanogen compounds through ammonia and
carbon bisulphide which is now being worked on a large scale
in England. Such double compounds are readily crystallised,
very stable, and easy of transport.
The Sulman-Teed Process. Among other suggestions for ren-
dering the presence of oxygen unnecessary, may be
mentioned
that made by Messrs. Sulman and Teed,* who use cyanogen
bromide, CNBr. The addition of this substance to a solution of
potassium cyanide makes it three or four times
more rapid in
dissolving gold. They put forward the equation
CyBr + 3KCy + 2Au = 2K.AuCy2 + KBr
*
Proc. Institute of Mining and Metallurgy, Feb., 1895.
in explanation of the action of their solvent, which is inopera-

tive except in the presence of an alkaline cyanide. Cyanogen
chloride and iodide give equally good results as far as rate of
solution is concerned, but are not convenient for use on a large
scale. In tests on concentrates, and on various complex ores,
the results obtained by Sulman, Teed, and others were remark-
ably good, high percentages of extraction being obtained in a
few hours from ores which yielded little or no gold to ordinary
cyanide solutions in the same time.
Description of Ores suitable to the Cyanide Process.
Up to the present the ores on which the most striking success
has been obtained on the large scale have been the tailings of the
free milling ores of the Witwatersrand Gold Field, in which the
gold is and the amount of pyrites present is very
finely divided,
small. In a large number of cases in South Africa, pyritic ores
have been treated with great success as for as the solution of
the gold is concerned, over 90 per cent, having been extracted,
but the consumption of cyanide is considerable. Ores contain-
ing sulphide of silver, mixed with base sulphides, often yield no
silver at all. It is, however, no longer doubtful that concen-
trates and other highly pyritic materials can often be treated
more cheaply by cyanide solutions than by roasting and chlorin-
ation.
The presence of decomposing marcasite, especially if some
copper contained in the sulphides, is frequently fatal to the
is
process, as great quantities of cyanide are destroyed by contact

with such ores, the amount often exceeding 50 Ibs. per ton,
even after careful washing with water and treatment with alkali.
However, the concentrates of many mines can be treated suc-
cessfully, even
they consist chiefly of sulphides of iron, lead,
if
zinc, &c. of treatment is in these cases often as much
The time
as three or four weeks, and the charges in the vats are some-
times drained and stirred up or transferred to other vats.
Nevertheless, high percentages of extraction are obtainable
from such materials.
The process is particularly applicable to low grade ores, con-
taining only finely-divided gold and small quantities of base
metals. to the necessity for comparatively coarse crushing
Owing
which exists all wet processes, and the difficulty of handling
with
great quantities of dilute solutions of the precious metals without
loss (which absolutely precludes the use of sufficient solution and
wash water to remove the whole of the soluble gold from any
ore), the percentage of extraction possible with the majority of
ores only amounts to from 70 to 90 per cent. When coarse
particles of gold are present, they must be removed by amalga-
mation, before treatment with cyanide. Ores containing coarse
gold cannot be treated by any wet process, no solution being
sufficiently rapid in its action.
PYRITIC SMELTING. 353
CHAPTER XVII.
PYRITIC SMELTING.
THE best known and most extensively practised smelting
processes for the treatment of gold and silver ores viz., lead
smelting, copper matte smelting, and smelting for the direct
production of copper bottoms, in which the precious metals are
concentrated may be best dealt with in the volumes in this
series devoted to Lead, Copper, and Silver, and will not be
described here. A brief account of iron matte smelting is
appended, however, as its main object is the treatment of purely
* to have
gold ores. This system is said by Eissler originated
in Hungary. It consists in fusing auriferous iron pyrites
in a blast furnace, with the object of obtaining a regulus of
iron, in which the gold is concentrated. The richness of the
regulus, under the original system, is increased by repeatedly
fusing it with fresh ores, or by alternately roasting and fusing

it until the percentage of gold has risen to a certain limit,
which varies in practice, but never exceeds 50 ozs. per ton, after
which the gold is extracted from this product by some other
method, either by roasting and chlorination, or by lead or copper
smelting. In the United States, the production of a rich matte
in one operation is effected by mixing the ore judiciously, and
burning out part of the sulphur in the furnace. It has been
proposed by Mr. "H. Lang f to restrict the use of the
term
" to the reduction of gold and silver ores in
pyritic smelting
blast furnaces, with the formation of a rich matte in one opera-
tion, the distinguishing feature of the work being the use
of the
the
sulphur in the ore as a fuel. This description applies to
system invented by Dr. W. L. Austin, of Denver, and now in
use in several localities in Western America. The definition,
however, is somewhat narrow, and, in this chapter, pyritic
smelting is taken as meaning iron matte smelting.
The method
is especially applicable in districts where no lead ores are to be
obtained, where fuel is cheap, and where there are available large
of gold,
quantities of iron pyrites containing a small quantity
with which purely quartzose ores can be mixed if it is desirable.
Iron pyrites, as is well known, on being heated with a limited
half its sulphur, and is
supply of air, may be made to lose about
then converted mainly into FeS, which is readily fusible. By
*
Metallurgy of Gold, London, 1891, p. 378.
t Eng. and Mng. Journ., Dec. 26, 1891, p. 721.
23
increasing the supply of air, the amount of matte produced may

be reduced to any required extent, the iron being oxidised and
slagged off. If auriferous pyritic ores containing a quartzose
matrix are in course of treatment, a flux must be added to slag
off the quartz. At the Dead wood and Delaware pyritic smelting
works in South Dakota, nearly pure quartzose ores are treated,
and here the flux (in this case pure limestone) is said to cost
about 25 cents per ton of ore.* A matte is" thus formed beneath
the layer of slag, and is drawn off for further treatment.
On account of the comparatively high temperature at which the
matte solidifies, the ordinary blast furnace, with a deep crucible,
used in lead smelting, is thought in Hungary to be unsuitable for
pyritic smelting, as the matte chills as soon as it sinks below the
smelting zone and freezes up the tap-hole, so that tapping is
rendered a difficult operation. Consequently, in Hungary, the
" "
Spur-Ofen has been substituted, in which there is no crucible,
the bottom of the furnace sloping to a tap-hole immediately
below the tuyers. The tap-hole is kept open continually, and
the matte, as fast as it is melted, flows through it by a narrow
channel into wells placad outside the furnaces, where it is
separated from the slag by gravity, and is tapped into moulds,
while the slag overflows into ordinary slag-pots on wheels.
There seems no reason, however, why the methods adopted for
keeping metallic copper molten and in a fit condition for tapping
at Terrazas, Chihuahua, Mexico, by Mr. H. F. Collins,f should
not be equally applicable to pyritic smelting.
The matte thus obtained is sometimes subjected to partial
roasting and then mixed with further quantities of crude auri-
ferous pyrites and smelted again, until the product is sufficiently
rich. An iron matte can be advantageously made richer than
lead without undue losses in the slag, but the limit is reached
when it is worth about 200 per ton. The losses in the slag at
Mineral City, Idaho, and at Deadwood are said to be about $1
per ton in gold and silver together, with mattes of nearly this
value. These mattes are produced by the Austin process, which
was introduced at Toston, Montana, in 1890, and is now in use
at Leadville, Colorado, and at Mineral, Idaho. Details of the
work at these places are given below.
If the sulphides treated contain a high percentage of sulphur,
this is of great value as fuel, and the supply of coke may be
diminished to a corresponding extent. At the Bimetallic
Smelting Company's works, at Leadville, no fuel at all other
than sulphur is used, except an occasional charge of coke when
some irregularity occurs. J JRaw sulphide ores, direct from the
*
Eng. andMny: Journ., Jan. 14, 1893, p. 28.
t Smelting Gold and Silver Ores, Proc. Inst. Civil Eng., vol. cxii. (1893),
part ii.
I Eng. and Mng. Journ., Feb. 4, 1893, p. 99.
PYRITIC SMELTING 355.
mine, are fed into the furnace, with the necessary proportions of
quartzose ore and limestone, to form slag. The furnace is of
peculiar construction, and was designed and built by the Colorado
Iron Works, Denver. The following results, however, were
obtained by the use of an old blast furnace, which had been
previously used for lead smelting. The internal dimensions of
this furnace were 36 inches by 80 inches, and it was altered and
adapted for the Austin system of pyritic smelting. hot-blast A
stove was erected, capable of delivering the required quantity of
air heated to 400C., and the other machinery consisted of one
100-H.P. Buckeye engine, two 40-H.P. boilers, and heavy line
shafting extending through the works conveying power to the
blowers, rock-breakers, slag hoist, &c. This machinery is stated
to have been enough to satisfy the requirements of six blast-
furnaces, but nevertheless it was found that one 40-H.P. boiler
was not quite sufficient when one furnace was at work. In a
trial rim in this furnace, in March, 1892, 1,206 tons of ore
were smelted with 216 tons of limestone as flux in twenty -five
days. The amount of coke burnt was 6 J per cent, of the weight
of the ore, its use being mainly to support the fine ores. The
hot-blast stove was heated by oil, which was also employed to
generate steam. The following is a summary of the cost per
ton of ore smelted in this run :
Coke, at $7-50 per ton, $0'47

Oil, at $1-10 per ton, 0'49
Limestone, at $1-90 per ton (using 18 per cent.), . 0*34
Labour and sundries, 2*62
Total cost per ton, . . . . $3'92
The ores used were stated to contain about 12 per cent, of

sulphur and 1 ton of matte was produced
from every 9 tons of
ore. The richest matte obtained contained gold 0'33 oz., silver
258 ozs., copper 14-2 per cent., the loss being, gold 3-43 per
in the ore. It
cent., and silver 4*15 per cent, of that contained
continuous work on a large
appears that, as a result of this trial,
scale has been begun and is meeting with considerable success.
At calcite
Mineral, ores containing approximately silica 30,
30, iron 10, sulphur 12*5, and arsenic 2-5,
with some zinc, are run
down in one operation into a very rich matte, without the use
of fluxes, and without admixture of other ore, using 7 per cent,
of coke. Most of the sulphur and arsenic is burnt off and the
corresponding proportion of iron
and zinc allowed to go into the
thus a desirable concentration of the matte, and at
slag, effecting
the same time utilising the heat of combustion of the elements
named.* Mr. the manager of these works, also states T
Lang,
that baryta isnot disadvantageous in pyritic smelting, and he
* 244.
Eng. and Mng. Journ., March 18, 1893, I-
id., April 22, 1893.
does not consider that the presence ot even 25 per cent, of heavy
spar would render an ore unsuitable, although it would be
almost hopeless to attempt to treat such an ore by lead smelting.
One of the main reasons for this difference is the fact that
sulphates do not increase the percentage of matte formed in
pyritic furnaces as they do in lead smelting, their acid being
volatilised unchanged in the former case, but reduced by the
coke in lead smelting. Mr. Lang has found in practice, at
Mineral, that the best smelting mixture contains approximately
silica 30 per cent., sulphur from 10 to 15 per cent, (the larger
the percentage of sulphur the less fuel is required), iron 10 per
cent., lime, magnesia, baryta, <fcc., 30 per cent. Afew per cent.
of zinc, lead, copper, &c., do no harm, and the lead and copper
will be retained in the matte. Arsenic is advantageous as it
economises the fuel. The cost of smelting such a mixture, and
refining the matte is about $3 per ton in Western America, at
points conveniently situated on railroads, within a moderate
distance of a coal-field.
Pyritic smelting, for treating gold ores, is as yet in its infancy,
and few details of working have been published by the managers
of the various works. It may possibly be found applicable to
the deep-level pyritic ores in South Africa and elsewhere.
CHAPTER XVIII.
THE REFINING AND PARTING OF GOLD BULLION.
General Considerations. By whatever process gold may have
been extracted from its ores, it is necessary to melt the crude
bullion and cast it into bars so that its value may be ascertained,
and that it may be put into a form convenient for transportation
and sale. The name " bullion " may be conveniently restricted
to the precious metals, refined or unrefined, in bars, ingots, or
any other uncoined condition, whether contaminated by admix-
ture with base metals or not. It is, however, often applied to
coin, and the appellation "base-bullion" is given to the pig-lead
or to copper bottoms or pig-copper, which have been obtained in
smelting operations, and which may only contain a few parts per
thousand of gold and silver, the main portion consisting of base
metals. The treatment of base-bullion, however, properly be-
longs to the metallurgy of argentiferous lead, and copper, and the
descriptions given in this chapter apply only to bullion which
consists chiefly of gold and silver. Refining operations which
REFINING. 357
involve cupellation on a large scale
may also be more con-
veniently considered under the heading of the Metallurgy of
Silver.
The operations to which the retorted metal, gold precipitate
or bars from the chlorination mills, &c., are subjected
may be
summarised as follows :
1. The bullion is melted in crucibles

(a rough refining opera-
tion being usually effected at the same time) and cast in
ingot-
moulds.
2. Assay-pieces are cut from the cast
ingots or dipped from
the molten metal before pouring, and assays are made on these,
by which the value and composition of the bars are ascertained.
3. The bars are then usually sold to the refineries, where the
base metals are eliminated and the gold and silver separated by
"
parting," and cast into bars separately. Both before and after
the parting it is sometimes necessary to subject the bullion to
further refining operations. The bars of gold and silver thus
obtained, being of a high degree of purity, are in a condition
to be used for minting, or for the various industrial purposes to
which they are applied.
Rough unrefined gold is frequently sold to the refineries
attached to the American and Australian mints, in the state of
retorted metal, <kc., without being previously melted and assayed,
the producing mills relying on the good faith of the officials at
these establishments.
REFINING.
Composition of Bullion. Bullion varies greatly in composi-

tion, and gold may be present in any proportion from zero up to
nearly 100 per cent. Native gold always contains more or less
silver, but silver quite free from gold is not uncommon. The
Mount Morgan gold is the finest gold which has yet been found;
this is 997 fine in gold, the alloying metals being chiefly copper
with a trace of iron. The gold obtained in most chlorination
mills is of a high degree of purity and rarely contains much
silver. This precipitated gold, however, generally makes brittle
bars owing to the presence of a few parts per thousand of lead,
bismuth, antimony, and other metals of high atomic volume.
From some chlorination mills the gold is far from pure, owing to
various causes, which include lack of care. If ferrous sulphate
is used as the precipitant, the precipitate may contain large
quantities of ferric hydrate from which some iron is reduced

in
the crucible, and if sulphuretted hydrogen is used and the gold
precipitated as sulphide, it is contaminated with
all the heavy
metals contained in the solution, copper, iron, and lead being
most often encountered. These may amount to several per cent.
Ketorted metal is of very different degrees of fineness, according

to the nature of the ore and the course of treatment. Placer
a
gold is usually finer than that derived from lodes, containing
smaller percentage of silver, while the nature of the material
treated and the methods used in placer operations are not
favourable to the contamination of the bullion with base metals,
which vary in amount only from to 20 parts per thousand and
seldom approach the latter figure. The average fineness of the
placer gold obtained in the United States is given in the following
table, which contains details concerning the chief producing
States *:
'
!
REFINING. 359
Some of that produced in California by the Reese River and

Washoe processes, which are not described in this volume, is
only 500 or 600 fine in silver or even lower, with a few parts of
gold per 1,000, and the remainder consisting chiefly of kad and
antimony, or copper.
The Furnace. The furnace used for melting the bullion is
of simple construction. It is usually square, with walls about
12 inches thick, consisting of an outer layer of
ordinary brick
and an inner layer, at least 4 inches thick, of the best fire-brick.
There is often a complete outer casing of iron, which is useful in
keeping the furnace from falling to pieces, but radiates more
heat than the bricks. The fire-box is about 1 foot square and
from 14 to 18 inches deep ; below it is an ashpit, provided
with a working iron door, through which the air-supply of the
furnace is made to pass, and by which it can be regulated.
The fire-bars are movable, and their ends rest loosely on iron
supports. The top of the furnace may be made flat or
sloping up towards the back at an angle of about 30. In
this case a wide flat ledge should be provided at the front, on
which crucibles and moulds can rest. The top is always made of
a cast-iron flanged plate, with an opening of the same area as
the fire-box. This opening is closed by a cast-iron sliding door
made in one or two pieces, and preferably lined with fire-brick
and running on rollers. The fiue is placed at the back of the
fire-box near the top its cross-section should have an area of
;
about 16 or 18 square inches 6 by 3 inches, and 4 inches square

are both convenient sizes. The flue communicates with a
chimney, which must be of brick for the distance of 2 or 3 feet
from the furnace, but may be of wrought-iron tubing in its upper
part. The height of the chimney will depend on the position of
the furnace, and should be as great as possible, 60 feet giving
better results than any less amount. Some authorities consider
that a height of 30 feet is the minimum that can be allowed in
order to ensure a good draught, but very satisfactory results can
be obtained with a chimney only 16 feet high. The furnace can
be built by any bricklayer acting under directions. No mortar
is used in its construction ;clay, mixed with an equal bulk of
sand, being substituted for it. A sliding damper in the flue at
a convenient height above the ground is necessary, so as to
regulate the draught. The fuel used in such a furnace may be
anthracite, charcoal, or good coke, made in ovens, not in gas-
retorts, and broken into pieces of about the size of a hen's egg.
If the coke is of high quality, it is the most satisfactory fuel,
making a hot fire and lasting for a long time, so that it does not
require very frequent replenishing. Neither dust, nor very
small, nor very large pieces must be used. Charcoal is preferred
in the United States Mints for small charges, and anthracite for
large ones.
The Crucibles. The bullion is melted in either graphite or

clay crucibles. If nitre is used to refine the metal, the graphite
pots are sometimes coated inside with clay. The amount of
refining that can conveniently be done in this way is limited
by the fact that the molten oxides produced rapidly corrode the
crucibles, and may perforate them in course of time, thus causing
loss. The Salamander crucibles, manufactured by the Battersea
Crucible Company, are the best graphite pots, as they require
very littleannealing, and will stand frost without being dis-
integrated. A No. 20 crucible, of 9 inches in height, and
holding 400 to 500 ozs. of bullion, can be used in the furnace
described above. The size of the crucibles and the weight of the
charges of bullion vary greatly, but in extraction mills, as a
general rule, a gold-charge does not exceed 400 ozs., and a silver-
charge 1,200 ozs. in weight. In mints and refineries, much larger
crucibles are employed, holding different amounts up to 6,000
ozs. of metal.
Melting the Bullion. All crucibles must be thoroughly
annealed before being used ; otherwise, the contained moisture
being suddenly converted into steam when the crucible is heated
rapidly, the pots fly in pieces. The crucible is kept on a shelf
near the flue, for as many days or weeks as convenient, before
being used. It is then placed on the top of the furnace or in the
ashpit for a few hours, when it will probably be safe to hold it
over the open furnace by means of the crucible tongs, until it
becomes gradually warm. After a few minutes, the crucible
being turned round at intervals, it can be lowered rim down-
wards upon the burning fuel, and as soon as the rim becomes
red-hot, the crucible is quite safe, and may be turned over and
placed in position for the reception of the gold. With Sala-
mander crucibles, a less degree of care in annealing will suffice,
as they are well annealed before being sold. The crucible rests
on a fire-brick about 3 inches thick, which is laid on the bars of
the grate. If the tire-brick were omitted, the bottom of the pot,
resting directly on the fire-bars, would be too cold, while a layer
of fuel, if placed below it, would soon be burnt out, and could
not readily be replaced, so that the pot would sink down to the
bars. The fuel is built up round the pot until it reaches to
its rim, and the fire urged slightly until the whole pot is at a full
red heat. One or two spoonfuls of borax are then thrown into
the crucible by means of a scoop or wrapped in paper ; this
flux not only assists the metal to fuse, but slags off the earthy
impurities, and makes the metallic oxides more liquid. As soon
as the borax is melted, the introduction of the bullion is com-
menced. The safest way to do this is to use the shoot shown
in fig. 57, which is held in position, its lower edge being
inside the crucible, with the left hand, while the metal is
transferred to it in a scoop by the right hand. In this way the
REFINING.
melter avoids all danger of loss which

might be encountered if
the metal scrap were wrapped in
paper and added by the ton^s.
Large pieces of metal are added by the crucible The
tongs.
cover, which must also have been
previously well annealed, is
kept on the crucible as much as possible, and the fuel pushed
down with the poker to avoid scaffolding, and fresh
pieces of
coke added when required. The crucible is not allowed to
become more than two-thirds full at any time, but more metal is
added when the first supply has been melted
down, and the
operation repeated until the pot is sufficiently full of molten
material.
Fig. 57.
Scale, 1 in. = 9 ins.
Refining the Bullion. If the bullion is of a high degree of
purity, containing but little dirt or base metals, not much flux
is added, a spoonful or so of carbonate of soda and nitre
being
enough. In this case the slag is not skimmed off but poured
with the metal. If the bullion is very base, however, it is usual
to refine it partially by adding nitre and borax, a little at a time,
and skimming off the slag when all action has ceased. The nitre
exercises a powerfully oxidising effect on the base metals in the
bullion, and the resulting oxides form a liquid slag with the
borax. When graphite crucibles are employed, the nitre must
be prevented from coming in contact with the sides, as in that
case the carbon would be oxidised and the pot rapidly corroded.
On the other hand, clay pots do not withstand the action of
molten oxides slagged with borax. For these reasons a favourite
practice is to use graphite pots, covering the surface of the molten
metal with bone-ash sprinkled on, the layer being thickest round
the sides. Holes are made near the centre of this cover with an
iron rod, and nitre introduced through them, in small amounts
at a time. As the fusible oxides are formed they are absorbed
by the bone-ash and prevented to some extent from attacking
the crucible. When sufficient nitre has been added, a point

which is judged by an experienced melter from the appearance
of the surface of the molten metal, the slag should be of a moder-
ately pasty consistency suitable for skimming. If it is too liquid
it is difficult to skim and must be thickened with bone-ash if it ;
is too pasty, shots of gold may become entangled in it, and it

must be thinned with borax.
Mr. Hanks, the late State mineralogist, describes the operation
of skimming, at the San Francisco Refinery, as follows * " : A
skimmer is prepared by bending the end of an iron rod of a
inch diameter into a spiral of about 1J inches in diameter,
shaped so that when the skimmer is let down vertically into the
crucible, the spiral will lie flat upon the surface of its contents.
When the slag and metal are perfectly fluid, the surface of the
former is touched by the skimmer, to which some slag adheres.
It is then withdrawn, quenched in a bucket of water, and at
once replaced in the crucible, thus causing a further small
portion of the slag to solidify. This operation is repeated until
the greater part of the slag has been removed from the crucible.
Care must be taken not to allow the bare iron to come in contact
with the molten gold, as in that case some of the latter would
adhere to it, and for this reason the slag is left adhering to the
skimmer during the latter part of the operation. The wet
skimmer must not be plunged below the surface of the molten
metal or an explosion would ensue. If too much slag accumu-
lates on the skimming tool it is detached by quenching and
hammering."
If the bullion is very base the addition of bone-ash, nitre, and
borax, followed by skimming, may have to be repeated two or
"
three times. The method was formerly known as the " Poussee
process. Lead is not readily removed from bullion by the action
of nitre, which is best adapted for the oxidation of copper.
When much lead is present, alternate additions of sal-ammoniac
and nitre are made, by which the lead is rapidly oxidised. It is
probable that the sal-ammoniac acts by decomposing basic com-
pounds of lead which resist the action of nitre. In obstinate
cases, a blast of air is directed upon the surface of the molten
metal, and the lead is in this way rapidly oxidised and slagged off.
Sometimes sal-ammoniac is sprinkled on, to remove lead, aftei
skimming.
Other fluxes which are sometimes used are sand, pearlashes.
and metallic iron. Sand is added to assist in forming a liquid
slag, and to protect the crucible from corrosion by the oxides,
especially when iron is present. Pearlashes are sometimes added
when the bullion contains tin.
Bars which contain antimony or arsenic can be rapidly refined
by stirring briskly with an iron bar, a little nitre being added to
*
Californian State Mineralogist's Report, 1884.
REFINING. 363
the charge. After three or four minutes'
stirring, the greater
part of the antimony will have been removed as antimonide of
iron.
The refined gold shouldnow be of a brilliant green colour, and
its surface should remain quiet, without
showing any iridescent
films or other signs of continued oxidation.
Toughening the Bullion. After melting with nitre, bullion
is sometimes toughened before
being poured, as small quantities
of lead, antimony, arsenic, and bismuth are still retained and
render it brittle. Nearly fine gold, which is the product of
chlorination mills or of parting operations, is
similarly treated.
The toughening is usually done in one of three ways, viz. :
1. Sal-ammoniac and corrosive sublimate are added to the
molten metal.
2. The metal melted with oxide of copper.
is
3. Chlorine passed through the molten metal.
is
The method of procedure in each case is as follows :
1. Sal-ammoniac is
sprinkled on to eliminate the lead and tin,
after which repeated small additions of powdered corrosive sub-
limate (mercuric chloride) are made. After each addition the
door of the furnace must be at once closed, as dense poisonous
fumes arise and must not be breathed by the workers. Volatile
chlorides of zinc, copper, antimony, bismuth, <fcc., are formed and
pass off, carrying with them some gold, of which there is an
appreciable loss. A
little corrosive sublimate sprinkled on the
surface of molten gold will completely toughen every part of it
without being mixed with it by stirring, even although the
crucible contains several hundred ounces of the metal.
When the metal is supposed to be tough, a small sample
is dipped out and made into a thin ingot, which, after it has
been cooled in water, is doubled up by hammering and its
degree of toughness thus tested. It is then often remelted with
copper to make up the standard alloy of the country, and again
cast and hammered or cut in two with a shearing machine. The
reason for doing this is that impure gold, although it may be
tough when unalloyed with copper, may make brittle standard
bars. If the gold is still found to be brittle, the main bulk of
it left in the crucible is
subjected to a repetition of its former
treatment as often as is necessary, and as soon as the toughening
is complete, the
gold is covered with a layer of charcoal in the
form of powder or lumps and thoroughly stirred before being
poured.
The melting under charcoal is sometimes necessary to render
silver bars for coinage when they have been treated for a long
fit
time by nitre. When

silver has been raised to a high degree
of fineness, it is affected by a peculiar bubbling which may be
due to the evolution of oxygen previously absorbed from the
nitre. If this continues it is necessary to stir continuously
with a graphite rod, keeping the surface covered with charcoal

powder, until the bubbling ceases. If the metal, while still
effervescing, is poured into a mould, it sprouts at the surface,
and a shower of extremely minute particles are projected, often
to some distance from the mould, requiring to be swept up ;
if the crucible is covered by a lid, very heavy effervescence
ensues when the lid is lifted. In this case the silver ingots
formed are not marketable, being brittle, of low density, and
covered by heavy efflorescences.
2. Black oxide of copper is less frequently employed. It is
stirred in with the molten metal, and the whole then allowed to
remain in the furnace for about half an hour, with occasional
stirring. Antimony, arsenic, bismuth, &c., are oxidised at the
expense of the oxide of copper, and volatilised or slagged off
with borax. It is stated that 2 per cent, of antimony can thus
be removed. The process is efficacious, but the pot is rapidly
corroded, and the gold is of course contaminated with the reduced
copper, a matter which, however, is of little consequence, for the
reason that the pure metal is seldom required.
3. The use of gaseous chlorine is described under the heading
of Miller's process of Parting, p. 383.
The time occupied in refining and toughening a crucible full
of gold of course varies greatly, but often occupies from one to
three hours after complete fusion has been effected.
Casting the Ingots. It is necessary to stir the charge
thoroughly before pouring, as the bar must be as homogeneous
as possible to insure a correct assay. Since segregation may occur
on cooling, assay pieces are often dipped out immediately alter
stirring. The subject of taking assay pieces in mrther con-
sidered in Chapter xx., p. 432. The stirring is usually done
with a peculiarly shaped graphite rod, made expressly for the
purpose. It is annealed carefully and raised to a full red heat
before being introduced into the crucible, and is held firmly by a
pair of tongs with special concave curved faces to its jaws, so as
to fit the round rod. In the case of very small meltings it is
sufficient to lift the crucible out of the furnace with the tongs
and to give it a rotary motion just previous to pouring. In
doing this, the metal must not be allowed to cool too much or
the casting will be defective. It is advisable to close the
dampers wholly or in part, so as to check the draught, when
stirring is being done.
Meanwhile the ingot-mould in which the gold is to be cast has
been prepared. It is cleaned thoroughly inside by rubbing with
emery paper and oil, or with pumice stone, and wiping with
an oily rag ; it may also be blackleaded inside, as this prevents
contact between the gold and the iron of the mould. It is then
warmed by being placed on the top of the furnace its temper-
;
ature must not be sufficiently high to ignite the oil, but it should
REFINING. 365
be too hot to touch with the hand. When the bullion is

ready
to pour, the mould is placed on a level surface, such as an iron
stool, at a height above the floor of about 12 or 18 inches and oil
poured into it. Any cheap non-volatile oil will do, whether animal
or vegetable. The mould is tilled to a depth of about a quarter of
an inch with oil, which is made to flow over all parts of the
interior.
The crucible is then carefully lifted from the furnace, usually
with basket tongs, and the contents poured rapidly but steadily
into the mould, the crucible being moved to and fro so that the
stream of molten metal is directed to all parts of the mould in
succession. The crucible is then held in the inverted position
for a short time, and jarred once or twice to cause the last portion
of the metal to flow from it. The oil is ignited, and burns on the
top of the cast metal, thus keeping it from tarnishing. In small
castings, the slag is allowed to flow out and remain on the top of
the metal in the mould ; in large castings, the slag is usually
skimmed off before pouring. Beads of metal are caught in a
large iron tray with raised edges, in the centre of which the
mould is placed. If the mould is clean and has been hot enough,
and if enough oil has been used, a clean untarnished bar is pro-
duced. It is turned out of the mould by inversion of the latter,
while still too hot to be handled, and the slag is separated by one
or two light taps with a hammer. The bar is then, in many
establishments, momentarily dipped into water to assist in the
complete removal of the last fragments of the slag, and it is also
a favourite practice to dip the bar, first into dilute sulphuric acid,
and then into clean water, the bar retaining warmth enough
after removal to expel all moisture. This treatment removes all
tarnish, and any adherent particles of slag are then chipped off
and assay pieces cut from the bar.
In some refineries large crucibles are used, and 3,000 or 4,000
ozs. of metal are refined at once. In this case, several ingot
moulds are filled successively from one pot, the weight of the
gold bars manufactured being usually either 200 or 400 ozs. each.
Silver bars, on the other hand, are made much larger, often
weighing 1,000 or 1,200 ozs. Mixed bars of bullion (containing
both gold and silver) are seldom cast of a greater weight than 600
ozs. in mills.
The method of refining which has been described above is
seldom attempted in extraction mills, and is often unnecessary
in refineries, before parting the silver from the gold. The object
at a mill is merely to melt down the bullion obtained, so as to
as possible.
bring it to a marketable form with as little loss
With this object in view, no nitre is used, but the metal is kept
covered by a layer of charcoal to prevent the formation of oxides,
and the crucible is poured as soon as the charge has been fused
and stirred.
In conducting all these furnace operations the use of a thick

pair of mittens, made of sacking or rubber, is to be recommended
for the protection of the hands.
Losses of Bullion Incurred in Melting. The losses sus-
tained in melting vary according to the composition of the
metal, but are usually very small. At the New York Assay
Office they are found to vary from -
5 to 1*5 per 1,000. The
losses may be divided into mechanical loss, and loss by volatili-
sation. The mechanical loss is reduced by care in the conduct
of the operation ; it may be due to a number of causes. The
crucible may break and its contents fall into the fire, or be
scattered over the floor of the melting house when on the point
of being poured. To avoid loss in this way, the ashpit may
be constructed of a cast-iron tray, which can be easily scraped out.
The floor of the room is made of carefully-laid flagstones, or,
better still, of iron plates, in which there are no cracks or
crannies capable of hiding metal beads. Projection by spirting out
of the crucible may be occasioned if certain impurities, such as
tellurium or antimony, are present in the bullion, and if the
nitre is added in large quantities at a time, the pot may "boil
over," and part be lost. When flux is to be added, the surface of
the metal must be at least 6 inches below the rim of the pot.
Recovery of any metal lost in the ashes is effected by panning.
The slags formed in the course of refining frequently contain
some small shots of metal, which may be recovered by grinding
the slag finely and washing down the product in a pan or on a
vanning shovel. At the San Francisco Mint, the residue from the
vannings are allowed to accumulate, and at intervals dried and
fused with borax in an old graphite crucible at a high tempera-
ture. The crucible is left in the furnace over night to cool, and
is then broken up. arid the shots of metal at the bottom picked
out. They are chiefly silver, very little gold being thus
recovered. Shots of metal can be recovered with greater
certainty from the slags by fusion with lead.
The crucibles, stirrers, lids, &c., also contain a certain quantity
of gold and silver. After each melting they are scraped, and the
scrapings panned, or, better still, calcined and fused with lead;
but, in spite of this treatment, precious metals accumulate in
the pots, and, when worn out, they are ground-up in a Chilian
mill or other grinder, and panned, in order to separate the shots
of metal. The tailings from this treatment will often pay for
fusion with lead, and subsequent cupellation. Certain refineries
treat large quantities of such residues, with which may be
included sweepings of the floor of the melting house.
At the Philadelphia Mint it is found worth while to recover
the gold from the iron tools used in stirring, dipping, &c. For
this purpose they are melted down in a graphite crucible with a
little charcoal to make grey-iron, and kept at a white heat for
REPINING. 367
some time, after which the charge

is allowed to cool
slowly.
Under this treatment the gold and
silver separate out (an alloy
containing three or four parts of silver to one part of gold being
better for the purpose than pure gold), and are found at the
bottom of the crucible sharply marked off from the surface of the
iron, which is now
quite freed from the precious metals.*
The mechanical loss in melting at the San Francisco mint,
where crude bullion is bought and valued, is stated to be only
about 005 oz. in a melt of 100 ozs. This is only one part in
-
twenty thousand, but the total- loss of gold is usually much

greater, and, as the crude bullion cannot be accurately valued
before it is melted-up, it is always difficult to determine how
much loss has actually occurred.
The loss by volatilisation is often more serious than the-
mechanical loss. It is increased by the passage of a rapid
current of air over the surface of the molten metal; to avoid
this, the fire-box is made deep enough to allow the top of the
pot to be sunk some distance below the flue, and the exposure
to the draught coming from the opening in the top of the furnace
is thus diminished. The metal is also sheltered by being kept
covered by the crucible lid as much as possible, and covers of
charcoal powder, of fragments of charcoal, or of bone-ash are in
general use. For the effect of different alloying metals on
the volatilisation of gold, see p. 7. Napier found f that much
volatilisation took place during pouring, so that if a wet beaker is-
held inverted over the mould during pouring, gold is condensed in
it and can be subsequently detected by the ordinary tests. As for
other conditions, it appears from the results on p. 6 that the
longer the gold is kept melted, the higher the temperature em-
ployed, and the smaller the mass of gold, the greater will be the
percentage loss by volatilisation. These results are in accordance
with the experience of those engaged in the operations on a large
scale.
The loss by volatilisation when the gold is toughened by the
use of corrosive sublimate is particularly high ; the total loss of
gold incurred when bars are subjected to this treatment is.
stated to be on an average about 0-85 per 1,000 or 850 per
million sterling, while the ordinary loss on melting, both
mechanical and by volatilisation, only amounts to about 0'15 or
0-17 per 1,000, part of which is recoverable.
In order to condense the gold and silver which are carried off
by the volatilised copper, Mr. J. Feix, foreman of the refining
department of the San Francisco Mint, devised an ingenious
flue-dust chamber, which is used there as an adjunct to the
melting furnaces. J The chambers are built of brick, and the
*Egleston's Metallurgy of Silver, Gold, and Mercury, vol. ii., p. 728.
tCV/em. Soc. Journ vol. x., p. 229.
,
J Ninth Report Col. State Min. 1SS9, p. 66.

,
horizontal flues, coming from the fire-box, open almost directly

into them. The chambers are 12 feet high, 1 foot wide, and of con-
siderable length ; the flues from them are placed within 1 foot
of the top and communicate with the stack. Iron doors, placed
at the end of the chamber, enable the sweepings and dust to be
readily withdrawn. It is stated that almost all the precious
metals which have been volatilised are condensed in this w ay. r
The sweepings are heated with borax to a high temperature in

an old graphite crucible, when the metal accumulates at the
bottom as a regulus. It is not stated how much gold and silver
is recovered in this way, and whether the amount has paid for
the erection of the chambers and the extra labour and fuel
required.
The loss of weight in melting will of course be frequently
much greater than the actual loss of gold. Thus mercury, if
present, will be expelled by volatilisation, and this metal usually
forms from 0-5 to 1 part per 1,000 of retorted gold. Zinc is not
so readily expelled, but, if any is present, some will be volatilised,
the amount depending on the temperature and duration of the
melting. Moreover, the earthy material and all the base metals
which have been oxidised by the nitre will be slagged off by
the borax, and the total diminution in weight may thus amount
to several per cent.
Refining by Means of Sulphur is said to be practised in
the United States Mints ;* the following is a brief account :It
is effected in plumbago crucibles, and has for its main object the
elimination from retorted metal of iron, when, as sometimes
happens, it is present in large quantities. The metal is kept just
above its melting point, the temperature being as low as possible
in order to avoid unnecessary waste of sulphur by volatilisation.
Sulphur is sprinkled round the edges of the molten mass, and
stirred in with a graphite stirrer. If sulphur is added near the
centre, particles of gold are lost by projection. Sulphide of iron
is formed with great energy, and sulphide of silver also, but the
latter is not produced rapidly until nearly all the iron has been
already converted into sulphide. The gold is unaffected by the
sulphur and subsides to the bottom. It is not usually cast by
pouring, but allowed to solidify in the pot, a better separation
between the gold and the matte being thus effected. The pot is
turned out as soon as solidification has taken place, and the
matte is broken off by a hammer, the gold being remelted and
cast into a bar. The small quantity of gold taken up by the
matte is separated by melting with metallic iron.
Osmiridium in Gold Bars. Gold from some districts of
California and from the Eraser River district of British Columbia
contains some platinum and palladium, and frequently some
osmium and iridium. As these metals remain together during
* Ninth Report CaL State Min., 1889, 6<l.
p.
PARTING. 309
the treatment, the mixture is commonly called osmiridium. If

this is present in perceptible quantities, a fact ^hich is not
usually detected until after the bars have been parted, the gold
is remelted in a clean crucible and
kept fused at a high tempera-
ture for about half an hour, when the osmiridium will settle
to the bottom of the crucible. This is due to the fact that
osmiridium does not seem to form a true alloy with gold, and,
being of high density and very infusible, the particles, unfused
or partially fused, settle through the liquid gold. The crucible
is then gently lifted out of the furnace, and the
greater portion
carefully but rapidly poured into a mould the remainder, which
;
contains almost all the osmiridium, is allowed to cool in the

crucible until it has solidified, and then assayed for osmiridium.
An alternative plan is to allow the whole charge to solidify in
the crucible, and then to cut off the lowest portion, which is set
aside. The osmiridium settles better from an alloy chiefly
consisting of silver than from pure gold, and the rich bottoms
are consequently melted several times with silver, the lowest
part being cut off each time. The gold is thus gradually
replaced by silver, which eventually forms by far the greater
part of the mass. It is then granulated and parted, and the
resulting powder of gold and osmiridium is treated with aqua
regia, by which the gold is dissolved, and the osmiridium
separated as a black powder. Such osmiridium is worth from
eight to twenty shillings per ounce, and selected grains are used to
make the "diamond points" of gold pens. Iridiuin is, however,
not invariably separated from the gold bars in which it is
contained, and traces can be observed in a considerable pro-
portion of the refined commercial bars met with in London.
Platinum and palladium are, in great part, extracted by the
nitre, and enter the slag.
PARTING.
Parting is the separation of silver from gold. During the

course of the operation the base metals are separated from both,
but, as the presence of these base metals is injurious to the
successful conduct of the processes which are chiefly in use, a
preliminary refining by one of the methods already described is
usually necessary. Only about 10 per cent, of base metals is
permissible in the alloys when sulphuric acid is used, although
a somewhat larger quantity does no harm if nitric acid is em-
ployed.
The processes of parting may be tabulated as follows :
1.Cementation.
2. Melting with sulphide of antimony.
3. Melting with sulphur, and precipitation of the gold from
the regulus by silver, iron, or litharge.
4. Parting by nitric acid.

5. Parting by sulphuric acid.
6. A combination of these last two methods.
7. The Gutzkow process.
8. The new Gutzkow process.
9. Parting by chlorine gas.
10. Parting by electrolysis.
The first two of these methods were *mown to the ancients,
and the third was described as early as the beginning of the
llth century.
Cementation. In this ancient and obsolete process, gold was
freed from small quantities of silver, copper, &c., contained in it.
The method was mentioned by Pliny and described by Geber,
who wrote in Arabic, probably in the eighth or ninth century ;
it is possibly still in use in Japan and in some parts of South
America. It consists in heating granulations of argentiferous
gold mixed with a cement, consisting of two parts of brick-dust,
or some similar material, and one of common salt, in pots of
porous earthenware. The temperature used
is a cherry-red heat,
which is melt the granulations.
insufficient to After about
thirty-six hours' treatment, the greater part of the silver is
converted into the state of chloride, and this, together with the
cement, can be removed from association with the granulations
by washing with water. The gold can in this way be raised to
a fineness of about 850 or 900. The silver is recovered from the
cement by amalgamation with mercury.*
Parting by Means of Sulphide of Antimony. This pro-
cess was also used to purify gold which contained only small
quantities of silver. The alloy was repeatedly melted with
sulphide of antimony, upon which the gold became alloyed with
the antimony and sank to the bottom of the mass, while the
silver was converted into sulphide and floated on the top, mixed
with the excess of antimony sulphide added. The gold was
subsequently refined by a blast of air directed upon it, the
antimony being thus oxidised and volatilised. The method is
now obsolete, but was in use at the Dresden Mint up to the
year 1846, and gold of the fineness 993 was said to be produced
in this way.
Parting by Means of Sulphur. This method was formerly
used for the purpose of concentrating the gold contained in
auriferous silver in order to obtain a richer alloy. The granu-
lated alloy was melted with sulphur and some of the silver
was thus converted into a matte. The gold was then precipitated
from the matte and collected in a smaller quantity of silver by
fusion with pure silver, or with iron, or litharge. No attempt
* For a full account of this
interesting process, as well as of the next
succeeding two methods, the student is referred to Percy's Metallurgy of
Silver and Gold, pp. 356-402.
PARTING. 371
was made to obtain pure gold in this way, and the enriched
alloy of gold and silver was parted by nitric acid. The silver
was recovered from the matte by fusion with iron. The method
was in use in several refineries in Europe at the beginning of the
present century. The employment of sulphur in refining at the
United States Mints has been already noticed, p. 368.
Parting by Nitric Acid. The first clear mention of the use
of nitric acid for parting silver from gold is made by Albertus
Magnus, who wrote in the thirteenth century, but the process
does not appear to have been employed on a large scale until two
centuries later in Venice. Here, according to an old tradition,*
some Germans were employed in separating gold from Spanish
silver in the fifteenth and sixteenth centuries, the art being kept
secret. These refiners were not inaptly named "gold makers "
by those who were unacquainted with their methods. The pro-
cess was fully described by Biringuccio in his treatise,! published
in 1540. and by Agricola J in 1556. It was first used in the
Paris Mint about the year 1514, and in London at least as early
as 1594, but for a long period the operations were conducted in
secret in both countries, and it is supposed that this method of
refining was not fully practised in England until about the
middle of the eighteenth century.
Parting by means of nitric acid is conducted on the large scale
in the same general manner as in the assaying of gold bullion.-
It consists of the following operations :
1. Granulation of the alloys.

2. Solution of the silver in nitric acid.
3. Treatment of the gold residues, viz. :
Sweetening by washing with
water, drying, melting, and casting into bars.
4. Precipitation of the silver as chloride by salt solution.
5. Reduction of the silver chloride by zinc and sulphuric acid.
Granulation of the Alloys. The gold to be parted must be

approximately free from, base metals, particularly from those
which are not soluble in nitric acid, such as tin, arsenic, anti-
mony, &c. If these were present they would form insoluble
oxides, which would remain with the gold, so that further refining
operations would be necessary: they would, moreover, cause a
great increase in the consumption of nitric acid, so, if they are
present, the gold is freed from them as far as possible by melting
with nitre, &c. Copper, lead, and other metals which are
readily soluble in nitric acid are less obnoxious, and small
percentages of these are allowed to remain, as they are not
difficult to separate from the silver when in solution with it,
while the presence of copper in particular is advantageous in
promoting rapid dissolution of the alloy. If present in large
*
Beckmann's History of Inventions, vol. iv., p. 578.
+ Da la Pirotechtiia, Florence, 1540.
1 De re Metallica.
quantities, however, even these metals would create difficulties

and expense, increasing the consumption of acid.
The bars are melted together to form an alloy which, it was
formerly believed, must contain one part of gold to three parts
of silver (hence the term "inquartation" applied to this process).*
This proportion is still adhered to in many English and European
establishments, and at the Philadelphia Mint. In some refineries,
however, the proportion of silver used is less, the minimum
being two parts to one part of gold. Dore bars containing small
quantities of gold are, of course, preferred to bars of fine silver
for the purpose of alloying with the argentiferous gold bars.
After the "inquarted" alloy has been thoroughly mixed by being
stirred while still in the furnace, the crucible containing it is
lifted out, and the metal is poured into copper tanks filled with
cold water, which is sometimes kept cool by ice, while, in some
refineries, a stream of water is kept constantly flowing through
the tank. The metal is poured with a circular and wavy motion
in a thin stream to prevent the formation of lumps ; leafy
granules and small hollow spheres are thus formed. The
pouring is done either directly from a crucible or from a dipper,
the vessel being held in either case about 3 feet above the
.surface of the water. In the tank is a perforated copper pan,
which is lifted out when the pouring is completed, and the
granulations allowed to drain.
Dissolving the Granulations. The granulated metal is heated
with nitric acid in vessels of earthenware, porcelain, or platinum.
The earthenware vessels are usually cylindrical. Those in use
at the Philadelphia Mint are 21 inches in diameter and 22 inches
deep, and contain 1,500 ozs. of granulations they are placed on
:
.a lattice work of wood, which is laid on the bottom of lead tanks,

and are surrounded by water 10 inches deep kept at the boiling
temperature by means of steam. The earthenware vessels are
covered by a closely -fit ting lid provided with a delivery tube to
-carry away the fumes. Messrs. Johnson & Matthey's platinum
vessels, which hold 800 ozs. of metal, are heated by separate
furnaces, the fuel being coke or coal gas.
The strength of the acid used varies from that of spec. gr. 1-33
(38 B.) to that of spec. gr. 1'2. When nitric acid alone is
used, about 3 Ibs. of acid (spec. gr. 1-2) are used to dissolve each
pound of granulations, but of this quantity the amount used in
the last boiling (about 20 per cent, of the whole) is available for
* Asmaller proportion of silver, however, was used at least as early

as the year 1627 in Paris. Thus Savot observes in his Discours sur lea
Medalles Antiques, Paris, 1627, chap, vii., p. 72: "S'il n'y a beaucoup
plus d'argent que d'or, Teau n'agira aucunement : de sorte qu'il faut qu'il
y ayt au moins les deux tiers d'argent, et un autre tiers d'or, et encore que
1'eau soit tres-bonne :
car, si elle est foible, elle n'operara point." Savot
did not seem to regard this proportion of 1 to 2 as of recent introduction.
PARTING. 373
further use; more acid is required if there is much

copper
present. The first addition of acid is of about 1| Ibs. to each
pound of metal it is kept boiling gently for about five or six
;
hours, by which time most of the silver will have been dissolved.
The solution is allowed time to settle, and the hot
supernatant
liquid is siphoned off by a gold or glass siphon, and diluted with
water to prevent the formation of crystals on The
cooling.
second addition consists, in some establishments, of
strong acid
(specific gravity 1-414), and in others of acid of the same strength
as before. The second boiling is for two or three hours only,
and the third boiling for only one or two hours, the liquid
being
siphoned off after each boiling.
The vessels are provided with hoods and small chambers in
the delivery tubes, in order to effect a partial condensation of
the acid, and also to recover the small amount of silver nitrate
which is carried over mechanically, owing to the violence of the
disengagement of gas bubbles. The fumes are conducted to the
melting furnace where they are consumed, giving up their
oxygen to the fuel.
The reactions that occur are partially expressed by the follow-
ing equations :
6Ag + 8HN0 3 = 6AgN0 3 + 2NO + 4H 2

3Cu + SHN0 3 = 3Cu(N0 3 2 + 2NO
) + 4H 2
4Zn + 10HN0 3 = 4Zn(N0 3 2 + N2
) + 5H 2
and similar reactions for other metals. The amount of nitrous
oxide evolved increases towards the end of the operation. It is
seen that silver decomposes less than its own weight of nitric
acid, while copper and zinc destroy nearly three times their
weight of the acid. Nitric acid of specific gravity 1*2 contains
about 32 per cent, of anhydrous HN0
3
so that the quantity of
,
acid of this strength theoretically required to dissolve 1 Ib. of

silver, copper and zinc is about 2-4 Ibs., 8-3 Ibs. and 7'6 Ibs.
respectively.
Treatment of the Gold Residue. The pulverulent gold is
" sweetened "
by being washed thoroughly in perforated earthen-
ware dishes with boiling distilled water, stirring being performed
with a spatula of wood, platinum, or porcelain. The gold is thus
freed from nitric acid and nitrate of silver, the operation being
continued until the washings show no signs of turbidity on the
addition of salt. The washings are added to the first silver solu-
tions, serving to dilute them, the dilution, as has already been
observed, being necessary to prevent crystallisation on cooling.
The sweetened gold is generally pressed, dried, melted, and cast
into bars, which are now made of a weight of either 200 or
400 ozs. The gold thus obtained is usually of a fineness of about
997 or 998, the remainder being chiefly silver, which would not
pay for extraction, although part of it could be separated with
a further expenditure of time, fuel, and acid. The gold is

pressed lav a hydraulic ram, the pressure exerted being about
800 Ibs. to the square inch. The cakes of metal are dried at a
cherry-red heat, and then broken up for melting.
Treatment of the Silver Solution. The solution of nitrate of
silver is diluted with water, allowed to cool, and then treated
with a strong solution of salt which is regulated so as not to be
in excess, a constant agitation being kept up by revolving wooden
agitators driven by steam power, or by hand paddles. When all
the silver has been precipitated as chloride, the whole is allowed
to settle overnight, and, in the morning, the clear solution of
nitrate of soda, containing most of the base metals originally
present in the alloys, is drawn off and filtered. The precipitated
chloride is washed several times by decantation and agitation,
and finally sweetened in wooden filters by boiling water, which
incidentally dissolves out the chloride of lead. The filters are
usually lined with linen or some similar material.
Reduction of the Silver Chloride. The silver chloride is then
reduced in lead-lined tanks by means of granulated zinc and
water acidulated with sulphuric acid. Thirty-three pounds of
commercial granulated zinc are stated to be enough to reduce 100
Ibs. of silver from the chloride.* The reactions involved are as
follows :
(1) 2AgCl + Zn = ZnCl 2 + 2Ag

(2) Zn + H 2 S0 4 = ZnS0 4 + H2
(3) H 2 + 2AgCl = 2Ag + 2HC1
(4) 2HC1 + Zn = ZnCl 2 + H2
These reactions explain the fact that, while zinc slowly reduces
silver chloride the presence of water only, the action is quick-
i*i
ened by the addition of free acid, by which the zinc is attacked
and hydrogen evolved. Nascent hydrogen is a powerful reducing
agent, and decomposes silver chloride much more rapidly than
zinc does, hydrochloric acid being formed and rendered available
for the production of more hydrogen. The result of this is that
the action, which is at first slow, becomes more and more rapid
as hydrochloric acid accumulates in the solution. The sulphuric
acid is only needed to start the reaction, but, of course, the more
that is added, the more quickly will the operation proceed, and if
much is added, the chemical action is more violent at first than
afterwards, the amount of free acid present in this case falling
off. At the San Francisco Mint 1 Ib. of acid of 60B. is added
for every 2 Ibs. of silver to be reduced. Hydrogen is evolved
copiously and is carried off by a hood and flue. Energetic
stirring with wooden paddles is desirable to prevent the forma-
tion of lumps of chloride, protected by a layer of silver powder.
* Ninth Report Cal. State Min., 1889, p. 70.
PARTING. 375
The white chloride of silver gradually turns black as the silver is

reduced. To test whether the reaction is complete, some of the
silver is taken out, washed well with ammonia, filtered, and the
clear solution acidified with nitric acid. A
white precipitate
signifies that undecomposed silver chloride is still present in the
vat.
When the reduction is complete, the vats are allowed to settle,
the solution drawn off, and a little sulphuric acid added to dis-
solve any residue of zinc that may be present. The dark grey
pulverulent silver is then washed by decantation, after which it
is removed to a wooden filter, and sweetened by washing with
boiling water, and finally pressed, dried, and melted into bars,
which are about 998 fine. The zinc and sulphuric acid used in
this process are lost, and a considerable quantity of undecom-
posed nitric acid is also run to waste, being contained in the
solution from which the silver chloride is precipitated.
The cost of refining and parting by the nitric acid process at
the United States Mints in Philadelphia and New York is some-
what less than 2 cents per oz. of the parting alloy, and in San
Francisco it is nearly 3 cents. The cost for dore silver is
considerably lower. In Europe, the cost is less than in the
United States.
Parting by Sulphuric Acid. This process has now, in the
majority of refineries, superseded the nitric acid method, which
is much more expensive, owing to the higher cost of the acid
used and of the plant required. The German chemist, Kunckel,
who lived in the seventeenth century, is said to have been the
first to employ sulphuric acid in parting, but it was not used on
the large scale until the year 1802, when it was introduced into
France by C. D'Arcet, and worked in a refinery built in Paris
for the purpose. It was established in London at the Mint
Eefinery in 1829 by Mr. Mathison, and has been in almost
continuous use there ever since, with little change, having been
leased to a member of the Rothschild family since 1852.
The method used varies considerably in different refineries,
but essentially consists of the following operations
1. Mixing and granulating the alloys.

2. Dissolving the silver from the granulations by sulphuric acid.
3. Washing and melting the gold residue.
4. Precipitating the silver from its solution by means of copper.
5. Recovering the copper sulphate by crystallisation.
The account given below is a general view of the operations in

various refineries, the modifications adopted not being described
in most cases.
Mixing and Granulating. The alloys must be carefully pre-
pared so as to be of suitable composition, as otherwise
difficulties
are encountered. The most suitable proportion of gold in the
* to be from 18 to 25
alloy is said by Dr. Percy per cent.,
including whatever copper there may be present but some
;
American authorities consider the proportion of one part of gold

to two and a-half parts of silver to be the most desirable, whilst
at a refinery at San Francisco the alloy consists of two parts of
gold to three parts of silver. This proportion was instituted
when alloying silver was scarce in California and has never been
abandoned, but the gold thus separated is only 990 fine, contain-
ing ten parts of silver, the maximum allowed by law in the
gold coins of the United States. If the ordinary proportion of
three parts of silver to one of gold is used, however, the gold can
be obtained about 996 fine, and the fineness of the gold can
always be increased to about 998 or 999 by fusing it, first with
bisulphate of potash and subsequently with nitre. If there are
only a few parts of gold per 1,000 of the alloy, it has been stated
that the silver left in the gold amounts to as much as 3 or 4 per
cent. ; nevertheless, such an alloy, when subjected in the form
of bars to the action of the acid, instead of being granulated,
yields gold at San Francisco of no less than 996 fine, after one
boiling only.
The amount of base metals present in the alloy must be care-
fully regulated, as their sulphates are little soluble in concentrated
sulphuric acid, and consequently are precipitated and interfere
with the progress of the operation. Bars consisting in great
part of copper are often received at the San Francisco works.
These are melted with fine bars so as to reduce the proportion
of copper, which must not be more than about 10 or 12 per cent.
of the whole ; a small amount of copper facilitates the solution
of the silver. A small quantity of lead is said to assist in the
solution of the copper, which is somewhat slowly attacked by
concentrated sulphuric acid, and a maximum amount of 5 per
cent, of lead does not interfere with the operation. From the
economy with which this system of parting can be practised,
silver containing only 0'5 part of gold per 1,000 can be separated
from it at a profit, while the nitric acid process is unremunerative
if applied to an auriferous silver alloy containing one part of
gold in a thousand. At the Vienna Mint, bars are parted con-

taining 0-9 part of gold per 1,000, and at Freiberg bars containing
only 04 part per 1,000 are profitably treated.
In England, silver bars are passed through the parting opera-
tion, if they contain at least 2 grains of gold per troy pound, or
0-35 part per 1,000.
The parting alloy is usually granulated, but at the San Fran-
cisco Hefinery the dore silver is not granulated but melted and
cast into bars J inch thick, 9 inches wide, and 15 inches long.
Solution of the Silver. This is usually effected in cast-iron
kettles, platinum having been abandoned on account of its
*
Metallurgy of Silver and Gold, p. 471.
PARTING. 377
high cost. The iron used is line-grained compact white iron,

preferably containing 3 or 4 per cent, of phosphorus, which
increases the durability, although 2 per cent, only of
phosphorus
is considered enough by some refiners. The kettle is slowly
dissolved by the acid, ferrous sulphate being formed, and, in the
course of about two years, the thickness of the vessel is reduced
from about 2 inches to from J to J inch, when it is discarded.
The perfect exclusion of air from the interior increases the
length of life, and dilute acid must not be allowed to come in.
contact with the iron, as the latter is freely dissolved by it.
The vessels are rectangular or cylindrical, with flat or hemi-
spherical bottoms, the latter being preferred in Europe and the
former in America. They are covered with cast-iron lids, about
^ inch in thickness, which are bolted tightly to the vessels, and
have bent leaden pipes fitted to them for carrying off the fumes,
which consist largely of SO 2 This is sometimes reconverted
.
into sulphuric acid in leaden chambers arranged for the purpose.

The cover has also an opening (supplied with a lid made air-tight
by a water-joint) through which the alloys and acids are added
and the operation watched. The heating is usually done by a
wood fire.
The charge for the pots varies from 200 to 1,000 Ibs. of alloy,
and the amount of acid required varies from 2 to 2J times the
weight of the alloy, depending on the composition of the latter.
About one-half of the acid, which is strong commercial acid of
66 Beaume (specific gravity 1'85) is added at first, and the
temperature cautiously raised to boiling point, when the pot is
closely watched, and, if the ebullition becomes too violent, the
temperature is lowered by regulating the fire and by adding cold
acid a little at a time. The charge is stirred occasionally with
an iron tool, particularly towards the end of the operation, when
the undissolved granules of metal must be freed from the
surrounding sediment, consisting of sulphates of the base metals,
and exposed to the action of the acid. The ebullition gradually
subsides and action ceases in about five or six hours, the
presence of a greater proportion of base metals increasing the
length of time required. The reactions are as follows :
(1) 2H 2 S0 4 + Ag 2
= Ag S0 4
2 + S0 2 + 2H 2
(2) 2H 2 S0 4 + Cu = CuS0 4 + S0 2 + 2H 2
and similar reactions with tin and lead. The re-actions with
antimony, bismuth, zinc, and iron are more complicated. It is
obvious that 63 parts of copper decompose as much sulphuric
acid as 216 parts of silver. It is clear, therefore, that an increase
in the percentage of copper present necessitates an increase in
theamount of sulphuric acid required.
One part of sulphate of silver is soluble in J part of boiling
concentrated sulphuric acid, but the solubility rapidly falls off as
the temperature and concentration diminish, so that 180 parts of

cold acid of specific gravity T08 are required for the same
purpose. Sulphate of copper dissolves slightly in the boiling
concentrated acid, but is almost all precipitated in the form of
the white anhydrous salt on cooling. Tin and zinc behave
similarly, and lead makes the solution turbid and milky. The
iron would not be so much attacked if it were not for the
increasing dilution of the acid during the process, owing to the
formation of water, which is, however, in great part boiled off as
fast as it forms, or taken up by the anhydrous sulphates. The
presence of copper checks the dissolution of the iron.
When the dissolution is complete, the fire is withdrawn and a
few pounds of cold acid of 55 B. are added to the charge, by
which the acid is cooled and diluted, and some crystals of silver
sulphate are formed. These, falling to the bottom, carry down
with them the suspended fine particles of gold, and so clarify the
solution. If much copper is present, however, this is not
necessary, as the slight cooling of the acid, caused by the
withdrawal of the fire, is enough to precipitate some sulphate of
copper, which falls to the bottom and adheres to it very firmly,
thus clarifying the liquid and enabling it to be poured oif or
ladled out very closely. The clear silver solution is then ladled
out with iron ladles into lead-lined rectangular wooden vats
already partly filled with hot water, in which the precipitation
is subsequently effected.
Washing and Melting the Gold Residue. The residue in the
dissolving pot, if the amount of base metals present is not large,
is then boiled twice more with fresh concentrated sulphuric acid
added hot, after which the gold residue is hard and heavy and
rapidly subsides to the bottom, and the liquors are ladled into
the precipitating vat. The gold is dipped out with an iron-
strainer and transferred to a lead-lined filter-box where it is
thoroughly washed, first with hot dilute sulphuric acid and
subsequently with boiling water, after which it is pressed, dried,
and melted. It is almost always brittle, from the occurrence in
it of traces of lead or tin which are difficult to separate by sul-
phuric acid owing to their insolubility. These metals are
eliminated by fusion with nitre or by a blast of air, and the bars
thus toughened.
If the amount of base metals present is very large, the gold
residues are ladled into a vessel of hot dilute sulphuric acid and
boiled with it by means of steam. In this way, most of the
sulphate of silver and the whole of the copper, zinc, iron, &c.,
remaining with the gold are rapidly dissolved. Care must be
taken, however, to add the residues a little at a time, as other-
wise the anhydrous sulphate of copper will form lumps, which
are only slowly dissolved. The gold is then allowed to settle,
and, after the solution has been drawn off, is boiled again with
PARTING. 379
acid if necessary, or if it is
already pure enough it is at once
washed, dried, and melted.
In Europe it is not customary to attempt to obtain pure
gold from auriferous silver in one operation, but the gold is con-
centrated in a small quantity of silver and then mixed with other
alloys rich in gold and parted again. The product of gold ^hus
obtained is purified by heating in a furnace in small iron pots
with about half its weight of bisulphate of potash, by which som
additional silver is converted into sulphate. The temperature is
not raised much above the fusion point of the salt. The fust d
mass is then boiled in sulphuric acid, and again washed, dried,
and melted. In the United States these methods are not used,
auriferous silver being cast into slabs and parted in one operatic 11
by boiling with sulphuric acid ; fusion with bisulphate of potash
israrely resorted to.
Precipitation of the Silver. On pouring the sulphuric acid
solution into water, most of the silver sulphate is precipitated at
once in the form of small crystals, consisting of bisulphate, and
the liquid must then be raised to boiling, by means of steam, in
order to redissolve them. When the original alloys contain
much lead this is not redissolved, and it is, therefore, necessary
to let the solution settle and transfer the clear liquid to another
vessel. Some particles of gold are usually found in the pre-
cipitate thus formed.
The reduction and precipitation of the silver is effected by
means of copper, which takes its place in solution. The copper is
usually added in the form of scrap while the liquid is being heated
up by steam. The precipitation is assisted by constant stirring
by means of wooden paddles. In San Francisco, however, the
copper is cast into slabs, which are suspended side by side in the
solution in a vertical position. The solution should be of about
24 B. if it is much more concentrated than this, the precipita-
;
tion of the silver is imperfect. The end of the reaction is detected

by testing with salt solution, and when complete, the stirring
is stopped, the solution allowed to settle for two hours, and the
clear liquid tapped into lead-lined vessels, where further settling
of the suspended particles of silver takes place. The precipitate
of silver is thoroughly washed with boiling water in wooden
filters lined with lead, until the reaction for copper can no longer
be obtained with ammonia. Care must be taken that no frag-
ments of metallic copper remain with the silver. The metal is
then pressed, dried, and melted, and is usually from 998 to 999
fine, even without fusion with nitre, when the copper plates
are
used for reduction. At the London refineries, the silver pro-
duced is only about 996 fine.
Crystallisation of the Sulphate of Copper.
This is effected by
alternate evaporation and crystallisation in lead-lined wooden
tanks. The solution, which is still of 24 B., is run from the pre-
cipitating tank into the evaporating pan and concentrated to

40 B. by heating with steam; thence it istransferred to the
crystallising tanks, where it is allowed to cool and remain for
from ten to twelve days. The mother liquor of 36 B. is then
run off and reconcentrated to 45 B., after which it is again
allowed to crystallise, reconcentrated to 55 B., and a third crop
of crystals obtained, which contain much iron. The clear acid
mother liquor can now be used to dilute the solution of sulphate
of silver in the dissolving pot as already described. The excess
of acid in surplus liquids is neutralised with oxide of copper,
more copper sulphate being thus formed.
The crystals of bluestone are found adhering to the sides and
bottom of the tanks. They are detached with copper chisels,
redissolved in pure water and recrystallised, the mother
liquors being eventually added to the first liquor from the pre-
cipitating vats. When the liquors become over-charged with
iron, the copper in them is precipitated by means of metallic
iron, and they are thrown away or evaporated to get the crystals
of sulphate of iron. The bluestone crystals are packed in barrels
for the market. One pound of metallic copper with 1/5 Ibs. of
sulphuric acid of 66 B. will make 4'5 Ibs. of crystallised sulphate
of copper.
The cost of the process of parting by sulphuric acid in Europe
is about one farthing per ounce troy of the parting alloy.
Combined Process. At the Philadelphia Mint a combined
process is used, nitric acid and sulphuric acid being employed
in succession. The alloys are granulated and digested with
nitric acid of 39 B. for four or five hours in the same manner as
has already been described ; the solution is then siphoned off,
and the gold washed two or three times with distilled water, by
decantation, and subsequently sweetened in lead-lined filters with
boiling water. The gold is then introduced into cast-iron cylin-
drical kettles and boiled for five hours with sulphuric acid of
60 B., the gold being stirred up with an iron rod every ten or
fifteen minutes to prevent agglomeration, and the solution is then
ladled out and treated as already described, p. 379.
The gold is again boiled in the same kettles with concentrated
sulphuric acid for two hours, after which it is washed thoroughly
and sweetened in wooden filters, boiling distilled water being
poured through it until the washings will no longer redden blue
litmus paper. The silver is precipitated from these washings as
chloride by the addition of salt. The gold is then pressed,
dried, melted, and cast into bars, which are from 996 to 998 fine.
By a third boiling in sulphuric acid it is said that they can be
raised to 999-5 fine, but according to English refiners this can
only be done by fusion with potassic bisulphate.
This process is much cheaper than the nitric acid process,
costing 20 per cent, less for acids, and saving some fuel. The
PARTING. 381
granulations contain 100 parts of gold in 285 of tlie alloy. After

the boiling in nitric acid only 6 per cent, of silver is left with
the gold. The cost of refining is a little over one cent, per oz.
The Gutzkow Process. This process of parting by sulphuric
acid was invented and patented by Mr. F. Gutzkow in 1867. It
has been extensively worked in Germany and in San Francisco,
and up to the year 1891 had been instrumental, on the authority
of Mr. Gutzkow, in refining one hundred million dollars' worth of
silver. It is fully described in Percy's Metallurgy of Silver
and Gold, p. 479, and only a brief account will be given here.
When the patent had expired, Mr. Gutzkow introduced and
patented several improvements on it, which will be described at
greater length. This improved process is now successfully at
work at the Consolidated Kansas City Smelting and Refining
Company's Works at Argentine, Kansas, where it was estab-
lished in the spring of the year 1892.
The original Gutzkow process, as employed at the San Fran-
cisco Assaying and Refining Works for many years, may be
summarised as follows The bullion treated is of three kinds,
:
viz., (1) Gold bars from retorted metal, containing about 900
parts of gold, 10 to 20 of base metals, and the remainder silver ;
(2) Comstock silver bars or dore bars, usually containing 20 to
100 parts of gold per 1,000 ; (3) base bars from the Reese River
districtand from pan-amalgamation of tailings, containing from
100 to 800 parts of silver, and the remainder chiefly copper, with
sometimes a little gold. The gold bars (1) are alloyed with
silver and granulated, but the others are cast into bars, and
parted in that form. The dore bars, when prepared for solution
in the acid, weigh about 100 Ibs. each, and are 12 inches long,
6 inches broad, and 5 inches thick. The base ingots are melted
with fine bars to reduce the average copper contents to 12 per
cent., and are cast into bars 1 inch thick, the gold from which is
only about 992 fine.
The boiling is done in flat-bottomed thin cast-iron kettles
(A, Fig. 58), of which the bottom is only f inch thick
when
new, and J inch when worn out. The solution can be rapidly
heated, owing to the thinness of the iron kettles, and 200 Ibs. of
alloy are dissolved in four hours by means of 300
Ibs. of
sulphuric acid, which comes from the tank C, and is forced

into
the kettle through the pipe / by the plunger d. The solution
is then siphoned off through the pipe into the tank E,
-///
and diluted with a large quantity of hot mother liquor from a

previous crystallisation, which is mainly sulphuric acid of
about
58 B. some water is also added, and the solution partially
;
to
cooled, so that some crystals of silver sulphate are enabled
of
separate out and carry down with them the milky precipitate
lead sulphate and any suspended particles of gold ; green basic
is then
sulphate of iron also settles firmly. The clear solution
siphoned off intoH and cooled to 80 R, and almost all the silver
sulphate thus crystallised out. If the acid is concentrated, white
PARTING. 383
soft crystals of bisulphate are formed, which is not desired if, ;
however, the acid is only at about 58 B., large hard yellow-

crystals of monosulphate, free from acid but contaminated with
copper, are deposited. The mother liquor is pumped back into
the tank E or to the original acid tank, the device
employed
for this purpose being to exhaust them of air, so that the acid is
sucked up without passing through any valves, which would
soon wear out. The crystals of sulphate of silver are transferred
to the filtering box I by iron shovels, and a hot solution of green
vitriol of 25 B. run on to them from G. This is at first mainly
occupied in dissolving the sulphate of copper, and the first portion
of the solution, after passing through the filtering box, is run into
a storing vat, where the silver, incidentally dissolved, is precipi-
tated by copper, and the latter subsequently recovered by means
of iron. After a time, the copper being dissolved, the silver
begins to be reduced, the green solution of iron turning coffee-
brown ; the reaction is as follows :
2FeS0 4 + Ag 2 S0 4 Fe 2 3 . 3S0 3 + 2Ag

The reduction may be effected by sheets of metallic iron,
also
which is first converted into ferrous sulphate and then into ferric
sulphate, the silver being simultaneously reduced to the metallic
state. The brown solution of ferric sulphate is boiled with
metallic iron in K, in order to regenerate the ferrous salt. The
silver is washed, pressed, dried, and melted. The gold from the
original dissolving kettle is also washed in a filter, pressed, dried,
and melted.
Such was the original Gutzkow process as employed in
,
treating dore bars. Its chief advantage over the ordinary

sulphuric acid process was the saving of acid. In the ordinary
process, none of the acid used is saved, so that it is reduced in
amount as much as possible, but does not fall below twice the
weight of the silver dissolved. This reduction in the amount of
acid used makes the finishing of the dissolving a difficult and
delicate operation. In the Gutzkow process, however, only the
acid decomposed by the silver is lost; the weight of this is about
equal to that of the metal, the rest of the acid being all recovered
and used over again in the boiling. Moreover, the long and
tedious crystallisation of copper sulphate is avoided, and the
space required for the crystallising vats saved. However,
several large lead-lined vessels are required for the storage of
the various solutions, and the expense of these, as well as
the space required, is greatly reduced by the recent improve-
ments described below.
The New Gutzkow Process.* Mr. Gutzkow has lately
*This is fully described by Mr. Gutzkow in the Enn. and Mng. Journ.,
Feb. 28, 1891, p. 257, and May 7, 1892, p. 497, from which this account is
summarised.
pointed out that if a large amount of acid is used for the boiling,
not only is the silver more completely dissolved and the operation
greatly expedited, but the presence of a high percentage of copper
does not hinder the parting, as it is kept in solution by the excess
of free acid. Thus, for ordinary dore silver, he uses four parts of
acid to one of bullion for bars containing 20 per cent, of copper
;
he uses six parts of acid ; for still baser bullion, more acid, and
so on, never losing more than one part of acid for one of bullion,
and recovering the remainder.
The charge for a pot 4 feet in diameter and 3 feet in depth
is 400 Ibs. of dore silver : the pot is flat-bottomed, with a
basin-shaped pocket or well in the centre which is useful for
the collection of the gold. The bullion is first attacked by
fresh acid of 66 B., run in by gravity from a large tank, and,
when most of the silver has been dissolved, mother liquor from
a former operation is added, a pitcher-full at a time, until the
charge is completely dissolved, which takes from four to six
hours. The lire is then moderated, and the pot tilled with
mother liquor to within 1 or 2 inches of the top, when the
temperature of the acid will have been so far reduced that only
faint fumes are discernible. If no fames are visible the acid is
too cold and some silver sulphate will be precipitated, but other-
wise the large excess of acid will keep it in solution. The
well-stirred charge is now allowed to settle, which is perfectly
accomplished in ten minutes, as the yellowish slowly-subsiding
persulphate of iron is transformed to a greenish flocculent com-
pound by the water in the mother liquor, and this settles quickly
and carries all suspended matter to the bottom. More iron is
dissolved from the kettle than in the ordinary process, owing to
the greater dilution of the acid used in boiling.
The solution is now siphoned from the kettle by means of a
j-inch gas pipe into a large cast-iron vessel, only about 1 foot
deep, standing in a larger vessel which can be filled with water
for cooling the charge. Steam is blown into the still hot acid
solution through a lead nozzle, J inch in diameter, pointing
vertically downwards. This both dilutes and warms the solution,
the heating being necessary in order to prevent crystallisation
of the silver consequent on the dilution. As soon as the dilu-
tion has proceeded sufficiently far to ensure the crystallisation
of the hard yellow monosulphate instead of the soft white bisul-
phate of silver, a point which is found by dipping out small
quantities at intervals, and observing their behaviour on cooling,
the steam is shut off and the vafc cooled with water and left all
night. The silver crystals form a coating of about 1 inch thick,
which is contaminated with copper sulphate if the mother liquor,
by repeated use, has become saturated with it. The mother
liquor is now pumped back into the acid storage tank by the
creation of a vacuum, and the crystals of sulphate of silver are
PARTING. 38.5
detached with an iron shovel and thrown into a filtering-box

provided with a false bottom. Cold distilled water is sprinkled
on the charge, and is allowed to filter through it and flow back
into the crystallising vat, until the greater part of the free
acid has been removed. The stream is then deflected into a
'silver filter" where any silver is precipitated that may have been
dissolved at the same time as the sulphates of iron and copper.
The silver filter is a lead-lined box, partly filled with precipitated
copper and provided with a false bottom. The silver separates
on the top of the copper as a spongy sheet, a corresponding
amount of copper being dissolved. When the crystals of silver
sulphate in the first-named filtering box have been completely
freed from acid, and from copper and iron sulphates, the stream
of water is discontinued. The spongy sheet of silver is then
"
removed from the " silver filter box and treated with hot
water and a few crystals of silver sulphate to dissolve the copper
still retained by the sheet. During this whole operation of
sweetening the crystals of silver sulphate, only about 3 per cent,
of it is dissolved, as it is little soluble in cold water. The copper
" silver
solution, after passing through the filter," is either run
to waste or precipitated by scrap iron in wooden tanks at a
nearly boiling temperature.
The crystals are now dried in an iron pan which is placed
above a furnace, and, after being mixed with about 5 per cent,
of charcoal, they are at once charged into a hot crucible in a
melting furnace. The silver sulphate is reduced at a low red
heat to metallic silver, carbonic and sulphurous acid gases being
evolved. By the time the temperature of melting silver is
reached, these gases will have all passed away. The silver is
toughened by adding nitre and borax until the so-called
"boiling" indicates that the sulphur has all been eliminated,
and the metal is then cast into bars.
The gold residue in the dissolving kettle contains insoluble
sulphates of lead, iron, antimony, mercury, and often some
copper and silver. It is ladled out and boiled with water to
dissolve out the sulphates of silver, copper, iron, &c., and, after
thorough filtering, it is stirred in a dish with hot water, and
decanted on to a filter-cloth until the insoluble sulphates of lead,
&c., have all been washed off, and the gold is left bright and
clean. The gold is stored until enough is collected to make a
200-oz. bar, which is usually brittle. The material collected on
the filter-cloth is re-washed once or twice to recover the particles of
gold from it, and can then be reduced with charcoal and cupelled.
If lead is present in the original alloy, part remains with the gold,
and is dealt with in the manner which has been already described,
but the greater part is carried off with the silver solution, and is
deposited both while the steam is being passed in, and
also-
subsequently during crystallisation of the sulphate of silver,

which is coated with it. The sulphate of lead is removed from

the crystals by stirring them well in a stream of cold water,
by which the light insoluble particles of lead and antimony
sulphate are carried away ; it can then be collected, reduced, and
cupelled. Any silver that may be dissolved in the course of this
washing is precipitated by copper as before.
The process is seen to differ from the original one in three
essential particulars : 1. The solution is diluted with steam
instead of with mother liquor, the amount of liquid in use, and
consequently the number of lead-lined vats required being thus
reduced. This is an important item, especially in the United
States where lead-burning is expensive, owing to the existence
of a powerful union. 2. The weak silver solution is precipitated
at once, instead of being stored in tanks to be used again or to be
precipitated at leisure. 3. The silver sulphate is reduced directly
with charcoal in a crucible in the furnace. This saves the
pressing of the silver and, what is of greater importance, avoids
the use of the solution of sulphate of iron, which needs to be
stored. The reduction in the crucible and subsequent melting
requires scarcely more fuel than would be used to melt the
pressed silver. One of the minor advantages of the process is
said to be that no stirring is required during the boiling, owing
to the large amount of acid used. This saves labour and enables
the acid fumes to be more easily condensed, as they are not
mixed with air, which in the ordinary way would enter through
the aperture left for stirring. The exclusion of air also helps to
prolong the life of the iron kettles by checking the attack on them
by sulphuric acid. Mr. Gutzkow also declares that, owing to the
excess of acid present, it is not necessary to specially prepare the
alloys for dissolution. Bars, retorted metal of any shape, scrap,
&c., may be added just as they are, provided that the amount of
gold in them is so small that they can be fairly called dore bars.
Finally, it is stated that all the silver and gold charged into the
kettles in the morning can be melted into bars and made ready
for assay before night. The cost of this process was stated by
the general manager of the Kansas City Works in April, 1892,
to be 0*35 cent per oz. of dore. The wages at these works are
from $3 to f>4 per day, and sulphuric acid costs ]i cents per Ib.
The refining charges in the Eastern States average 1 cent per oz.
of metal, and in California about 2 cents per oz. It is evident
that these charges can be greatly reduced by the new process.
Miller's Chlorine Process. The use of chlorine gas for the
purification of molten gold was first proposed by Mr. L. Thompson
in 1838, and the results of his investigations were published in
the Journal of the Society of Arts* two years later. He stated
that " it has long been known to chemists, that not only has
gold no afiinity to chlorine at red heat, but it actually parts with
*Vol. liii., parti., p. 17.
PARTING. 387
it at that temperature, although previously combined. .. .
This, however, is not the case with those metals with which it is
usually alloyed. It offers, therefore, at once an easy and certain

means of separation."
In 1867, Mr. F. B. Miller, Assayer of the Sydney Mint,
applied this property of chlorine to the separation of gold from
silver on the large scale, and his process has been in use at
Sydney ever since, being particularly suitable for the purpose
under the local conditions. Among these conditions may be
mentioned the facts that acid is very costly, and that there is a
scarcity of silver bullion containing small quantities of gold,
while the gold produced in Australia contains but little silver.
The result is that the ordinary parting processes would prove
very expensive, but the chlorine process can be applied cheaply,
as it requires very little acid, and is efficacious in removing
small quantities of silver from gold bullion which has not been
made up into alloys of definite composition. Before the intro-
duction of the chlorine process no attempt was made to extract
the silver from any of the native gold of Australia and New
Zealand which was coined at the Sydney Mint. Sovereigns
were manufactured containing several per cent, of silver, which
replaced part of the copper used as the alloying metal. These
sovereigns, some of which are still in existence, can be easily
recognised by their pale tint, due to the presence of silver. Such
sovereigns have not been manufactured since 1867. Besides
separating the silver, the chlorine process removes the small
quantities of lead, antimony, &c., which render most of the
Australian retorted gold brittle, and so in one operation prepares
the gold for coinage. Practically the whole of the gold produced
in Australia is now deposited in the mints of Sydney and Mel-
bourne, and refined by this process, the amount treated in 1892
having been 1,673,000 ozs.
The following is abridged from that given by the
description
lateMr. Miller,* and from later writings The furnace used is
:
an ordinary melting furnace, such as has been already described.

The tile cover of the furnace has a hole in the centre to allow
the chlorine tubes to pass through. French clay crucibles are
used, Nos. 17 and 18 being convenient sizes, holding about 600
or 700 ozs. of gold ; they are placed inside graphite pots to
prevent loss by cracking. They are glazed inside by melting
borax in them to prevent them from absorbing molten chloride
of silver. Graphite crucibles are said to be unsuitable, as silver
chloride appears to be reduced, presumably by the hydrogen
contained in them, as fast as it is formed. The crucibles are
covered by loosely fitting lids, through which the clay pipe-stems
of about ^--inch bore are passed to the bottom of the crucible for
* New South Wale*,
Journ. Chem. Soc., 1868; and Trans. Roy. Soc. of
1869.
the conveyance of chlorine. The pipe-stem is made red-hot

before being introduced into the molten metal, as otherwise it
would crack and break off. The chlorine generator consists of a
stoneware jar furnished with three necks, and capable of hold-
ing from 10 to 15 gallons of liquid. The three openings are
fitted with well-secured rubber plugs, through two of which
two tubes are passed, viz., the safety tube, which is 8 or 10 feet
high, with its open end bent over so as to deliver into a large
jar, and the eduction tube, which is closed by a stopcock till it
is required. The generator is partly filled with from 70 to 100
Ibs. of manganese dioxide in small lumps, an amount which will
suffice for many operations ; hydrochloric acid is introduced
through the safety tube when the gas is required. The
generator is warmed by a steam jacket.
The chlorine gas is conveyed in leaden pipes to the furnaces.
All joints and connections are made by well-wired india-rubber
tubes, which must be protected from direct radiation from the
furnace. Screw compression clamps on these rubber tubes enable
the supply of chlorine to be regulated to a nicety. When the
clamps are closed the gas accumulates and forces the acid up the
safety tube into the vessel placed overhead, and so the further
generation of gas is prevented. Two such generators and three
melting furnaces are enough to refine 2,000 ozs. of gold, containing
10 per cent, of silver.
The generators being in readiness, the crucibles are slowly
heated to redness, and the full charge of 600 or 700 ozs. of bullion
introduced and melted, 2 or 3 ozs. of borax being sprinkled on
its surface or poured on in a molten state. The chlorine is now-
allowed to pass slowly through the clay pipe to prevent metal
from entering it, and the pipe is plunged to the bottom of tho
molten metal and kept there by means of a weight attached to it.
The full stream of chlorine is now turned on and is heard to be
bubbling into the molten metal, by which it is completely
absorbed, so that no splashing and projection of the metal occurs.
A height of 16 to 18 inches in the safety tube corresponds
to and balances a height of 1 inch of gold in the refining crucible.
The safety tube acts as an index of the pressure in the generator
and of the rate of production of the gas any leakage or the
:
exhaustion of the acid is at once indicated by a fall of the liquid

in the tube. Fresh acid is added at intervals as it is required.
When the chlorine is introduced, dense fumes at once arise
from the surface of the metal owing to the formation of volatile
chlorides of the base metals, which are the first to be attacked :
lead gives especially dense fumes, which can be condensed on a
cold object held in them. After a time these fumes cease and
silver chloride is formed, very little chlorine escaping from the
crucible, even if an extremely rapid current is passed into it ;
consequently the operation is expedited by every increase in the
PARTING. 389
volume of the current. Towards the end of the operation

splashing is more noticeable, and dark brownish-yellow fumes
appear, consisting chiefly of free chlorine. The completion of
the refining, however, is indicated by a peculiar reddish or
brownish-yellow stain which is imparted to a piece of white
tobacco-pipe when exposed to the action of the fumes for a
moment. It is suggested by Prof. S. B. Christy that the stain
contains gold. This stain appears in about one hour and a-half
from the start, when 600 ozs. of gold, containing 10 per cent, of
Bilver, are being subjected to treatment. The current of gas is
then at once stopped, and the crucible lifted out of the furnace
and allowed to cool sufficiently for the gold to solidify. Pro-
bably, if the operation were continued after the appearance of
the brown stain, losses of gold by volatilisation would occur.
The chloride of silver, still molten, and floating on the top of
the gold, is then poured off into iron moulds, and the crucible
inverted on an iron table, when the red-hot cone of gold falls out.
This is now fine, and after any adherent chloride of silver has
been detached from it by scraping, it simply requires melting
into ingots, 98 per cent, of the gold being thus at once rendered
.available for use. The remainder of the gold is contained in the
chloride of silver, partly in the form of entangled shots of metal,
but chiefly as a double chloride of silver and gold. It was for-
merly recovered by melting the chloride with about 10 per cent,
of metallic silver, rolled to about J inch in thickness. The gold
is reduced by the silver and alloys with the excess, settling to the
bottom of the pot where it solidities after ten minutes cooling,
so that the chloride of silver can be poured off into large iron
moulds, slabs suitable for reduction being thus formed.
It was found at the Sydney Mint,* that the above method of
separating the gold from the silver chloride
was subject to several
disadvantages. In particular, although on a small scale the
amount of gold in the silver could be reduced to from 0-3 to I'O
with silver foil for
per 1,000 by careful and continuous stirring
-a great length of time, nevertheless in practice on a large scale
the results varied greatly, and the silver bullion produced usually
contained from 10 to 25 parts of gold per 1,000. Several re-
such as resin, hydrogen and coal gas were
ducing agents, argol,
to give good results.
successively tried but were not found
Finally, the application of soda carbonate,
which had been pro-
Leibius in 1868, was adopted, the method of procedure
posed by
being originally as follows f"The argentic chloride is covered
:
and when all is

by a layer of fused borax, about J inch thick,
well .fused, the powdered soda is sprinkled on the top of the
borax, without stirring, as rapidly as
the ensuing action will
* Fourth Annual Report of the Royal Mint, 1873. Report by A. Leibius,
p. 63.
t Loc. cit.
admit. Occasionally the top layer is dipped with a stirrer

slightly underneath the molten argentic chloride, without stirring
the latter. When all the necessary soda is added and the action
is nearly over, the pot is covered with a lid, and left for about
ten to twenty minutes to increased heat, and, when the contents
are quite liquid, the pot is lifted out of the fire without previous
stirring, and allowed to cool, so as to enable the argentic chloride
to be poured off from the gold button at the bottom of the pot.
"Although in several experiments all but O'l of gold per 1,000
was eliminated from the silver bullion produced, in no case is
every trace of gold removed in one operation. To free the
argentic chloride entirely from gold, producing therefore silver
bullion free from gold, was, however, accomplished by subjecting
the argentic chloride to a second treatment, with a small quantity
of soda, in a separate boraxed clay pot, similar to the first opera-
tion.
" A convenient quantity of argentic chloride, to be treated
in a No. 18 French clay pot, was found to be 230 ozs. The
amount of soda required for 230 ozs. of chloride may range
from 16 to 20 ozs. Less than 16 ozs. leaves too much gold in
the silver, while more than 20 ozs. produces a very silvery
gold button, and yet without completely freeing the argentia
chloride from gold.
"The use of 18 ozs. of soda for 230 ozs. of chloride pro-
duces a gold button weighing between 30 and 35 ozs., assaying
about 920 to 930, and leaves from 0-5 to 1-0 part of gold in 1,000
parts of silver bullion produced.
"With 20 ozs. of soda the results were :
Gold, about 35 ozs.,
assay 870-880 ; gold left in the silver bullion produced from 0'2.
to 0-5 per 1,000.
"With 16 ozs. of soda: Gold from 30 to 33 ozs., assay
940-950; gold left in the silver bullion, from 1-0 to 2-0 per
1,000, and sometimes as much as 6'0 per 1,000.
" To free the
argentic chloride from gold, a second treatment
with 3 ozs. of soda per pot of 200 ozs. chloride, containing but
a minute quantity of gold, will always be lound to answer, thfr
only care required being gradual application of the soda and
enough heat at the end of the operation."
The time required for the two operations is about half an
hour.
" The
presence of a large proportion of chloride of copper has
been found to prolong the operation considerably on account of
oxide of copper being formed on addition of soda, as a much
greater heat is required in order to
fuse the whole mass. The
argentic chloride produced from base gold alloys would contain
a large proportion of chloride of copper, &c., and it would be
better, therefore, to reduce it direct, and dissolve the
reduced
metals in acid, to separate gold and silver therefrom."
PARTING. 391
The may be assayed for gold by cupellation

silver chloride
with lead and subsequent parting.
foil
The method just described was adopted at the Sydney Mint
in 1872, and at the Melbourne Mint in the following year, with
excellent results.
The process of reduction of the silver chloride was devised by
the late Mr. A. Leibius, fellow-assayer of Mr. Miller at the Sydney
Mint, and was described in a paper communicated to the Royal
Society of New South Wales.* In this process, 1,400 ozs. of
argentic chloride are reduced in 24 hours by the apparatus, of
which the following is a brief description Seven zinc plates,
:
each 14 inches long, 12 inches wide and J inch thick, are sup-
ported about 1J inches apart in a vertical position in slots
in a wooden frame. Six slabs of argentic chloride, each 12
inches long, 10 inches wide and J inch thick, are suspended by
loops made of silver bands, in such a way that each slab is placed
between two of the zinc plates and separated from them by
spaces of about J inch. The silver loops are connected with silver
bands on which the zinc plates rest, so that there is metallic
connection between the slabs of chloride and the zinc plates.
The whole is now plunged into water, to which some of the liquor
from a previous operation containing chloride of zinc in solution
is added as an exciting agent. Galvanic action soon begins, the
liquor gets gradually warmer and a strong current is discernible.
The silver chloride is gradually reduced to metallic silver, the
slabs undergoing no alteration of form, and the zinc is dissolved.
The slabs of silver chloride are generally free from most of the
base metals, but copper, if present in the original alloy, is not
volatilised in the crucible, and its chloride remains mixed with
that of the silver. The two metals are now reduced together.
"When all action has ceased, the slabs of cupreous silver are lifted
out and boiled, first in acidulated water and then in pure water,
while still suspended in their silver loops. The porous metal is
now ready for melting. As no acid is used the amount of zinc
consumed is the theoretical quantity required by the equations -
2AgCl + Zn = ZnCl 2 + 2Ag

CuCla + Zn = ZnCl 2 + Cu
The weight of zinc consumed usually amounts to from 24 to 25

of fused chloride. The zinc
per cent, of the weight of the slabs
are used over again until worn too thin for safety, after
plates
which they are melted-up and cast into new plates. They suffer
no loss if the apparatus is left untouched for any length of
time after the whole of the silver has been reduced.
At the Melbourne Mint in the year 1889, the zinc plates
* Trans.
Roy. Soc. New South Wales, 1869.
The paper is given almost
at full length iu Percy's Metallurgy of Silver and Gold, p. 418.
employed as described above were replaced by sheets of iron

"
with satisfactory results. Upon the reducing bath being
heated with steam, the chloride of copper dissolving, dis-
engages itself freely from the slabs of chloride of silver, and
coming into contact with the iron is reduced, and the metallic
copper falls to the bottom of the bath in large quantities, leav-
ing the reduced silver in a much cleaner state than when zinc
was used. The noxious fumes which were formerly given off on
*
the melting of the reduced silver sponge are also avoided."
Fig. 58a.
The Chlorine Process as now practised at Melbourne. The

Mr. Francis
following description has been kindly supplied by
R. Power, the Assayer at the Royal Mint, Melbourne, by per-
mission ot the Deputy Master. It gives the exact methods and
apparatus in use in the early part
of 1896. As will be seen,
these differ considerably irom those described above, and from
the practice at Sydney.
* Twentieth Annual Report, Royal Mint, 1889, p. 126.

PARTING. 393
Furnaces. These are thirteen in number. They are built

cylindrical^ (see Fig. 58a, in which one of these furnaces is
shown in section, with crucible and pipe-stem in position), being
more compact in this form, more easily cleaned from clinker,
.and more economical in fuel than the square ones. They are
12 inches in diameter and 21 inches deep. The five firebars,
1|-
inches square and 18 inches long, are set in a cast-iron box,
D, 12 inches by 2 inches, which passes through the brickwork
in front of the furnace, the other ends of the firebars resting on
an iron bar set in the brickwork at the back of the furnace.
The bars are 6 inches above the floor. The draught is obtained
through a grating in the floor, which covers a portion of the ash-
pit, over which there slides
a cast-iron plate, M, inch thick, for
regulating the admission of air, and pivotted in one corner. The
flue, L, is 6 inches square, and communicates with a series of
five condensing chambers, 8 feet by 8 feet by 5 feet, running the
length of the furnaces (42 feet), all communicating and leading
to the stack, 80 feet high, common to refining and melting fur-
naces, which are twenty-one in all. There are three furnace
covers, two of them 20 inches by 6J inches, the third a little
smaller, and all are bound with iron. The middle one is per-
forated by a 1-inch hole, through which the chlorine delivery
pipe passes. Glenboig arched firebricks, B, 9 inches by 4J inches,
and tapering from 2f inches to 2 inches, are used for lining the
furnaces, and are set with touching joints in an iron cylinder, A,
21^ inches in diameter, and at least ^ to |^ of an inch thick,
which is supported by a cast-iron plate, C, J- of an inch thick,
and 22 inches in diameter, with a 12-inch hole in the centre.
This plate is supported by the brickwork which forms the
foundation. The ashpit is a cast-iron flanged box, easily
cleaned in case of an accident. Round the iron cylinder
concrete, N, is rammed, the front iron plate of the furnace
being shifted 2 or 3 inches in, until this is set sind then moved
out, thus providing an air space, E, and keeping the plates
cooler. The furnace top is a plate of cast iron and, so as to
facilitate repairs, should be in two pieces for each furnace, halved
into one another, the hole being slightly bossed at the edge so
that the tiretiles may run easily on them. One piece has a hole
6 inches in diameter over which the swing ventilating hood, P,
is placed by which the pot is covered when removed from the
fire. This hood communicates by a passage through the brick-
work with the flue. The cylindrical furnace is calculated to
last for three years, the square ones lasting only eighteen months
and taking three hours to reline, while the cylindrical ones
take one hour.
The CruciWes, <&c. The guard pot, placed lor safety under the
white pot and afterwards used for remelting the refined gold,
is a plumbago crucible 8 inches high, 6 inches inside diameter,
J-
of an inch thick at the top, and |- of an inch at the bottom,
which is flat inside and stands on a cylindrical firebrick 5 inches
in diameter and 2 inches deep. The white pots, fitting loosely
into the guards, are 10J inches high, 5 inches in diameter, and
f of an inch thick at the top, tapering from inch at the bottom.
I
They are covered by a closely-fitting lid, dished at the top to

catch any globules spirted out by too rapid a current of gas
and perforated by two holes ^ of an inch in diameter. A new
pattern of lid to be introduced shortly will have a notch in
the edge for the pipe-stem to pass through, the advantage of
this being the easy removal of the lid without withdrawing
the pipe-stem, as is necessary with the old lids.
Fig. 586.
The pipe-stem is 24 inches long, tapering from f to J an inch

at the end inserted into the gold, and is wedge-shaped to facilitate
the escape of the chlorine when resting on the bottom of the
pot. The bore of the pipe-stem is J- of an inch in diameter.
The thin end of the pipe-stem is attached to the branch delivery
-inch rubber about 2J inches long, which
pipe by a piece of
in length, with
connects with an ebonite junction, G, 3 inches
a bore of ^ of an inch, turned with a ring round the middle,
which acts as a rest for the 8-oz. weight, H, used as a sinker
for the pipe-steal. One end of the ebonite junction is J an inch
in diameter, the other j of an inch ; the latter being connected
PARTING. 395
by a stout rubber tube 3 or 4 inches long to a 14-inch lead pipe

Q an inch in diameter) which is attached
[by a rubber junction!
to a glass stopcock, I, from the
spigot of which a j-lb. lead
weight, J, is suspended to prevent the pressure of gas from
blowing it out. The glass stopcocks have replaced the com-
pressor clamps, which were not satisfactory owing to the rubber
cutting through, and chlorine leaking past.
^
The rubber joints
are sufficiently flexible to allow the
pipe-stem to bend down
into the pot or to be laid horizontally on a rest when not in use.
Each furnace is provided with one glass stopcock to control
the flow of gas. The cock is far enough away at the back of
the furnace, to be unafl'ected by the heat when the firetiles
Fig. 5Sc.
are removed, and is connected by a J-inch lead pipe with the

main pipe running along the wall. Thinner pipe-stems are
found to be as serviceable as the above and do not require
such careful annealing. The tubes, stopcock, &c., are some-
what diagrammatic in Fig. 58a, and can be studied in Fig. 586,
which is from a photograph of the furnaces by Mr. R. Law.
This also shows the crucibles, guard pots, ventilating hoods, and
chloride cakes.
The generators are shown in Fig. 58c, which is also from a
photograph by Mr. R. Law. The pressure regulator and the
reduction tank are also shown. The generators, eight in number,
.are three -necked cylindrical stoneware vessels with domed

tops, and having a flange round the middle by which they are
supported on the stoneware steam jacket 16 inches high, 16J
inches in diameter, and | of an inch thick. The domed vessel
is 2 feet high. The three necks have If-, 1|-, and If -inch holes,
the first lor charging-in the manganese ore and closed by an
indiarubber plug, the second lor the pipe leading to the main
chlorine pipe, and the third for a branch acid supply tube, an
inch in diameter, fitted with a glass stopcock a foot above the
neck, between which and the stopcock another J-inch tube,
the overflow, branches. The overflow tube, through which the
hot waste of the generators has to pass, is provided with an
.
ebonite stopcock which is turned off daring refining. Stout

combustion tubing is used. The bottom ot the generator is
covered by 4 inches of quartz pebbles, to prevent choking of
the acid delivery pipe, which reaches to within 1 inch of the
bottom of the vessel. 56 Ibs. of manganese ore (about 73 per
cent, peroxide), broken to ^ to J inch square, is placed on the
pebbles, and commercial hydrochloric acid, of specific gravity
1-16, is added as required through the acid delivery pipe by
turning the glass stopcock. The acid pipe is of glass, and leads
to the eight storage tanks 20 feet above the floor, which hold
320 Ibs. of acid each, and are interconnected by glass tubes luted
into the bottoms the delivery of acid, however, being from one
;
at a time. The gas delivery pipes from the generators all connect
with a 1 inch lead pipe, which leads to a distributing vessel
with two necks and partially filled with manganese chloride
solution. A pressure guage of 1 inch glass tube and 15 feet
high is luted into the bottom of this vessel, and is fixed to the
wall by brackets, 10 to 11 feet of the solution being required to
overcome the resistance of about 7 inches of metal in the pots.
The pressure in refining is equal to 5 Ibs. per square inch.
A four-way tube of lead or pottery is passed through the second
neck of the vessel, and each arm is connected by thick rubber
to glass stopcocks to which J-inch lead pipes are joined, these
pipes leading to sets of four, four and five furnaces, so that the
supply of gas can be delivered to a few or all the furnaces, as
desired, the subdivision being made for safety in case of a
leakage or for convenience if only a few furnaces are in use.
All the generators are used whether the quantity of gold to
be refined be large or small, the same quantity of acid being
run into each. \V hen the flow of chlorine through the gold is
stopped the acid in the generators is forced back through the
overflow pipe by opening the ebonite tap. It is found necessary
to have the main pipe in communication with another two-
necked earthenware vessel containing such a quantity of water
that when the pressure of gas exceeds the working pressure
required, the end of a glass tube, passing to the bottom of
PARTING. 397
the vessel and connected above the neck with an upright 4-inch
lead pipe 10 feet high, becomes unsealed, and the gas escapes
through the water in large bubbles, escaping through a glass
pipe, inclined at an angle at the top of the lead pipe, into the
air. When sufficient gas has escaped to reduce the pressure to
the working limit the pipe is sealed. Thus the pipe acts auto-
matically in keeping the pressure below such an amount as
would endanger the apparatus or cause joints to leak. It is
found expedient to cover all the rubber junctions in the
generating room with calico and then to paint it. Protected
in this way it will last until stopped up by the action of the
chlorine which fills it with lemon-yellow incrustation, at the
same time reducing its thickness. All junctions are secured with
copper wire where practicable.
Refining Operations. The guard, with the white pot in it
containing 2 or 3 ozs. of fused borax, is placed in the furnace,
and is heated gradually until the bottom of the white pot is
dull red. The ingots (of which the larger are slipper-shaped)
to be refined, amounting in all to 650 to 720 ozs. in weight, are
then placed loosely in the pot, the furnace filled with fuel, and
the dampers opened. As soon as the gold is melted, which
generally happens in about one and a-half hours, the boraxing
of the pots being also effected at the same time, the perforated
lid is put on, and the pipe-stem, previously brought carefully
to a red heat to prevent cracking or flaking, is pushed to the
bottom of the pot. As the pipe is being inserted, the chlorine
is gently turned on to avoid stoppage of the passage through the
stem by the solidification of metal in it. The supply of chlorine
is controlled by the glass stopcock over the furnace, and the
amount is adjusted so that the whole of the gas is absorbed and
no globules of metal can be thrown up. This can usually be
ascertained by feeling the pulsations of the gas through the
indiarubber connections as it escapes in bubbles out of the
bottom of the pipe-stem. When the gold contains much silver
or base metals, the absorption of the chlorine takes place rapidly
but gently, very little motion of the contents of the crucible
being apparent, but when the gold to be refined is of high assay
and also in all cases towards the end of refining, the gas is
admitted only in a small stream, and requires careful watching
to prevent spirting. When base metals are present in large
quantities (over 2 per cent.) dense characteristic
fumes of the
chlorides of these are given off, and the metal or metals present
may be generally identified by the fume or incrustation caused
by the condensation of the base chlorides on the pipe or lid.
Gold containing 2 per cent, of silver and 0*5 per cent, of base
metal is refined to about 995 fine in one and a-half hours, while
that containing 3-5 per cent, of silver and 1-5 per cent, of base
metals takes two hours. When larger percentages of silver or
base metals are to be dealt with, the time taken is not pro-
portionately longer, because, as mentioned above, a much greater
stream of gas may with safety be admitted, though, in all cases,
at the beginning the chlorine must be introduced gently on
account of there being air in the chlorine mains, and, also, at the
end of refining, the supply must be greatly reduced. When
" flame "
nearing completion, the issuing from the holes in the
lid becomes altered in appearance, and much smaller; it now
contains much chlorine mixed with small quantities of the
volatile chlorides. The actual completion of the operation is
generally known by the appearance of a very characteristic
"
flame," which is luminous, with a dark brown fringe. In case
of doubt, a piece of clean pipe-stem is used as a test. It is
placed, cold, for a few seconds in the issuing flame, and if the
refining is finished, a clear reddish-brown stain, tending to
yellow, is imparted to the test end. This stain consists of ferric
oxide and chloride from the oxidation of ferrous chloride, and
contains gold and sometimes chloride of silver, and is probably
caused by small quantities of chloride of iron retained by the
fused gold and non-volatile chlorides, from which it is freed by
the unabsorbed chlorine bubbling through. Traces of copper
and iron are always found in the refined gold, the bulk of the
alloy being silver. As soon as the stain is found to be of the
right colour, the current of gas is reduced, and is allowed to
pass for a further fifteen minutes, and the pipe is then with-
drawn and the clay pot lifted out of the guard. The pot is
allowed to stand under a hood (to carry off the fumes) until the
gold is set, which usually takes place in from five to seven
minutes, the fact that solidification has taken place being
observed by thrusting a piece of red-hot pipe-stem down through
the fused chlorides. The chloride is then poured into a mould
provided with a ventilating hood, which, in consequence of the
high density of the fumes necessitating a sharp draught to
remove them, is connected with the stack. Any borax poured
off with the chlorides is allowed to remain, as it is required
as a cover for the chlorides in the subsequent fusion for the
separation of their gold contents. The pot is then broken,
as the cone of gold will not fall out of it soon enough, and the
cone of refined gold is remelted in the guard and cast into two
flat ingots, 12 inches by 4 inches by 1J inches, which, when set
and still red hot, are placed on a copper lift, dipped in dilute
sulphuric acid and then in water, and after removal from the
water are still sufficiently hot to dry by their own heat. The
broken pots are ground in a small Chilian mill and panned off,
"
and the gold obtained is added to the " end that is returned at
the end of the day. 9,000 ozs., containing up to 10 per cent, of
silver and base metals, constitute a day's refining.
An improvement has recently been introduced, by which a
PARTING. 399
considerable saving of time and material is made. This is the

dipping of the fused chlorides and borax from the pot while it is
still in the fire, and without previous solidification of the
gold,
by a small clay crucible, from which they are poured into a
covered mould projecting over the furnace, the drops falling
back into the pot. This had previously been the practice when
the percentage of silver was large, as the silver doubles its bulk
on conversion to chloride, and would have overflowed. The last
" "
dip always contains some gold, and is poured into a separate
mould, in which the metal sets at once. The chloride is thence
poured into the larger mould, and the gold returned to the pot.
The chloride remaining in the pot is then made into a paste
with bone ash, after which the refined gold is stirred and cast
into ingots, the pot being at once returned to the furnace to be
used a second time.
The chlorides, which hold from 5 to 10 per cent, of gold in
feathery particles, are remelted during the day in plumbago
crucibles holding 300 ozs. of chloride. When fused, 7 per cent.
of their weight of bicarbonate of soda is added, cautiously and
without stirring, which produces a shower of globules of reduced
silver, and these falling through the chlorides carry down nearly
all the gold. As one addition of bicarbonate of soda does not
entirely free the chlorides from gold, a second addition is made,
without removing the crucible from the furnace, ten minutes
being allowed after each addition. The pot is then lifted out
and placed on one side to allow the metal to set, when the
chlorides are poured into a mould 12 inches by 10 inches by
2 inches, practically free from gold and ready for reduction.
The silvery button obtained contains from 40 to 60 per cent, of
gold, and is refined on the following day.
The silvery ingot and the refined gold contain 99*85 per
cent, of the gold issued in the morning for refining, the bulk of
the deficiency being in the pots. The amount of gold which
goes immediately into work after refining is 9 7 '6 per cent, (an
average of thirty days refining). The amount of gold left in the
silver after reduction from the chloride reaches a maximum of
1 part in 10,000, but is usually from J to ^ of this quantity.
The cakes of impure chloride, weighing about 250 ozs. each,
may be colourless and translucent to brown or chocolate colour
and opaque, the colour depending on the amount of copper salt
present. They consist of argentic and cuprous chlorides, with
traces of other chlorides and 9 per cent, of chloride of sodium
from the decomposition of silver chloride by bicarbonate of soda.
When cool, each cake is sewn up in a coarse flannel bag to
prevent loss of any silver which may become detached during
reduction, and they are then boiled with water in a wooden vat
for four or five hours. By exposure to air and moisture the cakes
become coated with a green deposit, owing to the conversion of the
cuprous chloride into cupric oxy chloride and hydrated cupric

chloride, and the successful removal of a large proportion of the
cuprous salt in the vat is due to its solubility in a hot solution
of common salt and of hydrated cupric chloride, from which it
is redeposited on dilution.
The cakes
for reduction are placed alternately with wrought-
iron plates inch thick in a cast-iron tank lined with similar
plates. The plates are prevented from touching the bags by
laths of wood, otherwise the copper would be reduced in the
bags, and would be difficult to separate from the silver. The
reduction is slow in starting, unless either some liquor is left
from a previous operation or some chloride of iron added. The
bath is heated by the direct injection of steam (this is absolutely
necessary), and the reduction is complete in from two to four
days, though sometimes it takes longer. The time may be
lessened to twenty-four hours by putting the chloride cakes in
metallic connection with the iron plates. The completion of the
reduction may be easily ascertained by feeling the cake, when, if
any chloride be unreduced, it is felt as a hard lump. The
reduced silver is taken out of the bags, washed in boiling water
for about an hour, and then melted, no fluxes being necessary.
The use of the flannel bags makes the reduced silver of high
standard, as the reduced copper is thus prevented from adhering
to the silver cake, from which it was found very hard to detach
it without tearing off silver as well. A
small percentage of
reduced silver and silver chloride is found at the bottom of the
tank, its presence being probably due to the solubility of the
chloride in solutions of the chlorides of copper, iron, and sodium.
In 1894, the mean standard of the refined silver was 982-2,
the lowest ingot being 955 fine, and the highest 998. In
1895, the mean standard was 981 6, the lowest being 936, and
-
the highest 995 fine.

The mean fineness was only about 930 prior to the year 1889.
In 1889, after the introduction of the iron plates in place of zinc,
the fineness rose to 948. The subsequent improvement was
probably due to the introduction of the flannel bags.
The silver contained in the gold operated on is distributed in
the following manner, the mean results of the last five years
(1891-95) being given :
Silver in ingots, 88 '77 per cent.

left in refined gold, . . . 7 '62
,, in "sweeps," 2'00
,, unaccounted for, . . . . 1'60
" "
The sweep from the condensing chambers amounts to about
per annum, and contains an average of 41 ozs.
3 cwts. fine
gold and 157 ozs. fine silver, which are carried over as globules
or volatilised as chloride and condensed.
PARTING. 40 ]
The mean amount of gold refined per annum during five years
(1891-95) was 949,527 ozs., containing gold 9377, silver 49-6,
and base metals (by difference) 12-7. The mean assay of the
refined gold for the same period was 995-9, and the mean loss of
gold in the refining operations for the same period was 0-175 per
thousand.
The approximate cost of refining per ounce gross weight
refined was as follows :
In 1894. In 1895.
Material, . 0'1397 of a penny. 0-1215 of a penny.
Wages, . 0-1485 0'1439
0-2882 0-2654
Half the cost for materials was for hydrochloric acid at 20, 15s.
per ton.
The amount of gold refined in 1894 was 1,049,529 ozs.,
containing in parts per thousand gold, 933*9; silver, 51-4;
base metals, 14*7 ;
and the gold refined in 1895 amounted to
1,083,243 ozs., containing gold, 932-0; silver, 52-3; and base
metals, 157 parts per thousand.
In some experiments made by Mr. Barton, who took gold
alloyed separately with 4J per cent, of copper, 4J per cent, of
lead, 4 per cent, of iron, and 4J per cent, of tin, when the cost
of hydrochloric acid was 2d. per Ib. and manganese, ore (70 per
cent, peroxide) Id. per Ib., the following results were obtained,
operating on quantities of gold containing 30 ozs. of each of
these metals :
Cost of extracting copper, . . 4-5 pence per oz. Troy.

lead, 1-4
,, iron, . . 3-9
tin, . . 2-5
The gold in each case was brought up to the usual fineness.

These results give approximately the cost of extracting these
metals in the quantities in which they ordinarily occur in
Australasian gold.
The Miller Process at the Sydney Mint. Mr. J. M'Cutcheon,
the Assayer at the Sydney Mint, writes that the process of
freeing the chlorides from gold now used by him is as
follows :-^
The chlorides produced during the operation are separated into
two classes, termed "balers" and "non-balers." The first is
that portion baled, or rather ladled, out during the operation, to
of 350 ozs.,
prevent overflow ; this is re-melted in quantities
and whilst in the molten state, half a pound of bicarbonate of
soda is projected on the surface. This has the effect of reducing
some of the chloride, and the metal in sinking to the bottom of
the pot carries with it all the gold. The "non-balers," or that
refined gold when it
portion of chloride which is poured off the
26
has set, is treated as above, but 7 ozs. of granulated zinc is
used instead of the bicarbonate.

The chlorides are poured into slabs, and are now ready for the
reduction process, in which the silver loops formerly used have
been abandoned, iron plates being now used instead of zinc ; the
water is acidulated with hydrochloric acid.
The bullion treated at the Sydney Mint during 1895 contained
gold, 832-1; silver, 135-5; and base metals, 32-4 parts per 1,000.
The average fineness of the gold produced by this method at
Sydney in the period of nine years from 1884 to 1892 was 99.V9.
The remainder is silver, which apparently cannot be profitably
removed by chlorine or by any other method. The average
fineness of the silver produced at Sydney was about 970 in 1893,
the fineness of individual bars varying from 917 to 987. The
alloying metal in the silver bars is almost all copper. The
gold retained by the silver formerly amounted on an average to
1*3 parts per 1,000, but it has since been materially reduced
and is now a mere trace. Analysis of the silver resulting from
the refinage of gold, known originally to have contained, among
the base metals in the alloy, copper, lead, antimony, arsenic,
and iron, gave the following results :
Silver, . . . .- 972 "3

Copper, . . . . 25-0
Gold 2-7
Zinc and iron, . . . Traces.
1000-0
The losses of gold in the course of the process are very small,
varying from O'll to 0'19 per 1,000; this is considerably less
than would have been lost by merely toughening the gold with
corrosive sublimate without parting it from the silver. The
loss of silver at Sydney was about 4-^5 per cent, in 1895. These
losses are reduced if the amounts recovered from the flue-dust
and from the ground-up crucibles are taken into account. The
total cost of the process is about Id. per oz. of crude gold at
Sydney, and was about 0*65d. per oz. in Melbourne in 1873,
but has since been reduced. The loss of gold by volatilisation
is probably prevented from reaching the large amounts which
might be expected from the results of Christy and of the author

(see p. 21) by the fact that, during the whole time that the
chlorine is being passed, silver and base metals are present,
and, by absorbing the gas, protect the gold from its action.
Professor Thomas Price, who treated some of the Californian
gold-bullion by this method on a working scale in his laboratory
in San Francisco, states* that, with Californian gold, which
generally contains more silver than Australian gold, the gold
*
Trans. Am. Inst. Mining Eng., 1888, vol. xvii., p. 30.
PARTING. 403
taken up by the chloride of silver amounted to 5 per cent.,

and even to 10 per cent, of the total weight of the
gold. For this reason, and on account of the large amount of
silver bullion in the San Francisco market requiring parting,
Professor Price considers that the Miller process, while techni-
cally successful with Californian gold, is hardly able to compete
commercially with the ordinary sulphuric acid process. He
suggests that it might be well adapted for refining the nearly
pure brittle gold produced at chlorination works, where chlorine
is at hand and other methods of refining are not convenient.
The idea of refining brittle gold in this way had previously
been acted on by Professor Roberts-Austen in the year 1871 in
performing the experiments and operations detailed below.
Refining Brittle Gold by Chlorine Gas. In spite of the fact
that brittle ingots of gold are not accepted for coinage at the
Royal Mint, a number of such ingots formerly found their way
into work there, their lack of ductility not being observable
until after they had been alloyed with the copper requisite to
make standard gold, which is 916'6 fine. In 1871, no less than
40,000 ozs. of such brittle standard gold had accumulated at the
Mint, having been set aside as unfit for coinage. These bars
contained as impurities traces of antimony, lead, bismuth, <fec.
In order to determine how far such impurities could be elimi-
nated by chlorine gas, applied as in Miller's process, the following
experiments were conducted by Professor Roberts- Austen at the
Royal Mint:*
1. A barof standard gold, weighing 241*20 ozs., which broke
with a slight tap from a hammer, was melted in a clay crucible,
and chlorine passed for three minutes. The metal was found to
be perfectly toughened and to have sustained a loss of weight of
only 0*17 oz., which proved to consist entirely of base metal.
2. Gold bars
containing 0*5 per 1,000 of antimony and 0-5 per
1,000 of arsenic, or a total of I'O per 1,000, were converted by a
stream of chlorine into gold of excellent quality in three and
a-half to four minutes.
3. Abar of standard gold, weighing 301 ozs., in which the
impurities noted below were contained, was operated on. The
base metals consisted of:
Antimony,
Lead, .
This gold was as brittle as loaf-sugar, and could be broken with

the lingers. After chlorine had been passed for eighteen
minutes, the metal was perfectly ductile. The loss of metal was
found to be
Copper,
Gold,
....
.....
Base metals (as above), . . 4*51 ozs.
3-44
-15
Total, . . 8-10
These experiments having been successful, the 40,000 ozs. of

brittle gold were treated similarly in graphite crucibles, the
charges being about 1,100 ozs. of gold in each. The time of
passage of the chlorine varied from five to seven minutes, and the
"
total loss of gold after recovery from the sweeps," was returned
as only T^ oz. With this loss and at a very slight expenditure,,
the whole of the stock of brittle gold was toughened and rendered
fit for coinage. Considering the success ot this venture, it ia
surprising that the method has not yet found wider application
in the treatment of brittle bars.
Parting by Electrolysis The Moebius Process. This process
was patented by Mr. Bernard Moebius, in England, as long ago
as the year 1884 (Dec. 16, No. 16,554), and is now in successful
operation in several localities in the United States and Germany.
It is said* to be specially suitable for refining copper bullion
containing large proportions of silver and gold with small
quantities of lead, platinum, and other metals, but is chiefly used
in parting dore silver.
The apparatus required consists of a number of wooden vats
coated inside with graphite paint, and filled with a solution
containing 1 per cent, of nitric acid, which constitutes the elec-
trolyte. The anodes consist of plates of bullion of about J inch
thick and 14 inches square, which are hung in muslin bags
destined to catch the insoluble impurities after the silver, copper,
<fec., have been dissolved. The cathodes consist of plates of pure
silver, slightly oiled to prevent adhesion of the deposited metal.
These plates are continually scrubbed by a mechanical arrange-
ment of brushes by which their surfaces are kept free from loose
crystals of electro-deposited silver. This loose silver falls on to
trays placed below, which are removed at intervals, and the
silver collected from them.
The current should have an electromotive force of from one
to three volts for each vat. The copper is not deposited unless
the solution becomes too weak in silver or too rich in copper,
and even if some happens to be deposited, it falls into the cathode
trays with the silver and does little harm, since it will be-
* Gore's 240.
JSlectroiytic Separation oj Aletals, London, 18l>0, p.
PARTING. 405
gradually redissolved if the conditions are corrected. When too

much copper has accumulated in the solution the latter must be
removed, the method being as follows :The bullion anodes are
replaced by carbon ones, and a weak current passed until all the
silver is deposited. The silver cathodes are then replaced by
copper ones, and a strong current passed so as to deposit the
copper as rapidly as possible as a loose powder, which falls into
a copper box. This box is connected with the cathode to prevent
corrosion by the acid which is set free. The liquid thus regen-
erated is used again as the new electrolyte.
The process is stated to be the cheapest parting process known.
If no copper were present in the bullion it is clear that there
would be no consumption of acid, and it would never be necessary
to change the electrolyte. Under the most favourable conditions,
therefore, with water power available and the amount of copper
in the bullion very small, the cost of parting dore silver would be
merely nominal.
The following description of the Moebius plant which is in
successful operation at the Pinos Altos Mine, Chihuahua, Mexico,
is an abstract of that given by Mr. George Maynard.* The
plant is capable of treating from 3,500 to 4,000 ozs. of dor6 silver
in 24 hours, and consists of a tank 12 feet long, 2 feet wide and
20 inches deep, divided into seven compartments. Four silver
plates (the cathodes) and six dore plates (the anodes) are placed
in each compartment in such a way that an anode is opposite to
each face of a cathode. The trays to catch the deposited silver
have perforated bottoms and are covered with asbestos cloth,
-and in order to facilitate cleaning-up, the electrodes, frames,
anode bags, trays and brushes can all be lifted up simultaneously
by a hoisting arrangement worked with a crank and handle, so
that the exciting liquid alone remains in the cells.
The bullion is from 800 to 900 fine in silver, and 25 to 50 in gold,
the rest being chiefly copper. The exciting liquid, which at first
contains 1 per cent, of nitric acid, is gradually converted into a
solution of silver and copper nitrates. The silver nitrate is con-
tinually decomposed by deposition of the metal, but the copper
nitrate accumulates, fresh acid being added at intervals to prevent
the copper from being deposited. The silver dissolves from the
anodes and is precipitated on the cathodes in the form of heavy
tree-like crystals. The action of the brushes or scrapers is of
vital importance to the success of the process, the advantages
derived from their action being summarised as follows :
1. Theliquid is agitated and so kept homogeneous.

:. Polarisation is prevented.
3. The electrodes can be brought very near to one another,
and the resistance thus reduced without any fear that short
circuiting v/ill ensue by the bridging over of the space between
them by crystals of deposited silver.
*
Eng. and Mng. Journ., May 9, 1891, p. 556.
The lead (as peroxide), platinum metals, antimony and other

impurities remain with the gold in the bag surrounding the
anodes. When the exciting liquid becomes too highly charged
with copper, the solution is used instead of bluestone in the
amalgamating pans of the mill, but the copper could of course be
recovered as usual if it were desirable. The manual labour is
all performed by the assayer and his assistant ; the cathode trays
are hinged, and by letting them turn on their hinges the silver is
let fall into a movable tank on castors furnished with a false
bottom, on which it is washed and dried, and it is then ready to
be melted into bars of about 999 fine. The gold slimes are simi-
larly washed and filtered ; when they are fused, the lead is
almost all slagged off.
The electric currentemployed is 170 amperes, the electro-
motive force being 8 volts this requires an expenditure of
;
2J H.P. to refine from 3,500 to 4,000 ozs. per day. The cost of
parting is said to be less than ^ cent per gross oz. of bullion, and
the royalty is ^ cent per oz.
The
original cost of the plant is said to have been about $6,000,
including $1,000 for the silver in the cathode plates. This sum
included the cost of conveying the plant to the mine, which was
very high, as a journey lasting several weeks on mule-back had
to be performed. The amount of silver contained at any one
time in solution in the bath is about 300 ozs. The weight of
the forty-two anode plates is about 4,200 ozs. when they are
fresh, and this silver can be melted up and recovered about forty-
eight hours after tHe operation has been started, so that in
this plant only about 10,000 ozs. of bullion are necessarily locked
up continuously in the apparatus. The clean-up takes place
once a month.
The Moebius process has also been in successful operation at
the works of the Pennsylvania Lead Company at Pittsburg since
September, 1886 ; here it is said that from 30,000 to 40,000 ozs.
of dore bullion are refined daily at a cost of three-fourths of a cent
per oz. The silver produced is from 999 to 999*5 fine. A small
plant was also erected at the Kansas City Smelting and Refining
Company's works, and a large one in 1891 at St. Louis. The
process is also worked by the Norddeutsche Refining Company
at Hamburg.
THE ASSAY OF GOLD ORES. 407
CHAPTER XIX.
THE ASSAY OF GOLD ORES.
THE assay of gold ores is almost universally conducted in the
dry way i.e., by furnace methods. Exceptions will be noted
later. The plan of operation is to concentrate the precious metal
in a button of lead in one of two ways, viz. :
(1) By fusion in a
crucible; or, more rarely, (2) By scorification. The button of lead
obtained by either method is then subjected to cupellation, by
which the lead is oxidised and removed, and the resulting bead
of precious metal is weighed. Since in these operations silver
and the metals of the platinum group remain with the gold,
they are subsequently separated by inquartation and parting, and
in the case of platinum and its allies by further special methods.
The exact method to be used in any particular case varies
with the richness of the ore, the nature of its gangue and the
presence or absence of compounds of the base metals. As a
general rule, poor ores i.e., those containing less than 2 ozs. gold
per ton are better assayed by the fusion process so that a com-
paratively large quantity of material may be operated on. Rich
ores may be assayed either by fusion or scorification, the errors
arising from the small amount of material
used in- the latter
Telluride ores,
process being less important in their case.
arsenical and antimonial ores, and ores containing tin, nickel, or
cobalt must all be scorified if possible, but in these cases it is
better to make the ordinary crucible assay, and then to scorify
the lead button obtained. Either method can, however, be used
for any ore.
Assay by means of the Blowpipe* This method, though less
exact than that made in a furnace, is of importance, because in
its means not only to
prospecting expeditions it is possible by
detect the gold and silver in any ore, but also to determine its
amount quantitatively with fair accuracy. On such expeditions
to
it is impossible to carry the cumbrous apparatus required
make an ordinary The amount of powdered ore taken is
assay.
usually 100 milligrammes, and this is
mixed with borax and
about 1 gramme of granulated lead. The whole is wrapped in
flame of a blow-
paper and heated on charcoal in the reducing
pipe until the fusion is complete, and
then for a short time with
the oxidising flame. The lead is then separated from the slag
* For a full
description of this method, see
Plattner's Manual of
Analysis with the Blowpipe, pp. 360-406. London, 1875.
and heated on a bone-ash cupel until it is all converted into

litharge. The diameter of the button of silver and gold thus
obtained is carefully measured on an ivory scale, which at once
gives the percentage amount in the ore. The gold is usually
separated from the silver by parting in nitric acid, but Richards
has recently stated* that the silver can be distilled off by the
blowpipe, leaving a bead of pure gold, which can be measured.
Sampling the Ore and Preparation of the Sample for
Assay. The value of an assay depends largely on the care with
which a sample of the ore is selected. The sampling is done by
processes which are as far as possible automatic, and independent
of the will or judgment of the assay er. The best and most
widely applicable method is that of dividing and intercepting a
part of a stream of ore falling vertically downwards or sliding
down an inclined plane. This is done in many ways in mines
and mills. A convenient method in use in Colorado may be
described as follows : The mineral, having been passed through
a stone-crusher and one or two pairs of rolls so as to be reduced
to the size of coffee beans, falls through a vertical tube, on to the
apex of a cone. The surface of the cone is inclined at an angle
of 45 q to the vertical and has one or more openings or windows
bounded by straight lines drawn from the apex to the base. The
ore, falling on the cone, slides over its surface into the ore-bin,
excepting a portion, usually one-sixteenth, which passes through
the openings in the cone and is conveyed down a tube to another
receptacle. The operation may be repeated on the sample thus
taken, after it has been further crushed.
In the assay office the sample, however obtained, is further
reduced in bulk by the implement known as the sampling tin.
This consists of a series of troughs arranged side by side and
fastened at equal distances from each other (the width of the
spaces being equal to that of the troughs) by strips of metal
soldered on to their ends. An even stream of ore being let fall
from a shovel on to this sampler, half is retained in the troughs
-while half passes through. Careful experiments have proved
that each half is representative of the whole. A
repetition of
the process reduces the sample to any required extent. A
sampler with troughs 1 inch wide is suitable for treating
materials which include lumps of not more than J to J inch in
diameter. For finely ground materials a convenient width for
the troughs is f inch.
In taking a sample from a large bulk of ore, or where
sampling tins cannot be procured, the method of dividing into
quarters may be employed. The ore is piled into a heap,
thoroughly mixed and divided into four by a shovel by two
lines at right angles to each other. The two opposite quarters
are removed, and the remainder again mixed and divided as
* Journal
of the Franklin Institute, June, 1896.
THE ASSAY OP GOLD ORES. 409
"before. When the sample is reduced to a few pounds in weight

it isbroken down to the size of coffee beans, reduced in bulk by
quartering or the sampling tin, crushed finer, again sampled and
so on, until finally, when reduced to a mass of from
J Ib. to 1 lb.,
it is all crushed fine
enough to pass through a 60-mesh or
80-mesh sieve (i.e., one containing 80 holes per linear inch).
Before this fine crushing can be done, it is usually necessary to
dry the sample ; the percentage of moisture should always be
estimated at the same time by weighing the sample both before
and after it is dried.
The implements employed for crushing samples of ore are
various. A
small rock-breaker, with reciprocating jaws, similar
to those used on the large scale, and worked by hand or steam
power, is useful for breaking down large lumps, which otherwise
may be broken by a hammer. For finer pulverisation a small
pair of steel-faced high-speed rolls may be used if steam power is
available. This method is adopted at large smelting or sampling
works, where great numbers of samples are crushed daily. In
smaller works or offices the buckboard (Fig. 59) is most suitable.
Fig. 59.
Scale, f in. = 1 ft.
It is a smooth plate of iron about 2 feet square with a 1-inch
rim surrounding it on two or three sides. On this a bucking
hammer is worked a heavy piece of iron 5 to 10 Ibs. in weight,
with a large smooth curved face and a handle 30 inches long.
It is moved about on the iron plate (on which the ore is spread)
with both hands, one holding the handle, the other pressing the
head downwards, the curved face being below, while an oscillatory
movement is imparted by the handle. The instrument is very
effective if the ore is previously broken down to the size of
coarse sand in a mortar. The pestle and mortar are of value in
breaking down samples from the size of nuts to that of coarse
sand. In grinding down siliceous material, so as to enable it to
pass through an 80-mesh sieve, the pestle and mortar is far
inferior to the buckboard.
" Metallics." In many ores, both gold and auriferous silver
"
occur native in grains or threads. These " metallics are not
readily reducible to a fine state of division, and, though a part
always passes through the sieve, some of the larger pieces which
have resisted abrasion fail to do so. In some assay offices part
of the pulverised ore is thrown back into the mortar with the
metallics, and grinding is continued until everything passes the

sieve. This is a dangerous practice, as it is impossible to ensure
the even distribution of metallics through the sample. The
smaller pieces which pass through the sieve in the first instance
constitute an unavoidable evil which is increased by every piece
of metal that follows them. The safer plan is to cupel by
themselves the whole of the metallics left on the sieve, and
calculate their value per ton of ore independently of the result
obtained by the ordinary assay. The total value of the ore is
found by adding these two results together.
The prepared sample may be stored in tin boxes or glass jars,
which should be labelled by numbers, none of which are ever
repeated. Before weighing out the powdered ore for assay, the
whole sample should be turned out into a wide bowl or on to
glazed paper, or, better still, rubber cloth (which does not crack
and wear out like paper), and thoroughly mixed with a spatula.
The sample should never be mixed by shaking, and care should
be taken to avoid jarring it after mixing, as metallics in that
case tend to settle to the bottom from their superior density, and
a fair sample cannot then be easily obtained. For the same
reason the part of the mixed sample which is taken for assay
should not be hastily shovelled on to the balance pan from the
top of the pile, but a vertical slice should be taken, some of the
lowest layer being carefully scraped up from the rubber cloth.
FUSION OR CRUCIBLE PROCESS OP ASSAY.
This process is divided into three parts, viz. :

(1) Fusion; (2)
Cupellation; (3) Parting.
(1) Fusion. The object of this operation is to concentrate
the precious metals in a button of lead, while the whole of the
remainder of the ore forms a fusible slag with suitable reagents,
in which lead sinks. The fusion is made in clay (or rarely iron)
crucibles in a wind furnace, the fuel being coke, charcoal or
anthracite. The size of the furnace depends on the amount of
work to be done and on the fuel employed, charcoal requiring
more space than coke. If only one fusion is to be made at once,
the internal dimensions of the fire-box may be 8 inches square,
while a furnace 14 inches square will hold nine crucibles. The
furnace is exactly similar to that used for melting bullion
described on p. 359.
In the United States it is usual to perform fusions in a muffle
furnace, similar to that described below under cupellatlon, p. 406.
The temperature required is about the same as that used in
scorification. The advantages claimed are greater cleanliness
and neatness and more uniformity in the conditions, the tem-
constant and uniform
perature of a muffle being more easily kept
FUSION OR CRUCIBLE PROCESS OF ASSAY. 411
than that of an ordinary fusion furnace. Six or eight fusions

can be performed at one time in a large muffle.
There are three shapes of clay crucibles in common use, namely
the French, the Battersea, and the Colorado. The French pot
differs from the others in the thickness of its walls near the
bottom, and consequently it requires very careful annealing*
The Battersea pot is of coarser texture than the others, having a.
larger percentage of sand in its composition. The Colorado-
crucibles, manufactured by the Battersea Company, are specially
made for fusion in the muffle; the "20 gram" size is the most
generally useful.
In weighing the materials for a crucible charge the use of a.
set of assay-ton weights saves much labour in calculation. The
assay-ton is a weight which contains as many milligrammes as a.
ton contains ounces. Thus an English ton of 2,240 Ibs. contains
32,666 Troy ounces, so that the corresponding assay-ton must
weigh 32,666 milligrammes or 32*666 grammes. Now suppose
the weight of the resulting bead of gold (or silver) from an assay-
ton of ore to be 1*5 milligrammes, it is obvious that the ore
contains 1*5 ozs. gold per statute ton. If the value per ton of
2,000 Ibs. is required, the weight of the assay-ton is 29-166
grammes, since there are 29,166 Troy ounces in 2,000 Ibs. avoir-
dupois. If grain weights are preferred the weight of the assay-
ton may conveniently be taken as 326-66 grains (or 291*66 grains
for the short ton) ; the weight of "the resulting bead of gold in
hundredths of a grain then gives the value of the ore in ounces per
ton. Sets of weights ranging from i A.T. (assay-ton) to 4 A.T.
can be bought, or they can be made up from an ordinary box
of decimal weights. In the following pages this system is used
in giving the weights of fluxes, &c., required the grammes or
:
grains of the various authors referred to are converted into

the approximately corresponding number of assay-tons.
General Charges. The charge for a fusion assay varies accord-
ing to the nature of the ore. The following proportions are
supposed by their respective authors to be suitable to all gold
and silver ores
*
Mitchell's Formula
Ore, 1 A.T.
Soda carbonate,
Litharge, ,
Borax glass,
Salt to cover, ,
and the amount of argol or of nitre necessary to ensure the

formation of a button of lead weighing about 200 grains. These
quantities can only be determined by a preliminary assay.
* Manual of Assaying,
Mitchell's 1881, p. 546.
Aaron's Formula*
Ore, 1 A.T.
Soda carbonate, 3
Litharge, 1
Borax, .
Sulphur,
Flour, .
A3 nails.
Iron,
Glass.
Salt to cover.
*{
Melt and leave in a hot fire about twenty minutes after fusion."
When ores contain only small quantities of base metals the
following formula is recommended by Brown & Griffiths f
Ore, 1 A.T.
Soda carbonate, .
If
Litharge,
Borax glass,
.
H
Carbonate of potash,
Silica, .
Charcoal, . '6 gramme.

Salt to cover.
Percy's Formula J for ores containing only small proportions of base

fi-f Q O
metals 1
Ore, i to 1 A.T.
Red lead,
Soda carbonate and borax, 1 together.
Charcoal, 13 to 17 grains.
The charges given above are varied according to circumstances.

One A.T. of the ore is a suitable quantity if the value in gold is
from *5 oz. to 10 oz.s. or more. With very poor ores 2, 3, or 4
A.T. may be taken, and with very rich ores J A.T. may suffice.
The fluxes are altered in proportion. Formerly the usual
practice of assayers was not to reduce all the red lead or
litharge; part was left in the slag, forming easily fusible silicates.
Aaron was the first to describe the method of reducing the
whole of the litharge employed; the gangue is in this case slagged
off by soda carbonate, borax or silica, and the base metals
separated as a matte. The advantages claimed for this method
are that the button is never much contaminated by copper, and
that the crucible is but little attacked. Brown & Griffiths
observe that beginners find Aaron's method less easy than the
other; but Aaron's method is now always employed at the Royal
College of Science, London, except in rare cases, such as, for
instance,when much copper is present.
The following remarks on the use of the fluxes may serve as
an aid in making up charges in particular cases Litharge or :
*
Aaron's Assaying, 1884, p. 53.
t Brown and Griffiths' Manual of Assaying, p. 174.
I Percy's Metallurgy of Silver and Gold, p. 245.
red lead is added in the proportion of at least one part to two of

ore ; if too much litharge is used the slags are not clean,
as a slag containing lead may mean a loss of silver and gold.
Whatever method is used, the amount of lead to be reduced
should be from 250 to 450 grains. Raw ores or regulus contain-
ing much sulphide of copper may be fused with from 4 to 6 A.T.
of litharge to 1 A.T. of ore. In this case the other fluxes, except
sand, may be omitted. Some assayers prefer to concentrate
the copper as a regulus, and then to treat this over again little :
more than the ordinary quantity of litharge is then used, and

not much iron.
The amount of the charcoal added varies with the reducing
power (percentage of ash, &c.) of the particular sample which is
employed, as well as the degree of oxidation of the ore. In some
highly basic oxidised ores as much as 30 to 40 grains of charcoal
powder is required for 1 A.T. of ore. If there is much Fe 2 O 3 in
the ore (e.g., roasted pyrites) the slag is often rich. The oxide
must be completely reduced to FeO, hence the need for the large
quantity of charcoal. In such cases 3 or 4 per cent, of the gold
may be lost.* The remedy usually adopted is to increase
largely the quantity of soda carbonate, while sand must also be
added to prevent the crucible from being perforated by the
scouring action of the ferrous oxide. If ores contain much
sulphur no charcoal is used, and it may be necessary to add nitre
to burn off the excess of sulphur ; otherwise a rich matte may
be formed, or, if the old practice of adding much red lead is
followed, the amount of lead reduced may become too great.
The addition of much nitre is, however, to be deprecated since
the pot is liable to boil over; with very large quantities of
sulphides it is better to make a matte and treat
the latter again.
Carbonate of soda is used to flux silica, while borax is valuable
in basic ores to prevent corrosion of the crucible, and render the
are judged from
slag more liquid. The relative amounts required
the appearance of the ore in the first place, and afterwards
modified according to the success of the fusion. Even when the
ore is of silica, some borax is added. The
entirely composed
most convenient form is borax glass.
Silica is used only for ores full of lime, baryta, compounds
of the base metals, (fee., or generally whenever the ore does
not contain much quartz. It aids fusion in these cases,
and protects the crucible from corrosion. Fluorspar is added
to the charge when the ore contains sulphates of barium or
it increases the
calcium, and in the fusion of cupels. Like borax,
of almost but it attacks the crucible, and care
fluidity any charge,
must be taken to avoid deficiency of silica when it is used.
In general, it may be noted that for basic impurities an acid
flux is used, and for an acid gangue a basic flux.
*C. & J. J. Beringer's Assaying, p. 125.

The ore is weighed accurately to ^ grain unless very poor,

when a less exact approximation is sufficient. The charcoal is
always weighed with great care ; the litharge is best measured
by a ladle or shot measurer ; the fluxes may also be measured
out by a ladle. By practice the assay er becomes very rapid in
measuring out reagents. The various ingredients are thoroughly
mixed together on rubber cloth or in the crucible in which the
fusion is to take place. Part of the borax is kept separate and
used as a cover, being put on the top of the rest of the charge
after transference to the crucible. Acover of common salt is
used by German and American assay ers for its employment
;
at the Royal College of Science see Dr. Ball's remarks on

p. 416. Iron is added to the charge in the shape of large nails
or hoop-iron, or even scrap, if sulphur or arsenic is present.
Sulphur is thus kept out of the lead button.
The crucible is carefully annealed in the ashpit of the furnace
before using. It is lowered into a hollow in the fuel of the
furnace made by piling the coke round an old pot and then care-
withdrawing the
fully latter. The tongs
most used are shown at B, Fig. 60, A
being those used when the fusion is per-
formed in a muffle. The fire should be at
a good red heat on charging-in, and care
must be taken that no coke in the furnace
is black. If the upper layer of coke is
much colder than that below, the top of
the charge in the crucible remains un-
melted while the bottom begins to effer-
vesce, and the crucible may froth over
and part of its contents be lost. As the
charge melts the fire is urged, and brisk
effervescence occurs chiefly from the escape
of carbonic acid from the soda carbonate
as it unites with silica. After a lapse of
from twenty to forty minutes the charge
is in a state of tranquil fusion, with the
Fig. 60.
exception perhaps of slight action round
the sides or next the iron, if any is used. The crucible is then
lifted out of the fire by the tongs, the nails withdrawn and any
lead adhering to them shaken off into the pot. The pot may
now be tapped on the floor to assist the lead to settle in the
slag, allowed to cool and broken by a hammer to extract the
lead button, or the charge may be at once poured into an iron
mould. The mould must be cleaned, blackleaded and warmed
before being used, and, after pouring, it is tapped on the bench
to collect the lead at the bottom.
"If the charge does not fuse completely, so that the slag is pasty
or has lumps in it, it is advisable to recommence the assay,
making such alterations in the charge as experience

suggests,
having regard to the considerations mentioned above, p. 413.
Wnen the mould is quite cold its contents are readily separated
from it if the precautions mentioned above are taken, and the
slag is detached from the lead button with a hammer on an
anvil. The slag should be glassy and homogeneous ;
if it is
streaked it isprobable that the fusion h-is not been perfect. It
is most often black and opaque, but is red owing to the presence
of cuprous oxide if the ore contains much copper.
The lead button should be soft and malleable. If it is hard
or brittle it may contain sulphur, arsenic, antimony, copper, &c.
Sulphur and arsenic are kept from entering the lead by the
addition of iron, and then form with the iron a regulus or speiss,
which separates as hard blackish-grey or white layers found
just above the lead. They are richer than the slag, and may
often yield appreciable quantities of gold on further treatment
by scorification or roasting and fusion. Antimony makes the
lead hard, white, brittle and sonorous ; it can be removed by
adding nitre to the fusion charge, but, according to Rivot,* it is not
a source of loss if forming less than 1 per cent, of the lead button
(see under Cupellation below).
The lead must be completely
freed from the slag, very small quantities of the latter interfering
seriously with the cupellation, forming a scoria and occasioning
loss.
On the subject of fusing gold ores, Dr. E. J. Ball, who was
until lately the Instructor in Assaying at the Royal College of
Science, London, writes as follows :
" I find a
very good plan in assaying a gold (or a silver) ore to
be as follows, noting the points :
"(1) That the main aim at the beginning of the assay is to

produce an intimate contact between the molten reduced lead
and the gold freed by the original grinding of the ore.
"
(2) That when this has
been effected, the particles of ground-
up ore must be, as it were, 'atomically' ground-up by solution, in
order to liberate enclosed particles of gold.
" The first of these a
stages should not be accompanied by
fusion of the charge, because immediately that happens the lead
sinks to the bottom of the crucible, and is at once removed from
the possibility of contact with the gold floating about in the
charge. To bring about this contact in the second stage, the
bath must be as thin-fluid as possible, and convection currents
should be induced by irregular heating of the crucible, in the
the particles of
hope that, in the course of one of their gyrations,
gold liberated by the solution of the quartz particles may strike
against the surface of the molten lead at the bottom of the
crucible.
" The maximum in
quantity of lead which the ordinary cupels
*
Traitt de Docimasie.
use at the Royal College of Science will take is about 450 grains.
I therefore recommend (when treating fairly pure quartz, con-
taining say 1 oz. of gold per ton) that the charge should be made
up as follows :
Ore (passed through an 80-sieve), . 1,000 grains.

Red lead, '
. . . 500 ,, \'
Charcoal, '. . . . 25-30
The charcoal and red lead are first mixed together, and the ore
is then carefully incorporated with them. Then 250 grains of
sodium carbonate are roughly stirred in, so as to prevent the
formation of a sort of sand-bottom, which would not dissolve in
stage 2.
" The charge is then maintained at as high a temperature as
possible, actual fusion being avoided for fifteen minutes, when
another 1,000 or 1,200 grains of sodium carbonate are charged-in
little by little, and the temperature raised to about 950. At
this stage, any necessary addition may be made in order to make
the bath fluid, borax for instance, but borax seems to give low
results if added at the beginning of the assay. I always add a
piece of hoop iron to help to decompose any lead silicate or
sulphide.
" I never recommend the use of
salt, but sometimes when a
large excess of sodium carbonate has been added, the 'boil' at
the end seems never likely to stop, owing to the action of the
acid crucible on the basic charge. In that case, a little salt
stirred into the bath seems to volatilise, prevent the contact of
the charge with the walls of the crucible for a moment or two,
and to quiet the bath, enabling the charge to be poured
properly."
Roasting before Fusion. Ores containing large quantities of
sulphur, arsenic or antimony may often be roasted with advan-
tage as a preliminary to fusion. Roasting is effected in shallow
circular clay dishes, in a muffle, or in the crucibles in which
the fusion is afterwards performed. The temperature must be
kept low at first and the ore frequently stirred with an iron wire
or spatula, to prevent fritting, and to expose fresh surfaces to
the air. The roasting takes place in two stages at first, sulphur
:
dioxide, arsenious oxide (As 2 O 3 ) and antimonious oxide (Sb O 3 )

are formed and volatilised, the sulphur burning with a blue flame.
The formation of lumps is most to be feared during the first few
minutes of the operation, and can scarcely be prevented if much
sulphide of antimony is present; in this case an equal weight of
pure silver sand is mixed with the crushed ore before charging it
into the muffle.
After a time the blue flame disappears, the odour becomes less
strong, and sulphates, arseniates :md antimoniates form. By
raising the temperature sulphates are decomposed, but arseniates
FUSION OR CRUCIBLE PROCESS OF ASSAY. 41J
and antimoniates are stable at high temperatures and cause loss

of silver in the fusion. To prevent their formation the ore
should be roasted in a coke furnace, starting to heat it very
gradually and admitting a limited supply of air. In all cases the
roasting is nearly complete when the glow caused by stirring is
shown only by a few specks of ore; the temperature may then be
raised to a strong red heat without danger of fusion. The opera-
tion is complete when the ore remains of a uniform colour on
stirring. The fusion of roasted ores requires less lead than raw
ores and more charcoal powder, the amount of the latter needed
being sometimes as much as 40 grains per A.T. of ore.
Cleaning the Slag. The slags of very rich ores may retain
enough gold to necessitate further treatment. The slag is roughly
crushed and fused with 300 to 500 grains of litharge, 15 to 20
grains of charcoal and a little carbonate of soda, the same crucible
being used again. If any regulus or speiss forms during the
first fusion it must be preserved with special care and re-fused,
charcoal being omitted. The button of lead from the second
fusion is cupelled with the first button, or alone ; the slag is
almost always poor enough to be thrown away. It is not
usually necessary to clean the slag of ores containing less than
5 ozs. gold per ton.
The Treatment of Base Ores. The difficulty of roasting
arsenical and antimonial ores may be avoided * by taking ore, 1
A.T. ; red lead, 1,000 grains; sodic carbonate, 500 grains; potassic
ferrocyanide, 550 grains, with a cover of salt or borax. The
button is scorified, together with the matte formed, before cupella-
tion. It is perhaps better to scorify these ores as well as those
containing much copper if they are rich enough. If poor, copper
ores may be treated in three ways,f so that each method may
serve as a check on the others, viz. :
(a) Fusion with iron nails:
the lead button becomes cupriferous, and should be scorified
together with the matte ; (b) roasting, followed by fusion and
scorification ; (c) treatment with nitric acid, by which all the
sulphur and copper are removed. The silver dissolved in the
liquid is then precipitated by a solution of common salt, of
which a large excess should be avoided. The insoluble residue
is dried, and can now be readily fused and cupelled. By
treatment (c) the lead button is kept free from copper, the
presence of which in the lead obtained by methods (a) and (b)
renders cupellation difficult and unsatisfactory. As already
observed, some to concentrate the copper as a
assayers prefer
regulus. The regulus first obtained still retains gold, and must
be re-treated.
Cupellation. This operation is conducted in a muffle furnace,
the construction of which is shown in Figs. 62-64, p. 436. The
*
Rickett's Notes on Assaying, New York, 1887, p. 77.
t Percy's Metallurgy of Gold and Silver, p. 247.
27
fire is lighted, a little bone-ash is sprinkled on the floor of the

muffle to prevent its corrosion by litharge in case of the upsetting
of a cupel, and the cupels introduced as soon as a bright red
heat is attained. Cupels are little cups made of bone-ash, and
are either round or square. In their manufacture the bone-ash is
finely powdered so that it will pass a 40-mesh sieve, then slightly
moistened with water (to which a little carbonate of potash is
sometimes added), put into a mould (Fig. 65) and compressed
by the blows of a mallet so as to cohere firmly. Cupels must be
dried very carefully and slowly but completely, as otherwise
cracks appear when they are heated, and loss is thereby occa-
sioned. The cupels at the Royal Mint are made two years
before being used, and are dried slowly on shelves at some
distance from the furnaces. If too much water is used in
mixing the bone-ash, the cupels lose part of their porosity if :
too little is used they are too soft and crumble readily. About
1 oz. water to each
troy pound of bone-ash answers very well.
The cupels having attained the same temperature as the
muffle, the lead buttons obtained as described on p. 415 are
charged in by the tongs (A, Pig. 61). The buttons collapse
r.
Fig. 61.
and lose their shape almost instantly if the temperature is

sufficiently high, but the molten mass formed is covered by a
black crust for several seconds later. The crust then breaks up,
and the brilliant surface of a liquid bath is seen. The muffle
door should be closed immediately after the charging-in is
completed, and kept closed until all the assays have thus
" uncovered." The door is then
opened a little way to let in a
current of air by which the lead is oxidised, and the litharge,
floating to the edge of the bath, is absorbed by the cupel,
together with the oxides of other base metals, which are not
taken up by bone-ash if they are in a state of purity. Uncom-
bined metals are not readily absorbed by the cupel, and only
traces of gold and silver are carried into it by the litharge.
These quantities depend partly on the nature of the cupels,
coarse-grained bone-ash absorbing more of the precious metals

than fine-grained. By a rise in temperature the absorption of
gold is increased. The presence of other base metals also has
some influence on the amounts of precious metals absorbed, oxide
of copper in particular carrying with it much gold into the cupel.
As the cupellation advances the lead bath is reduced in size
by oxidation and absorption ; reddish patches float slowly over
its surface, appearing earlier in the richer assays ; it becomes
more convex and brighter the red spots move more quickly
;
and finally whirl round with great speed and then disappear ;
moving iridescent bands take their place for a moment and then
disappear likewise, and the bead becomes suddenly much duller
in appearance, thus indicating that the cupellation is at an end.
The temperature must be raised towards the end of the operation
to remove the last traces of lead, and the beads left for from
three to ten minutes, after all apparent oxidation is at an
end. The cupels may then be removed from the muffle, provided
the ore is poor or has little silver in it. It must, however, be
remembered that silver absorbs oxygen when molten and gives
it off suddenly when solidifying, so that if the bead weighs more
than O'Ol gramme (Rivot) little fountains of metal are thrown
up and some "part may be projected
"
out of the cupel.
"
This
"
sprouting," spitting," or vegetation may take place in
argentiferous-gold beads if the gold does not exceed one-third of
the silver (Levol). At the Royal Mint it is found that a still
larger percentage of gold does not prevent spitting unless a trace
of copper is present. Where spitting is to be feared, therefore,
either some copper is left in the bead (vide infra, p. 440), a course
which is inadmissible if the silver is to be estimated, or else
slow cooling is resorted to, the muffle being carefully closed and
luted up and the fire withdrawn. The door is not opened again
until the beads have solidified; under these circumstances, no
sprouting occurs. Slow cooling may also be obtained by cover-
ing the cupel containing the silver bead with a red-hot empty
cupel.
When cooled the beads often "flash" i.e., brighten suddenly
at the moment of solidification. This is due to the fact that the
latent heat of fusion being released raises the temperature of the
bead enormously, the metal having been in a state of surfusion
at a temperature many degrees below its melting point. The
flashing of small beads can rarely be observed.
The proper temperature for cupellation of gold ores is higher
than for that of silver, as loss by volatilisation is less to be
feared. The muffle should be at a bright orange-red heat, the
cupel red, and the melted lead much more luminous than the
cupel ;
the fumes should rise slowly to the crown of the muffle.
In Western America in both silver and gold assays the heat is
kept low enough to enable crystals of litharge to form in a ring
round the cupel, but the results thus obtained are too high, the
lead not being completely eliminated. The whole of the litharge
should be completely absorbed. The formation of scales, due to
low temperature, is accompanied by a sluggish heavy movement
of the fumes, which fall in the muffle. On the other hand, the
heat is too great when the cupels are whitish, the fused metal
is seen with difficulty, and the scarcely visible fumes rise rapidly
in the muffle (Mitchell). If the assays are long in uncovering
they may sometimes be started by dropping on them a little
charcoal powder wrapped in tissue paper, or still better by
placing a piece of charcoal near them. If one freezes before
completion it is restarted in the same way, or fresh lead is added
or the temperature is raised, but the results are not good.
The admission of the current of air to the muffle is carefully
regulated. Too much air may cause the bath of lead to spirt
and so occasion loss ; too little air delays the operation and
causes increased loss by volatilisation and absorption by the
cupel.
The bead thus obtained should be well rounded and bright,
loosely adherent to the cupel and slightly crystalline although
malleable. If it contains lead it is more globular and brittle and
its surface is very brilliant, while it does not adhere at all to the
bone-ash. If copper is present the bead adheres firmly to the
cupel, and in extreme cases its surface is blackened. Khodinm
and iridium occasion black patches at the bottom of the bead;
platinum makes the surface of the bead crystalline and rugose.
If the bead cracks on being flattened Van Riemsdijk recommends
a second cupellation with some more lead and 10 per cent, of
cupric chloride. In this way the metals with volatile chlorides
are eliminated.
Influence of Base Metals on Cupellation Iron. Its oxide is not
readily fusible with lead oxide the button is long in melting,
:
and a brown scoria is left on the cupel, which may entangle lead
globules and so contain gold. Care must be taken not to dis-
turb the cupel during the operation, as, if the molten lead is
moved so as to touch the scoria, part is entangled. The cupel is
stained red.
Zinc burns with a blue flame at first and volatilises, taking
gold with it
'
y
it forms a pale yellow scoria, having the same
effects as that consisting of oxide of iron. The button is slightly
crystalline.
Tin forms a grey scoria.
Copper carries gold into the cupel and is usually not wholly
removed from the bead.
Nickel and Cobalt are not so easily oxidised as copper they :
may form a dark green scoria and always stain the cupel green.
The button is crystalline.
Antimony does not interfere if less than 1 per cent, is present
FUSION OR CRUCIBLE PROCESS OP ASSAY. 421
If there is more, in volatilising it takes

(Rivofc). gold with it,
and forms antimoniate of lead, giving a pale yellow scoria and
it
causing the cupel to crack.

Arsenic has a similar effect. The last two metals mentioned
are removed by scorification if present in large quantities.
Manganese causes black stains and corrosion of the cupel, and
forms a dark scoria.
Chromium gives a brick-red stain and scoria on the cupel, and
aluminium a grey scoria both these metals delay the course of
;
cupellation.
Cadmium causes a black sooty ring to form inside the cupel
near its margin, and gives a brown scoria.
Tellurium causes loss by volatilisation, and also, in common
with some other metals, has the effect of causing subdivision of
the cupelled bead.
Bismuth has less prejudicial effects on cupellation than the
metals mentioned above, and may be substituted for lead. How-
ever, according to E. A. Smith,* the losses experienced in this
case, especially by absorption by the cupel, are much greater
than if lead is used.
The loss during cupellation by volatilisation was proved by
Making. It is never absent, but is insignificant with ores assay-
ing below 10 ozs. per ton, especially if highly volatile metals are
absent. The absorption by the cupel is more serious (see under
Bullion assaying, p. 448). Rivot states that gold is oxidised to
some extent at a red heat in the presence of antimonic oxide,,
litharge or cupric oxide, and that it is the oxidised part which
is absorbed by the cupel. This contention is as yet hardly
supported by sufficient proof.
Inquartation and Parting. The bead of silver ar.d gold
obtained by cupellation is squeezed between pliers, or flattened
by a hammer on a clean anvil, to loosen the bone-ash adhering to
its lower surface, and is then cleaned by a brush of wires or stiff
bristles. It is then weighed, the silver removed by solution in
nitric acid, and the weight of the residual gold taken, when the
difference between the two weighings represents the silver. If
the bead contains more than one-fourth its weight of gold,
enough silver is added to it to make an alloy of about this
composition, otherwise some of the silver will remain undis-
solved, being protected from the action of the acid by the outer
layers of gold. The amount of silver to be added is calculated
from the (approximately) known composition of the bead, or
guessed from its colour. A pale yellow bead always contains
more than 60 per cent, of gold, but a perfectly white bead may
"
not " part completely. The addition of the silver is effected in
the case of small beads bv fusion on charcoal by the blowpipe,
but large beads are, together with the silver, wrapped in as small
*
Joum. C/tem. Soc., vol. Ixv., p. 624 (1894).
a piece of lead foil as possible and cupelled. This is called

"
inquartation" The resulting bead is cleaned, flattened by the
hammer and, if it is large, by passing through the rolls, rolled up
into a cornet (vide Bullion assaying) if of convenient size, and
dropped into nitric acid. If the gold in the original bead did
not exc-jed one-quarter of the whole, inquartation is omitted.
The parting is effected in a test tube or a porcelain crucible if
the bead is small, in a "parting flask" if large. The strength of
acid used for the first boiling varies with the composition of the
alloy. Nitric acid of specific gravity 1'25 is used for the
inquarted alloy, and more dilute acids for poorer alloys. In
all cases the acids may be previously warmed with advantage, as
the gold does not in that case break up into such fine particles as
if cold acid is used. The acid must be of sufficient strength to
-attack the bead instantly, turning it black nnd giving off nitrous
fumes. The temperature is gradually raised to boiling point,
and maintained at it for two or three minutes, or until all apparent
action of the acid on the metal has ceased. The acid is then
poured off, the residue washed once with boiling water by decan-
tation, and if the bead is very large fresh acid added of sp. gr. 1 -32
(one part of strong nitric acid to one-half part of water). The
boiling is then continued for five to ten minutes longer, when
almost all the silver will have been dissolved. Avery small amount
of silver, weighing from '05 to '1 per cent, of the gold, obstinately
resists the action of the acid, and remains as a surcharge which
may be neglected in almost all ores. The gold may be left as a
single piece if it constitutes not less than one part in 10 to 20
of the bead if present in less than this proportion it breaks up,
:
and the finer particles may float and be lost in decantation.

Particles floating on the surface may be sunk by touching with
a glass rod or by a drop of water let fall on them. Continued
heating of the gold in acid makes it agglomerate to some extent,
so that it is easier to wash. The second acid is rarely necessary.
After decantation of the second acid and washing twice with
water, the water is drained off and the porcelain crucibles dried
at a gentle heat, and then gradually raised to redness. The
gold which was previously black and soft, being in a fine state of
division, now resumes its usual yellow colour, and hardens so
that it can be removed to the pan of a balance and weighed.
If a parting flask or test tube is used for the boiling, the parted
gold is transferred to an unglazed Wedgewood crucible. To
effect this the flask is filled with water, and the crucible placed
over its mouth. On inverting both together the gold falls into
the crucible, and the flask is removed in such a way as not to
disturb the precious metal. The water which lias filled the
crucible is then poured off, and the crucible heated as before.
The chief difficulty in parting by the last-named method is
encountered in transferring the gold from the glass vessel to the
FUSION OR CRUCIBLE PROCESS OF' ASSAY. 423
\\ edgewood crucible minute

; particles of gold may adhere to the
glass and are then left behind. The loss of these is the cause of
the low results occasionally observed when this method is used.
If the glazed porcelain crucibles are used for both
boiling and
annealing, no transference from one vessel to another of the gold
in its soft state is necessary, so that the source of error mentioned
above is avoided. On the other hand, the difficulty of boiling
the acid in a small porcelain crucible without sustaining loss by
projection may prevent the assay er from continuing the boiling
for a sufficient length of time, and some silver may thus be left
undissolved ; in consequence of this the results obtained are
sometimes too high by as much as 2 or 3 per cent, of the weight
of the gold. Such errors would, however, be inappreciable in
the assay of ores containing le.ss than 1 ounce of gold per ton, and
are the less serious for the reason that all the other sources of
&c.), tend to make the
error (unclean slugs, absorption by cupel,
result too low.
By whatever method the parting is performed, rery finely
divided gold may remain suspended in the liquids, and may thus
be lost in the course of decantation.
To prevent " bumping," one or two small pieces of capillary
glass tube or of clay pipe or other porous body are put into the
acid together with the alloy to be parted.
By the operation of parting, silver, palladium and some plati-
num are removed in solution, but the greater part of the
platinum, and all the rhodium, iridium, &c., remain with the
gold. If the presence of these metals is suspected they must be
looked for and removed by special methods (vide infra, p. 455).
Examination of Assay Materials. The fluxes are usually
free from the precious metztls, but the litharge, red-lead, and
granulated lead contain a small amount of silver and less gold
(see p. 30). These are estimated by fusion with charcoal (or in
the case of granulated lead by scorification) and cupellation.
Examination of the Cupel. When rich ores are assayed,
appreciable quantities of gold are carried into the cupel, especially
if much copper is present in the lead button. To assay the
cupel, all clean bone-ash is detached and thrown away, and the
remainder crushed so as to pass through an 80-mesh sieve, and
the charge made up as follows :
Cupel, .
Fluorspar, ....
. . . 100 parts.
75
Sand, . .
Soda carbonate,
Borax,
...
. . '. 75
100
50
Litharge,
Charcoal, ....
. . . . 50
4
The fusion is performed in an earthen crucible, and the resulting

button of lead cupelled. The slag of the original fusion may be
cleaned by adding it to the charge, when less fluxes will be

required.
ASSAY BY SCORIFICATIOX.
As stated above, this process is especially applicable to (a)

complex ores, (6) very rich ores, (c) ores which are mainly valuable
for their silver contents. The losses are small since the slag is
highly basic, consisting chiefly of litharge, while the bath of lead
is always poor, so that there is little loss of the precious metals
by volatilisation. Moreover the operations are easy to conduct,

and need not be varied much for different classes of ore. For
these reasons the process is generally preferred to the crucible
process in Germany and the United States ; if the ore is poor
several assays are made and the lead buttons scorified together.
The chief disadvantage of the process lies in the small quantity
of ore that can be treated, so that the presence of one or two
metallic particles of gold may cause the result to be erroneous.
Scorification is conducted in a muffle at a much higher tem-
perature than that required for cupellation. It must be high
enough to melt litharge when contaminated by silica and oxides
of copper, iron, manganese, &c. A temperature of 1,050 to
1,100 C. is usually enough. The charge is placed in a scarifier,
a shallow circular fire-clay dish 2 to 3 inches in diameter, which
is charged in by the tongs shown at B, Fig. 61. The charge
consists of about 50 grains of ore (or -^ A.T.), 500 to 1,000
grains of granulated lead and a few grains of borax glass. The
ore is mixed with half the granulated lead, the mixture put in the
scorifierand smoothed down, the rest of the lead spread over
evenly and the borax put on the top. The amount of lead to be
used varies with the nature of the ore. Rlcketts gives the fol-
lowing table* as a guide :
Character of Gangue.
ASSAY BY SCORIFICATION. 425
slag more liquid, but its quantity is kept as low as possible to-
prevent the slag from completely covering the bath of metal too-
soon. After charging-in, the door of the muffle is closed until
fusion takes place. As soon as the lead is melted the door is
opened, and a current of air allowed to pass over the bath of
metal. Some of the ore is now seen to be floating on the surface
of the lead, and is rapidly oxidised, partly by the air and partly
by the litharge which immediately begins to form. The sulphur,
arsenic, antimony, <fcc., are thus soon eliminated, while copper,
iron and other bases slag oft' with the borax, and the silica, and
other acids form fusible compounds with the litharge. Efferves-
cence and spirting may occur, especially if the scorifier
has not been well dried by warming before it is used. The slag
soon forms a ring completely incircling the bath of metal. As
oxidation of the lead proceeds, the litharge flows to the sides and
increases the quantity of slag until at length the ring closes
* 5
completely over the metal, leaving a "flat uniform surface

(Percy). This usually happens after from thirty to forty min-
utes. The slag should be " cleaned" before withdrawal. This
is done by placing 3 grains of charcoal powder wrapped in
tissue paper on the surface of the slag, with a pair of cupel
tongs, and closing the muffle door. A
number of globules of
lead are formed by the reduction of the litharge, and these, falling
through the slag, extract and carry down with them any gold
and silver which it contain, and concentrate them in
may still
the molten lead below. The fusion being quiet again, the
charge is poured into an iron mould by the scorifying tongs, and
the lead button cleaned from slag by a hammer. If the button
is too large to cupel at once it is re-scorified in the same dish,
fresh lead being added if it is not soft and malleable. Less loss
of the precious metals is incurred by scorifying lead than by
cupelling it, and consequently it is better to reduce any lead
button weighing more than 200 to 300 grains by re-scorification
before cupellation.
The slag from the scorifier should always be separately re-
treated in the case of very rich ores, but seldom contains much
gold.
The losses incurred in this process are chiefly due to improper
temperatures. If the muffle is too cold at first, gold is retained
by the slag. This initial low temperature is often indicated by
the occurrence of white patches of sulphate of lead on the surface
of the slag after pouring. If the slag is pasty, borax is added,
but the slag is then rich. It is better to begin again with less
ore or more lead. When extremely rich sulphides are assayed,
results are often obtained which do not agree well. Stetefeldt*
recommends in all such cases that the sulphides be attacked by
nitric acid and the residues dried and scorified.
*
Lixiviaiion of Silver Ores, New York, p. 106.
426 TUB METALLURGY OF GOLD.
The subsequent treatment of the lead button is as already

described. Tellurium ores often yield erroneous results, usually
ascribed to volatilisation of the gold. Eicketts asserts * that it is
not volatile in presence of tellurium (but see page 7), and that
no difficulty occurs in either crucible or scorincation method if
plenty of litharge or lead is used. For rich tellurides 60 to 100
parts of lead are required, a large scorifier being used. The
tellurium must be driven off before cupellation, as otherwise the
gold button will be subdivided into minute particles on the cupel.
Detection of Gold in Minerals. The detection of gold in
loose alluvial ground has been already described, p. 44. A
similar method may be employed in the case of auriferous quartz
after grinding it. If the concentrates obtained in either case
contain sulphides, these are collected, roasted or treated by nitric
acid and reground. The light particles of oxide of iron can now
be separated from any gold that may be present by washing.
" Colour"
may often be obtained thus when none could be seen
after the first concentration. The washing is made easier by
removing from the concentrates the magnetic oxides and iron
from the grinding tools by a magnet. All the finest particles of
gold are lost in the process of washing, and consequently many
auriferous ores cannot be made to " show colour."
The following method devised by Wm.
Skey, analyst to the
Geological Survey of New Zealand, is said by him to give good
results. The sample of ore is carefully roasted, then digested
with an equal volume of an alcoholic solution of iodine for a
length of time varying from twenty minutes to twelve hours, the
longer time being allowed if the ore is poor. A
piece of Swedish
filterpaper is then saturated with the clear supernatant liquid
and afterwards burned to an ash if gold is present in the ore
;
the ash is coloured purple, and the colouring matter can be

quickly removed by bromine. This method is said to show the
presence of as little as 2 dwts. gold per ton in certain ores, but
is not uniformly successful. It depends for its success on the
solubility of gold in a solution of iodine, and the alcohol must be
very pure to prevent reprecipitation of the gold. Iodine, how-
ever, is a slow and ineffective solvent for gold,f and the use of
bromine is preferable, since it fails less frequently and acts
more speedily than iodine. A mixture of 5 to 10 parts of
bromine with 100 of water may be used to 100 parts of ore. The
ore must be fine enough to pass an 80-mesh sieve and should be
reground after roasting. After leaving the mixture to stand for
some hours with occasional stirring, the liquid is filtered and the
excess of bromine evaporated from t>he clear solution, which may
then be tested by stannous chloride. Occasionally gold ores are
met with which give negative results with all these methods, but
yield small quantities of precious metal by crucible assay.
*
Notes on Assaying, p. 79. t Vide p. 27-
SPECIAL METHODS OF ASSAY. 427
Estimation of Gold in Dilute Solution. A. Carnot has

shown * that the rose colour produced in presence of arseniate of
iron is very sensitive, and can be used for colorimetric estimation
of small quantities of gold. To attain this end a neutral solution
of chloride of gold of known strength is prepared, and some
drops of solution of arsenic acid added slowly to it. Then
after a time two or three drops of dilute ferric chloride and some
hydrochloric acid are added. If the liquid is not acidulated, a
flocculent purple precipitate forms ; if it is too acid the reaction
fails, and only a faint blue colour is seen. The liquid is made
up to 100 c.c. with distilled water, a pinch of zinc dust added,
and the mixture shaken in a flask. A colour is produced,
varying from rose to purple, according to the amount of gold
present. The solution is clear can be filtered unchanged, and
kept without alteration for sonit^ time. If more than one milli-
gramme of gold is present (1 in 100,000), the colour becomes too
intense for small differences to be noticeable ;
if less than
one-tenth of this amount is present (1 in 1,000,000), the colour
becomes too faint. Between these proportions the amount of
gold present in a liquid can be determined by comparison with a
series of prepared coloured solutions.
For more dilute solutions, the test described on p. 26, depend-
ing on the use of stannous chloride, may be used. A number of
precipitates are prepared from solutions of gold containing
known amounts, and compared with that given by the solution
to be estimated. Suitable volumes of liquid from which the
test precipitates are obtained are as follows :
Strength of Solution.
Parts of Water present to each
part of Gold.
matte or regulus. All the gold is found concentrated in the

matte, which consists of sulphide of antimony, and forms about
one-tenth of the weight of the original ore. The matte is fused
with litharge and then dissolved in aqua regia, tartaric acid being
added, and the solution diluted by boiling water. The whole of
the silver separates out on cooling as AgCl, which is dried, fused
with lead, and cupelled. The gold is precipitated by sulphurous
acid or a little sodic hyposulphite, and is cupelled and weighed.
The method is said by its devisers to be accurate, Lut grey
antinionial-copper is difficult to obtain free from gold and silver.
2. Amalgamation Assay. This is useful only in determining
the amount of gold pi'esent in the ore capable of being extracted
by mercury. The sample of ore is pulverised fine enough to pass
through a 60 or 80-mesh sieve, made into a paste with a little
water, mercury added, and the whole ground in an iron mortar
with a pestle for from two to four hours, small additional
amounts of water or mercury being added from time to time
according to the appearance of the triturated mass. The con-
sistency of the mass must be such that the globules of mercury
do not sink in it but are broken up into very small particles.
A little sodium amalgam dissolved in the mercury prevents it
from flouring. The grinding is continued until the particles of
ore are all impalpably fine. A machine for the purpose called
the "arrastra" mortar has a large pear-shaped muller loosely
fitting the inside of the mortar, and capable of being revolved in
it by means of a handle. Complete amalgamation is performed
in this machine much more rapidly than in an ordinary mortar.
When the operator judges that amalgamation is complete,
enough water is added to reduce the mass to a thin pulp and
stirring is continued for a few minutes to collect the mercury
at the bottom. The contents of the mortar are then " washed
down " in a pan, the mercury collected and distilled, and the
residue, consisting of gold, silver and base metals, cupelled and
parted.
Mr. J. M. Merrick gives a similar method for the assay of
pyrites.* He roasts from 8 ozs. to 1 Ib. or more of the finely
pulverised pyrites, mixed with marble-dust to prevent caking,
and then amalgamates it in a stoneware pot or wooden bucket,
collects the mercury and distils it. It is stated that gold may
be detected in this way in samples showing none when assayed
by fusion. Many auriferous pyritic ores, however, yield but little
gold to mercury after roasting, so that the method is uncertain
as well as laborious. Such ores, if very poor, are better treated
either by chlorination or by Whitehead's method, which will be
described subsequently (see p. 430).
3. Assay by Chlorination. Plattner's method for the assay
of roasted pyrites consists in placing the mineral moistened with
*
Mitchell's Manual of Assaying, 1381, p. 694.
SPECIAL METHODS OF ASSAY. 429
water in a glass cylinder, 200-250 millimetres deep and 20-30

millimetres in diameter, and introducing; a current of chlorine
gas at the bottom. When the odour of chlorine is noticed above
the ore, a cover is put on, the stream of chlorine stopped and the
whole lelt for twenty-four hours, after which the reaction is
complete if chlorine is still in excess. Boiling water is now run
through the ore until all soluble salts have been washed out,
and the gold contained in the solution is precipitated by ferrous
sulphate, collected, cupelled and parted. According to Balling
this method fails if much silver is present, as the chloride of silver
formed encrusts the gold and protects it from the action of the
chlorine.* He recommends the addition of common salt to
dissolve the chloride of silver, but found that the telluride ores
of Nagyag yielded only 85 per cent, of their silver and 92 per
cent, of their gold when successively treated with chlorine and
sodium chloride.
A better method is to place the
completely roasted sample in an
ordinary soda-water bottle with enough water to make the whole
of the consistency of thin mud. The ore and water should
together occupy about two-thirds of the bottle. Bleaching
powder and a thin glass bulb filled with dilute sulphuric acid
are then added and the bottle securely closed. As cork is
attacked by chlorine, glass or vulcanite stoppers are better, and
the screw-stoppered bottles are most convenient ; if corks are
used they must be wired down. The bottle is then shaken so as
to break the sulphuric acid bulb and mix its contents with the
bleaching powder, when chlorine
is evolved. The bottle is now
left for several hours in a warm place, being shaken occasionally
by hand to mix its contents. At the end of a period of eight to
twelve hours the bottle is opened, and if excess of chlorine is
still present the liquid is separated from the ore and the latter
washed thoroughly by filtration or decantation. The liquid and
washings, whether clear or muddy, are warmed to expel free
chlorine and an excess of ferrous sulphate is then added to them.
The precipitate is collected, scorified with lead and cupelled.
In all cases it is better to keep the first liquid separate from the
washings, which should be concentrated by evaporation, since,
if
this is not done, the precipitate of gold may be too fine to settle
and will pass through filter-paper. A
better method of precipi-
tation is to boil the liquid with iron filings for a few minutes,
decant through wash the filings and dissolve them
filter-paper,
in dilute sulphuric acid when a residue of gold is obtained which
is easy to filter. Bromine may be used instead of the materials
will
geneiating chlorine. The quantities of chemicals required
be such as are sufficient to generate a volume of chlorine equal
to twice the capacity of the bottle used, or to make a solution of
2 per cent, of bromine in water. Only finely-divided gold is
extracted by this method.
*
Balling, V Art de VEssayeur, p. 433.
4, Whitehead's Method of Assay. A

useful method of
determining small quantities of gold and silver in base metals,
such as crude copper, or in mattes or base ores, is described by
Mr. Whitehead,* as follows Weigh out from 1 to 4 A.T. of
:
the auriferous material, place it in a beaker of 500 c.c. capacity,

and add gradually enough nitric acid to dissolve it completely ;
heat until red fumes cease to come off, dilute with water, and add
50 grammes of lead acetate, stir, and when dissolved add 1 c.c.
dilute sulphuric acid and allow the lead sulphate to settle.
Filter into a 1,000 c.c. flask and fill to the mark with distilled
water. The filter contains the gold which has been collected
and carried down by the sulphate of lead. The filter-paper and
precipitate are dried, the paper burned, and the ash and lead
sulphate scorified with test-lead. The button is cupelled and
the gold with any trace of silver it may contain is weighed and
then parted.
The solution is divided into two parts, and precipitated by
sodium bromide with constant stirring. The precipitates, which
consist of a mixture of the bromides of silver and lead, settle
quickly, and filter and wash well, only cold water being used.
When dry, the precipitate can easily be brushed from the paper,
and so the trouble of burning the latter is avoided. The
bromides are now mixed with three times their weight of
carbonate of soda and a little flour or charcoal, placed in a small
crucible, overed with borax, and melted down.
< The lead thus
obtained should be free from copper, and is easy to cupel. Dupli-
cate assays usually agree within two-tenths of an ounce of silver
per ton. The use of the lead acetate is to cause the precipitates
of gold and silver to settle quickly, and to enable them to be
filtered effectively. Sodium bromide is used instead of the
chloride, on account of the greater insolubility of the silver salt.
5. Assay of Pyrites. Schwartz's Method.^ 100 grammes of
pyrites is fused with 46'6 grammes of iron turnings under a
cover of salt. The protosulphide of iron formed is crushed and
dissolved in dilute sulphuric acid; the gold remains undissolved.
The liquid is filtered, the residue roasted, fused with borax and
cranulated lead, cupelled, and parted. It has been suggested
that the iron turnings are really unnecessary, and only serve to
increase the amount of regulus. Simple fusion with suitable
fluxes gives a rich regulus which contains all the gold.
Stapjfs Method.^ The pyrites is fused with sulphur and an
alkaline sulphate, and the alkaline aurosulphite thus formed is
dissolved in water. The gold is precipitated from the filtered
liquid by acidifying with sulphuric acid, and is cupelled and
parted.
*
Chem. News, 1892, vol. Ixvi., p. 19.
1" Dingler's Polyt. Journal^ vol. ccxviii.
Fremy's Encydopccdle Chimique, vol. iii., 16e cahier, p 202.
THE ASSAY OP GOLD BULLION. 431
6. Assay of Purple of Cassius. One part of purple of

Cassius is fused with three parts of carbonate of soda, cooled
and dissolved in water. The gold remains undissolved, and is
collected on a filter and cupelled after incineration.*
7. Assay of a Mint Sweep. The assay of sweepings of the
floors of jewellers' workshops, refineries and mints, and that of
pulverised crucibles, stirrers, tfcc., which have been used in

melting gold bullion, are included in this section. The larger
pieces of metal are removed from the sweepings by sieving and
washing. The sweep then still contains a large number of
metallic particles which will not pass through a fine sieve ; a
considerable percentage of sawdust, charcoal and graphite is also
present. The sample must be selected with great care and with
due regard to the fact that the lower parts of a heap are richer
than the top, owing to subsidence of metallic particles through
the mass. The amount of moisture present is first estimated,
and the sample dried and passed through an 80-mesh sieve, after
which it is roasted in a muffle and scorified with eight to twelve
times its weight of lead and a few grains of borax. The
metallics are scorified separately, or passed at once to the cupel.
The remaining operations do not differ from those required in
the case of a rich ore. Double parting is necessary, as both
silver and gold are usually present. From to ^ A.T. of the
unroasted material is a convenient weight to take for assay,,
but the metallics must be separated from a much larger quantity
and estimated by themselves.
CHAPTER XX.
THE ASSAY OF GOLD BULLION.
Assay of Gold Bullion. The assay of gold bullion, as described
in this chapter, has for its sole object the estimation of the
percentage of gold present in the alloy, all other constituents
being disregarded. In the first instance, the simple case of the
of only
assay of gold alloys containing appreciable quantities
copper and silver will be dealt with. Refined gold ingots
and
the alloys used for coinage, and for almost all jewellery, come
under this head. The effect of large quantities of other impurities
and the precautions thereby rendered necessary will be discussed
later.
The method universally employed is that of cupellation and
subsequent parting. The gold bullion is cupelled with silver
* xxxix.
Ann. de Pharmacie, vol.
and lead, by which the greater part of the base metals present
is removed as oxides dissolved in litharge, and an alloy of gold
and silver left on the cupel. This is " parted " by nitric acid,
which dissolves the silver and leaves the gold unattacked.
In the following pages the practice at the Royal Mint is
described, but the same description would apply, with very
slight alterations, to the methods used at other mints and assay
offices.
The "parting assay" was first mentioned in a decree of
King Philippe of Valois, published in the year 1343.* The
methods of procedure in the 17th century have been briefly de-
scribed by Savotf and by J. Reynolds,! and more fully in the
Compleat Chymist. In 1666 Pepys saw the parting assay being
practised at the Mint in the Tower of London, and from his
description it is clear that the method then employed bears a sur-
prisingly strong resemblance to that of the present day. In 1829
.a
Royal Commission was appointed in France to examine into
all questions relating to the methods of assaying gold and silver.
The results of the labours of that Commission are to be found in
great part in the Appendix to the Report of the Select Committee
on the Royal Mint, 1837. The Committee arrived at the con-
clusion that the method adopted for assaying gold often over-
stated the amount of precious metal by 1 part per 1,000. The
Mint Conference held in Vienna in 1857,|| resulted in the almost
universal adoption of a more uniform method of manipulation.
The degree of accuracy now attained in most assay offices
reduces the probable error in the report of an assay to (H per
1,000, but, to prevent the error from rising above this amount,
all weighings must be correct to 0*05 per 1,000, which is not
usually the case in ordinary bullion assays (see p. 451).

The system may be conveniently regarded as comprising six
distinct operations, viz. :
1. Selection of the sample.

2. Preparation of the assay piece for cupellation.
3. Cupellation.
4. Preparation of the assay piece for parting.
5. Parting and annealing the cornets.
6. The final weighing and reporting.
1. Selection of the Sample. At the Royal Mint the con-
tents of the pots are poured into moulds, each pot yielding about
nine bars, and a piece is cut from the middle of the end of each
of the first and last bars formed. Experience proves that these
samples are usually representative of the whole mass viz., 1,200
*
First Report of theRoyal Mint, 1870, p. 103.
t Discours sur Medalles Antiques, Paris, 1627, p. 72.
les
A New Touchstone for Gold and Silver Wares, London, 1679, p. 362.
Report on the Royal Mint, 1837, Appendix B, p. 123.
J| Kunst-und Gewerbeblatt Baiern, 1857, p. 151.
ozs. In the case of the ingots, weighing 400 ozs., which are
received from the tank of England for coinage, a
single sample
is cut 1'rom the middle of one of the lower
edges of the ingot.
When the sample is not representative of the whole ingo? (a
somewhat rare event, judging from results), other base metals
besides copper are probably present.* Amore certain way of
finding the composition of an ingot is to melt it under charcoal,
stir well and dip out samples from the top and bottom of the
mass, with an iron ladle. The metal thus obtained is granu-
lated by pouring slowly in a thin stream into a porcelain-lined
kettle filled with warm water, which must be constantly stirred.
The dip-sample may also be taken by a little charcoal crucible
fastened to an iron rod, a lid of charcoal being put on directly it
emerges from the molten metal. Dip-samples should always be
taken when very impure and base ingots are to be assayed. In
place of taking a cutting from the edge of an ingot a drilling
machine is used in the Paris Mint and in many other offices by ;
this instrument portions are taken for assay from any part of the
ingot near its surface.
2. Preparation of the Assay Piece for Cupellation. If the
" flatted "
sample to be assayed is a single piece cut from a bar it is
on a clean anvil by a hammer with a rounded face, weighing
about 11 Ibs. A convenient thickness for the sample is about
^2 inch. The piece is then wrapped in paper which is marked dis-
tinctively, and an assistant then performs the operation of "bring-
ing to weight," or obtaining a piece approximately ^ gramme in
weight. This he does by cutting with shears and filing.
The use of the shears can only be learned by practice, but the
following remarks may be of use to a beginner. The metal to be
cut is held firmly between the fore-finger and thumb of the left
hand. Care must be taken, and considerable force exercised by
the fingers if necessary, to keep the plane of the piece of metal
to be cut perpendicular to the cutting faces of the shears, other-
wise damage is done to the latter. Only clean portions of metal
must be used.
The weights used in gold assaying are the J gramme, which is
"
stamped 1000," and decimal subsidiary weights stamped 900,
800, &c., and 90, 80, 70 and so on down to 0-5. These stamped
numbers denote the number of J milligrammes ("milliemes")
contained in the weight. Ordinary weights in the gramme
system may accordingly be used, each milligram me corresponding
to two milliemes in the assay system. The report finally made
gives the number of parts (in milliemes and tenths) of pure gold
in 1 ,000 parts of the alloy.
Since the assay is reported to -nj-J^ny part, it is evident that
the balance used must clearly indicate a difference in weight
of 0-1 per 1,000 or -05 milligramme. It is convenient to havo
*
See under Liquation of Gold Alloys, p. 17.
28
the balance adjusted so that one subdivision on the ivory scale

traversed by the pointer corresponds to this quantity. In some
" "
offices, however, preparing balances less delicate than this
are used. The weight of alloy must not differ from the 1,000
by more than four subdivisions of the scale. The weighed
portion is folded in a small square of paper and tucked inside
the paper containing the rest, until it can be checked by the
assayer. Weighing becomes a very rapid operation with prac-
tice, a skilful workman being able to prepare forty to fifty pieces
in an hour.
The assayer now checks the weighed piece on a more delicate
balance. Those used at the Royal Mint have beams only 6
inches long, weighing 156 grains, and requiring only ten seconds
for a complete double oscillation. Suspension of the pans was
formerly effected by means of two hard steel points instead of
a knife-edge; in the year 1892 light agate knife-edges were
introduced. Point suspension gave greater lightness, but the
wear of the points was rapid, and slight blunting soon impaired
the accuracy of the balance, necessitating repairs. These
balances are released by a lever working in a vertical direction.
The excess or deficiency in weight is recorded on the assay
paper in tenths of a millieme, the weight of the piece being
adjusted by the assayer if it shows a deviation greater than four
of these tenths. The piece is wrapped up in lead foil, which
is very pure, and weighs 4 grammes or eight times as much as
the assay piece. The lead is twisted into a form resembling
that of a conical sugar bag, and the gold together with the silver
necessary for parting is dropped in and wrapped up. The lead
packets are put in order in the numbered compartments of the
wooden tray shown at c, Fig. 67, p. 444, their position being
noted on the assay paper. The corners of the packets are
squeezed down so as to fit the cupels by pliers specially designed
for the purpose, and the assays are then ready to be charged
into the furnace.
The silver added to the assays is in the form of square pieces
cut off by means of gauged shears from a strip rolled to uniform
thickness ; the squares weigh about 1^ grammes each or two
and a-half times the weight of the alloy. If silver is present
in the alloy it must be allowed for. In the case of fine gold
(used as proofs) or of any sample of gold of high standard, an
alloy of silver and copper, 966-6 fine, is used instead of fine
silver. The use of the copper is to make these assay pieces of
about the same composition as the standard alloy (9 16 '6 fine)
worked with them. Fine silver is added to all alloys containing
more than 50 milliemes of copper. All the silver added for
purposes of parting must of course be free from gold.
It was formerly considered necessary for the metals to be pre-
sent in the proportion of 1 of gold to 3 of silver, but, as early as
the year 1627, Savot relates th-at the proportion of 1 to 2 was

used, and strong acid employed in the boiling, " quand on veut
faire quelque essay curieux et exact."* Both Cbaudet and
Kandelhardt recommend the proportion of 1 to on the
2J,
ground that less silver is then retained by the cornel than in
any other case. If much more than three parts of silver are
present the gold breaks up in the acid. Pettenkoferf found
that the proportion of 1 to 1'75 could be employed if the
assays
were boiled in concentrated nitric acid for some time.
The amount of lead used varies with the proportion of base
metals present. The following table shows the proportions
recommended by D'Arcet,| Cumenge & Fuchs, and Kandel-
hardt respectively
||
:
Gold in 1,000
parts : the Alloying
Metal being Copper.
but to obtain a well-formed, clean and bright button suitable for

parting in. each case silver is added equal to two and a-half
:
times the weight of gold present.
Amount of Gold per

1,000 parts.
THE ASSAY OF GOLD BULLION. 437
Mint is shown in elevation in Fig. 62. It consists of an outer

casing of wrought-iron plates about & inch thick, united by angle
iron. This casing is connected with a chimney 60 feet high by
means of a wrought-iron hood (a) and flue which is provided
with a damper (6). The lining consists of Stourbridge fire-bricks,
and there are five openings in the front of the furnace. Fuel
is introduced through the uppermost one of these under the
nap, shown partly open, the position of which can be varied by

means of two saw-edged inclined planes. The opening (d) com-
municates directly with the muffle (e) and may be closed by
sliding iron doors, or by the firebrick front (g) in conjunction
with a sliding plate, not shown in the figure, covering the upper
part of the opening (d). There are also two small openings,
closed by sliding doors, at the side of (d), for introducing and
withdrawing the fire-bars ; these are seldom used to regulate
the draught, for which latter purpose the doors (k) of the ash
pit (/) are usually employed. The muffle is formed of fire-clay,
and is pierced with four holes at the upper part of the back, by
means of which a draught is established ; these slope from within
downwards, in order to prevent particles of fuel from finding
their way into the muffle. The muffle is shown in front elevation
in Fig. 63.
There are eleven fire-bars, but only the three outer ones on
each side are covered with fuel. On the other five bars rest the
cast-iron girder-plate (Fig. 64), which is flat on the upper surface,
but is strengthened with ribs on the under-surface in order to
prevent buckling. Above the girder-plate the muffle is fixed in
an inclined position, so that all the cupels may be readily visible
from the front. The muffle rests on a bed placed on the girder-
over with fire-
plate consisting of pieces of fire-brick plastered
clay. It is convenient to cover the top of the muffle with a thick
luting, composed of fire-clay and
a little graphite in order to
check radiation, and to protect the muffle from the coke let fall
on it from above, and a fire-brick (m) is placed behind to prevent
excessive heating of the back of the muffle. Charcoal, anthracite
and coke might all be used in such a furnace, which would be
built narrower if either of the first two were employed, but on
account of the expense they are now rarely employed, and in the
the size
Royal Mint coke is always"used. It is broken to about
of hens' eggs and carefully screened.
in use at the
Assay furnaces, heated by gas, have long been
Berlin and Utrecht mints, at the Sheffield assay office and at
many other places. Their use is becoming more general, especi-
ally in America.
Furnaces to burn soft coal are used in Western America
where good coke is very expensive. They differ little from coke
furnaces in construction, but have less space between the muffle
and the side walls. The flame of the coal is chiefly instru-
1
43S THE METALLURGY OF GOLD.
mental in heating the muffle, a comparatively thin bed of fuel

being employed.
The cupels in use at the lioyal Mint are in sets of four, the
outer margin being square, as shown in Fig. 65. This form was
LEVEL OF TK BONE
SECTION ON LIME X, Y. THE CLTIUJL.
Fig. 65.
suggested by Mr. Joseph Groves of the Royal Mint, in 1872,*

and is found to facilitate charging-in and withdrawal. It will
be observed that the cupel mould A
B is in three pieces, which
are separated to take out the finished cupel. The furnace tools
in use are shown in Fig. 66, where a represents the cupel tongs
7,1'"
above the top of the muffle. The coke is at a full red heat
throughout before the cupellation is begun and no fresh fuel is
added during the operation. The muffle is usually ready for
work in about an hour from the lighting of the furnace. The
charcoal is then removed from the muffle and all dust and ashes
blown out by a pair of hand bellows.
The packets of lead containing the silver and gold are now
transferred to the cupels, arranged in rows corresponding to
those on the tray, the muffle being at a bright orange-red heat.
This charging-in is performed by the tongs, b, in Fig. 66, over
the top of the plumbago front (g, Fig. 62), the object of which is
to render the temperature of the muffle as equable as possible.
The cupels at the back are filled first, and by the time six assay
pieces have "been introduced, the first one should be melted and
" uncovered
by the removal of the black crust which forms at
first. At the Royal Mint the full fire of 72 assays takes about
five minutes to charge-in. The slider is now put in place over
the upper part of the opening (d), and the air supplied to the
muffle increased by closing the damper. While the muffle
door is open the indraught is diminished by opening this
damper so as to prevent the cooling action of the current of air
from proceeding too rapidly. The air supplied to the muffle
and furnace may be entirely regulated by this damper. The
furnace operation should be performed rapidly and in such a
way that all the cupellations may be completed at as nearly as
possible the same time.
Distinct stages may be noted in the action which now takes
place on the cupel. Almost immediately the surface of the
molten metal becomes covered with greasy-looking drops of
litharge, which are rapidly absorbed by the porous cupel
and
replaced by others. They pass over the surface at first slowly,
but as the operation continues move with greater rapidity. In
from eight to fifteen minutes the metal suddenly becomes
imiformly dull and glowing except for iridescent bands, pro-
duced by extremely thin films of fluid litharge, which are seen to
pass over it. On the disappearance of these bands a bright
liquid globule of a greenish tint is left, but
the cupels are not
withdrawn from the furnace until the expiration of another
fifteen to twenty minutes so that the last traces of lead may
be oxidised and absorbed. The completion of cupellation takes
place first in the front rows and proceeds regularly backwards.
The cupels are withdrawn from the furnace while the assay
in a few seconds.
pieces are still fluid, and "flashing" ensues
" is most marked in the buttons, in which but
Flashing" purer
occur in
little copper or lead remains. Slight effervescence may
these cases, but the buttons are never sufficiently freed from
"
base metals for " sprouting to take place.
Some assayers do not remove the assays from the furnace
until "brightening" or "flashing" has taken place and the

buttons have become solid. The muffle door must be closed
and the temperature reduced to effect this. Removal of liquid
buttons saves much time and is always performed at the Royal
Mint, but has two dangers. A "
cupel may be upset or spitting
"
may take place. Under the conditions given above, however, a

trace of copper is always retained by the cupelled button and all
fear of spitting thus removed.
The " "
flashing of gold assays was shown by Van Riemsdijk
*
to be due to solidification after superfusion. The temper-
ature of the fluid metal falls until a certain point is reached
when it solidifies and the sudden
disengagement of the latent
heat of fusion reheats the cooling globule to its true melting
point viz., 950 and a peculiarly intense light is emitted which
rapidly fades as the temperature again falls. Asudden jar at
any moment causes the flashing to occur instantly. If the alloy
contains a minute quantity of iridium, rhodium, ruthenium,
osmium or osmium-iridium (the metals of the platinum group
which are non-malleable, refractory to heat and resist the action
of acids), the tendency of the cupelled metal to preserve its liquid
state below the melting point, and therefore to flash during the
final solidification is entirely prevented. These researches point
to a simple method for detecting the presence of metals of the
platinum group (except platinum and palladium) in ingots of
commercial gold j for, if assays made on them solidify directly
they are withdrawn from the muffle without flashing, it will be
safe to conclude that the metal contains some of the above-
named impurities.!
The buttons (which are of the form represented at a, Fig. 68,
p. 444) are removed from the cupels, after cooling, by a pair of
sharp-nosed pliers, cleaned by means of a stiff brush or by immer-
sion in warm dilute hydrochloric acid, and are
placed in the
compartments corresponding to their cupels in the tray (d, Fig.
67). If the bone-ash is not completely removed from their lower
surface it is of little moment since bone-ash is
readily dissolved
by nitric acid on parting. The surface of the cupels must be
carefully examined for minute beads of metal due to spirting of
the lead bath which sometimes happens if there is too
strong a
If any such beads are found in a
draught. cupel the fact is
noted and the assay repeated,
If traces of lead remain in the button it is more
globular,
separates more easily from the bone-ash of the cupel and has a
brilliant steely surface. The effect of the presence of other
metals is discussed on pp. 453-457.
Temperature. The exact temperature suitable for cupellation
can only be ascertained by practice, and the varying light of
*
Chemical Newts, vol. xli., 1879, p. 126.
tProf. Roberts-Austen in Tenth Annual Mint Report, 1879, p. 43.
the day may occasion error in judging the degree of heat. The
remarks on temperature in the cupellation of buttons from ores
(p. 419) apply
here. Care should be taken to ensure that the
"
heat is so high before " charging-in that the chilling which
necessarily takes place during this operation shall not cool the
muffle below the requisite temperature. It is of more conse-
quence that the muffle should be uniformly hot throughout than
that any absolute degree should be attained, as the checks used
eliminate uniform errors due to high temperature.
The measurement of the temperature of the assay muffle has
not been frequently attempted. In a paper read before the Royal
Society (Phil. Trans., 1828, 79-96), Mr. James Prinsep, Assay
Master of the Mint at Benares, gives an account of experiments on
the subject by observing differences in the behaviour of a number
of silver-gold and gold-platinum alloys when heated. He "made
trials in different parts of the same (muffle) furnace. The
" is
disparity of heat," he remarks, greater than might be
supposed, and where, as in assaying the precious metals, so-
much depends on the temperature at which the operation is
performed, it would be useful to know every
difference in this
respect obtaining in different countries, and its effect upon

the
or standard of bullion." His results were as follows :
quality
Maximum alloy
melted.
Muffle of an assay furnace : front . . .
pure silver.
: middle (average)
: behind (average)
The temperatures at which these alloys melt are

Silver, . . . 945 (Violle).
Silver, 70 \ ofin
Gold, 30 )'
Silver, 50 }
' '
QQI \o
Gold, 50 /'
No further experiments were made until the year 1892, when

the temperature of the muffle in the furnace at the Royal Mint,
described on p. 436, was measured by the author by means of
the Le Chatelier pyrometer.* The arrangements were as
follows :The wires, sheathed in clay pipe-stems, were fixed in
a porcelain tube, the junction being protected by plates of mica.
This tube was introduced into the furnace by sliding through
another of wider bore fixed in the brick door of the muffle. The
inner end of the narrow tube was plugged by clay and coated
over with a fusible mixture of sand, pipeclay and felspar. The
from the furnace, and
galvanometer was shielded from radiation
remained at about a constant temperature, but the zero was
determined in order to detect any changes. In
frequently
* 1893, p. 707.
Journ. Chem. Soc., vol. Ixiii.,
calibrating the wire, the effects of heating various lengths of it

were noted, and allowance made for the alterations in the
deflections of the needle due to this cause. The junction was
kept stationary at each point in the muffle until its temperature
had become constant, as indicated by the cessation of movement
of the spot of light on the scale.
The muffle is 15 inches long and 6^ inches wide, and its full
charge consists of seventy-two assays arranged in twelve rows,
the distance between the front and back rows being 10^ inches.
The position of the junction was about 1 inch above the cupels
in all cases. The error in observation in taking the readings
was probably not more than O'l mm. on the scale, which was
equivalent to 0-32.
c.
Place.
the flattened buttons are passed in succession through a pair

of jeweller's rolls. The rolls are adjusted so that one passage
v Scale,
E=D
,. ,|l
lA in. =1 ft.
ecu
Fig. 67.
through them reduces the buttons to the required thickness viz.,

about that of an ordinary visiting card. The " fillets " (c, Fig. 68)
thus obtained should all be
x \ of uniform size and thickness,
-*
/
/
\
\
<2
m
with " wire edges," as ragged
edges expose them to loss
during the boiling. After
being rolled they are replaced
in the tray, f, and annealed
at a dull red heat. The object
of the first annealing is to
soften the buttons and facili-
tate their passage through the
rolls, while that of the second
is to enable the fillets to be
rolled into " cornets " or
spirals, d, between the finger

and thumb. Care is taken in
the latter operation to make
Fi". 68. Scale, full-size.
that which was formerly the
lower side of the button form
the outer face of the coil, for a reason given below (p. 457). This
face is easily recognised, as it is less brilliant than the other and
is marked with undetached flattened particles of bone-ash. In
the Mints of India, Scandinavia and some other countries, the
fillets are rolled tightly round a glass rod and then slipped oft*
the end. This operation occupies a longer time than the coiling
between finger and thumb, and leaves the cornet less open to
rapid attack by the acid than if the spiral form is used.
5. Parting. This was formerly effected by boiling with
nitric acid in glass "parting flasks." Platinum boiling trays
save time, and are now used whenever possible. The silver is
dissolved by the acid which must be free from chlorine in any
form, sulphuric and sulphurous acids, or sulphide from which
sulphuric acid may be formed. A small quantity of silver is
kept in solution, in the stock of acid, so that chlorine, if present,
would instantly be detected.
When the flasks are used, 2 ozs. of nitric acid of specific
gravity 1-2 are put into each flask and raised to boiling point.
A cornet is then introduced and boiling continued for 15 or 20
minutes (i.e., for about 10 minutes after nitrous fumes cease to
be given off). Hot distilled water is then added, the solution of
nitrate of silver decanted off, and the flask washed by filling
with hot water, and decanting. Two ounces of hot nitric
acid of specific gravity 1-3 is now poured in and the boiling
continued for fifteen or twenty minutes, a parched pea or
piece of charcoal being added to prevent bumping ; after
which decantation and washing is twice performed. Another
boiling with acid of specific gravity 1-3 is recommended by
Chaudet with the object of dissolving out the last traces
of silver and leaving the gold quite pure. This practice
has been adopted by many assayers, but is useless and causes
loss of gold. If any small particles of gold have become detached
from the cornet, time must be allowed for them to settle before
each decantation. After the last decantation the flask is filled
with hot water, the top covered by a small porous crucible, and
the whole is carefully inverted ; the pure gold, which is of a
dark brown colour and exceedingly fragile, falls through the
liquid and rests in the crucible, the water which enters with it
being afterwards poured off. The crucible is dried and then
annealed at a red heat over gas or in the muffle, when the gold
shrinks greatly, though still preserving its shape, hardens and
regains its ordinary pale yellow colour. It can now be weighed.
When a platinum boiler is used the cornets are put on
platinum pins, as at the Sydney Mint, or more usually into
platinum cups, one of which is shown in Fig. 69. These cups
are supported in a platinum tray (which holds 144
cups at the Royal Mint) and the whole lowered by
a platinum hook into a platinum vessel containing
about 80 ozs. of hot nitric acid. Great attention
must be paid to the temperature of the acid. At
the Royal Mint acid of specific gravity 1 -26 (in which, ill
however, a small quantity of silver is already dis-
solved) is used, and the temperature at the moment
of introduction of the tray is 90 C. If the acid is
colder than this the cornets tend to break up, some
pieces being usually detached in the boiling of 144
cornets, even if the temperature is only 1 or 2 lower
than the correct point. If the acid is much hotter than 90*
446 TIIE METALLURGY OF GOLD
(93 or more) the action is so energetic that the vessel may

boil over. If the temperature of the acid is as low as 85,
the cornets are not immediately attacked ; they turn a coppery
hue and then bubbles of gas form slowly on them. In such
cases (where the action does not begin within 1 or 2 seconds
after entrance into the acid) the only chance of saving the cornets
is to withdraw them instantly and raise the temperature further.
Boiling is kept up vigorously for twenty to thirty minutes,

and the tray is then withdrawn, drained, washed by dipping
vertically in and out of a vessel of hot distilled water, drained
again, and placed in a second platinum boiler filled with boiling
nitric acid of specific gravity 1-32 (also containing a little nitrate
of silver). In this the cornets remain for a period of thirty to
forty minutes, when they are drained and washed as before, and
are then ready for annealing.
The platinum tray may be dried on a hot plate, or may be
introduced into the muffle at once while still wet. The iron
peel, shown in Fig. 66, is used to support the tray in the muffle.
Rotating peels tire sometimes used so that the temperature of
annealing can be equalised by rotating the tray without with-
drawing it from the furnace. A thin platinum plate may be
placed between the tray and the peel to prevent scales of oxide
of iron from entering the cups. The floor of the muffle should be
scraped clean of bone-ash to prevent projection of the latter into
the cups owing to the fall of drops of water from the wet tray when
it is first introduced, or a clean tile may be put in the muffle to
support the tray. The cornets are not injured by the slightly
explosive evaporation of small quantities of water contained in
their porous substance. Annealing should be conducted at as
high a temperature as possible, consistent with the safety of the
cornets. They fuse at 1,045 G. (having the same melting point
as pure gold), at which temperature the muffle appears orange-
red. If annealed at a low temperature, the cornets are rough in
texture, dull and fragile, being crushed easily between the finger
and thumb. In this condition they adhere to the platinum, and,
in detaching them, fragments are often left sticking fast inside
the cups. If annealed properly the cornets are smooth, lustrous
and hard, showing signs of incipient fusion under a magnifying
glass, and only yielding to considerable force exercised by the
finger and thumb. Under these conditions they can always be
detached from the platinum entire. By being annealed the
cornets, which after boiling are very soft and fragile, and dark-
brown in colour, shrink and harden, and regain the ordinary
yellow colour of gold (e, Fig. 68).
Relative advantages of Parting in Flasks and in Platinmn
Boilers. The use of platinum trays and boilers effects a great
saving of time in decanting and washing, as one operation takes
the place of as many as 144, If the standard of an alloy is
unknown, so that it is not certain that the cornet will remain

entire in the acid, it must, of course, be boiled separately, as, if
one cornet in the tray breaks up, fragments may adhere to a,
number of others. The manipulation of the platinum tray is
easier than that of parting flasks, and, in addition, the treatment
of.he cornets is more uniform, so that the correction afforded
by
t* use of checks is more trustworthy. On the other hand, if
platinum or palladium is present in an individual cornet, it
imparts a straw-yellow or orange colour to the acid ; but where
a number of cornets are boiled together, it is obviously impossible
to say from which the colour is derived, so that less informa-
tion on the subject is obtained. A saving of acid is effected by
the platinum vessels, 144 assays being boiled in 80 ozs. of acid
at the Royal Mint, while 288 ozs. would be required if parting
flasks were used.
6. Weighing the Cornets. The final weighing is performed
on a bnlance capable of indicating differences of 0'05 per 1,000,,
or one-fortieth of a milligramme, with a weight of ^ gramme
in each pan. The "checks" or "proofs" (vide infra) are weighed
first, and their mean excess or deficiency in weight applied as a
correction to all the cornets worked with them. The " weighing -
in" correction (p. 434) is also allowed for, and the report is at
once indicated by the marks on the weights without further
calculation.
Surcharge. The gold cornet does not actually contain the
whole of the gold present in the original alloy and nothing else.
Gold is lost by (a) volatilisation (b) absorption by the cupel \
;
(c) solution
in the acid. On the other hand, the cornet always
retains (1) some silver; (2) occluded gases. The algebraical sum
of these losses and gains is called the "surcharge," since the
cornet usually weighs more than the gold originally present in
the assay piece ; if the reverse is the case, the work is regarded
as less accurate by some assayers. The various losses and gains,
are discussed in detail below.
Losses of Gold in Bullion Assaying. The losses of gold.
can only be incurred in three ways, namely :
1. Absorption by the cupel.

2. Volatilisation in the muffle.
3. Solution in the acid.
The two latter clauses of loss were discussed by the late Mr. G.
H. Makins in a paper " On certain Sources of Loss of Precious
Metals in some Operations of Assaying," which was read before
the Chemical Society and published in their Journal (1860, 13,
77). He found gold and silver in the proportion of about 1 to-
9 in the dust taken from a flue used only in gold and silver
ascertain the percentage loss
cupellation, but did not attempt to
of gold by volatilisation. He also showed that large amounts of
the course of assaying, and
gold were dissolved by nitric acid in
attributed the dissolution of gold to the presence of nitrous acid,

but supposed that it would not be dissolved in the weaker acid,
where nitrous acid was formed in larger quantities, owing to the
protective action exercised by undissolved silver, which formed
the positive element in the gold-silver couple. The fact that
gold is actually dissolved in nitric acid, presumably as oxide, and
remains dissolved even after dilution with water, was proved by
Mr. A. H. Allen (Chem. News, vol. xxv., p. 85).
According to Bruno Kcrl,* an alloy which requires from four
to eight times its weight of lead for cupellation, sustains a loss
which may be as high as 1 millieme of gold if sixteen to
;
thirty-two parts of lead are required the loss reaches 2 milliemes.

According to Rossler the losses are heavier, varying from 1 to
-3 milliemes of
gold when four to five parts of lead are required
for cupellation. As the result of a number of experiments con-
ducted by the author at the Royal Mint,f it was found that the
losses there in the assay of standard gold (916-6 fine) vary from
0-4 to 0-8 per 1,000, of which about 82 per cent, is absorbed by
the cupel, 8 per cent, is dissolved in the acid, and the remaining
10 per cent., which is unaccounted for, is probably volatilised.
These ratios, however, vary considerably, as a hot fire increases
the loss by absorption and volatilisation, while by prolonged
boiling in acid, and especially by annealing the fillets at a high
temperature, the amount of gold dissolved in the acids is
increased. An increase in the percentage of copper in the assay
piece is accompanied by an increase in the loss under all three
heads. This increase of loss has long been known. The follow-
ing table of the results of experiments on synthetic alloys of gold
and copper is given by Pelouze and Fremy| :
Actual Standard Difference

of the Alloy. Surcharge.
yoo + 025
800 + 0-50
700 o-oo
600 000
500 - 0-50
400 - 0-50
300 - 0-50
200 - 0-50
100 - 050
The high surcharge assigned in this table to the 800-alloy is

difficult tounderstand. The usual experience is that an increase
of copper in an alloy causes greater loss of gold. The following
table gives the relative surcharges obtained in the parting assay
of gold of different standards it is compiled from a number of
;
results obtained at the Royal Mint by the author :
*
Metallurfjische Probirkunst., 1880.
\-Joiirn. Chem. i$bc.,lxiii , 1893, p. 710.
J TraitS de Chimie., Sine. Edition, vol. iii., p. 1230
THE ASSAY OF GOLD BULLION.
449
Standard of Alloy.
The amount of silver retained by the cornets varies with the

proportion of silver to gold in the fillet. Thus in an experiment
conducted by the author* fillets in which the relative proportions
were 2*47 1 and 2-4:1 yielded cornets containing respectively
:
0-6 and 0-7 per 1,000 of silver. This is a common source of

error, as the variation of 0-1 would appear in the result, while
such small differences in the proportion of parting silver added
to the gold are usually disregarded. Even if the initial weight
of silver is constant, the weight in the fillets may vary from
unequal losses in the furnace. This loss was found to be about
1 per cent, when standard gold is being assayed, 1'6 per cent, in
the case of gold 800 fine, and 3 per cent, if it is 666 fine.
Occluded Gases. Graham proved f that cornets retained twice
their volume of gases (mainly carbonic monoxide) in occlu-
sion after annealing. This amounts to two parts by weight in
10,000, and is reckoned as silver in the preceding paragraph.
According to Varrentrapp, the gas retained varies with the
temperature at which annealing takes place.
Checks or Proofs. Since the losses and gains detailed above
are dependent on so many conditions, it is always necessary to
subject check-pieces of known composition to the same operations
as the alloys under examination. The use of checks in the Royal
Mint was prescribed by law as early as the 14th century. |
Standard trial plates (916-6 fine) were made and used for this
purpose. Since, however, it is impossible to guarantee that a
mass of alloyed metal shall have absolutely the same composition
throughout, it is better to use pieces of pure gold, a corresponding
amount of pure copper being added in order to make the assays
absolutely comparable. The correction to be applied to a gold
assay is given by the following formula :
Let 1,000 represent the weight of the alloy originally taken.

p the weight of the piece of gold finally obtained.
x the actual amount of gold in the alloy expressed in thousandths.
a the weight of pure gold taken as a check, approximately equal to x.
b the "surcharge," i.e., the loss or gain in weight experienced by a
during the process of assay expressed in thousandths.
Then x f** p or the corresponding loss or gain experienced by x is equal
to *^
Example
;
so thatj, =x
Let p
~ (i) or x
= 9167 thousandths.
=
^ (ii).
a = lOOO'O
b = 0*3 ,, gain in weight.
Then 6 is a positive change, and therefore has the + sign.
916-7 x 1000
Hence - * =
1000 + 0-3
= 916-424
*T. K. Rose, Accuracy of Bullion Assay, Journ. Chem. Soc. (1893),
p. 706.
tPAtY. Trans. Roy. Soc., 1866, p. 433.
Mint Report for 1873, p. 38.
This result would be reported as 9 16 '4, and, therefore, the
following rule
is approximately correct ; if there is a
gain in weight by the checks, the
amount is deducted from the weight of each of the other cornets if a loss, ;
it is added. A piece of pure gold weighing 1000 may thus be taken, without
appreciable error, as a check on assays of all alloys over 900 fine, provided
that the "surcharge" docs not exceed + 0'5. This error
may be reduced
to zero by taking as a check- piece an amount of
gold equal in weight to
that present in the alloy under examination, an
unnecessarily laborious
process when assay pieces of various degrees of fineness are present.
In the above calculation it was assumed that absolutely pure

gold was used for proofs. It is doubtful whether it has ever
been obtained. The method of preparation of proof gold in use
at the Royal Mint is that given at p. 10. Since the assay of
an alloy only gives the relative fineness of proof gold and the
alloy, it follows that, if the proof gold is not quite pure, the
amount found in the alloy will be in excess of the truth. If a
sample of proof gold is less pure than the finest yet obtained, an
allowance is made. Thus, if it is 999-9 fine, a deduction of Ol
is made from all results of assays checked by it. This deduction
is readily proved to be a very close approximation to the correct
one.
" "
Lastly, the weigh ing-iii correction is applied. If the
original weight taken was say 1000-4 (recorded as + 4), it is
sufficient to deduct 0'4 from the final weighing. In this correc-
tion 0'4 of alloy is reckoned as fine gold, but the error is
inappreciable, and will remain so if the difference from 1000 is
kept less than 0*5.
Limits of Accuracy in Gold Bullion Assay. Attention
may here be drawn to the errors introduced by the lack of
delicacy of the very finest assay balances in ordinary use. It
has elsewhere been shown by the author * that by weighing in
the way indicated above, errors not greater than 0-15 per 1,000
may be introduced. It is, therefore, clear that this amount
represents the limit of accuracy when such balances are
used.
By weighing correctly to O'Ol per 1,000, however, and performing

all other operations with scrupulous care, then in the determina-
tion of gold in high -standard alloys of gold and copper, or of
gold, silver and copper, whether pure
or contaminated by small
nickel and some
quantities of lead, bismuth, zinc, antimony,
other elements, the error does not exceed 0-02 per 1,000, if
the mean of three results is taken,
Parting by Sulphuric Acid. The use of sulphuric acid of
66 B. instead of nitric acid for parting is recommended by some
of gold by dissolution in
assayers on the ground that the losses
nitric acid are variable, while sulphuric acid does not dissolve
gold. The inconveniences suffered by the use of sulphuric acid
are that (1) lead and platinum are left undissolved by it; (2)
violent bumping of the liquid occurs during ebullition ; and (3)
*
Journ. Chem. Soc., vol. Ixiii., pp. 700-714.
sulphate of silver is not very soluble in water, and the washing

is consequently done with dilute sulphuric acid. However, less
silver is left undissolved in the cornets than in the parting by
nitric acid if the proportion of gold to silver is between 1 2 :
and 1 3. :
Preliminary Assay. If the composition of an alloy is quite

unknown, a preliminary assay is necessary in order to determine
the right quantities of silver, copper and lead to be added. This
determination may be made by the touchstone, by considerations
of the colour and hardness of the alloy, or by cupelling 2 grains
of it with 6 grains of silver and 30 grains of lead, and parting
the button in a flask. Simple cupellation with lead gives satis-
factory results if silver is absent or insignificant in quantity ;
according to Fremy this method, in which parting is dispensed
with, is accurate to 3 milliemes if carefully performed with proofs.
Assay of Gold by Means of Cadmium. Balling has shown*
that cadmium may be substituted for silver in the operation of
parting. The J gramme of gold alloy is placed in a porcelain
crucible in which a little fragment of potassic cyanide has been
previously fused in order to protect the metal from the air.
Cadmium is then added in the proportion of 2 J to 1 of gold. If
silver is present in addition, the combined weight of cadmium
and silver must be 2^ times that of the gold. The whole is
fused and then cooled and plunged into hot water to clean the
button, which is then parted in nitric acid (specific gravity 1-3),
dried and weighed. The silver, if any, can be estimated by pre-
cipitation as chloride from the acid solution. By this method
the losses of gold and silver incidental to cupellation are entirely
avoided. A
similar method, employing zinc in place of cadmium,
had previously been recommended by Giiptner.f
Alloys of Gold, Silver and Copper. These may be assayed
by the method just given, the copper being estimated as differ-
ence ;
or the gold may be estimated as usual, and other assay
pieces cupelled with enough lead to remove all the copper. The
buttons thus obtained contain silver and gold only, and the pro-
portion of silver is found by difference. The method of double
cupellation, by which the button of silver and gold is weighed
and then subjected to inquartation and parting is less accurate.
The cupellation designed to remove the copper is made with
less lead than the quantities given on p. 435. If little gold is
present, half the amount of lead there given is used, with
increasing proportions as the amount of gold present increases.
The temperature of cupellation must also be lower than for gold,
approximating more to that used for silver. Proofs of similar
composition must be used and the operations require much
practice before the necessary skill is acquired. If less than
* Oestr.
Zeitsch.fiir Berg, und Httnwesn., 1879, p. 597.
t Zeitsch. fiir Anal. Chem., 1879, p. 104.
THE ASSAY OF GOLD 13ULLION'. 453
30 per cent, of gold is present original!}', the alloy may be

parted at once without cupellation and the silver estimated by
Aveighing as chloride.
Effects of the Presence of Other Metals on G-old
Bullion Assay. The effects on cupellation are the same as those
given under ore assay, p. 420. In general, if a scoria is formed
owing to the presence of large quantities of antimony, arsenic,
oobalt, nickel, iron, tin, zinc, or aluminium, there is a loss of
gold. The alloy should in that case be scorified with lead as a
preliminary to cupellation. If mercury is present gold is vola-
tilised (vide p. 455). The presence of tellurium is indicated by
the formation of numbers of minute beads of precious metal
dispersed over the surface of the cupel. Tellurium compounds
are best analysed in the wet way, see p. 457.
If members of the platinum group are present they remain
undissolved by the parting acid, and hinder the solution of the
silver, and the assay is consequently rendered unreliable. The
treatment of the alloys is discussed below, p. 455. The effects
of the presence of small quantities of various metals on the
surcharge in the ordinary parting assay is shown below. The
table is the result of experiments made by the author.* The
presence of 5 per cent, bismuth does not affect the surcharge.
All assay pieces contained 1,000 parts of gold, 2,500 parts of
silver, and 91 parts of copper, other metals being added in the
proportions indicated.
Metal added,
lead and a half part of borax. If the slag becomes pasty towards
the end of the operation more borax is added, a little at a time.
If the lead button obtained is hard, a second scorification is
necessary, with the addition of more lead. There is considerable
loss of gold from volatilisation, and therefore wet methods of
analysis are preferable.
Iron or Manganese Alloys. The operation is tedious and
difficult with these alloys, as they are difficult to fuse, having
higher melting points than pure gold,"* and the oxides of iron
do not form easily fusible compounds with the litharge. An
extremely high temperature and much borax is required \ ten
parts of lead and one of borax usually suffice.
Cobalt and Nickel. Twenty parts of lead are used, but no borax
at first, so that the oxidation of the nickel may not be hindered.
A very high temperature and the subsequent addition of two
parts of borax are necessary. Several successive scorifications
are required as nickel and cobalt are difficult to oxidise.
Zinc. Oxide of zinc does not form a fusible mixture with
litharge, and the slag is only rendered pasty by borax, unless it
is added in large quantities. Gold is lost by volatilisation, but
the loss is minimised by slagging off the zinc as rapidly as
possible. Use fifteen to twenty parts of lead and two to three
parts of borax added little by little, until the slag is fluid. f
Tin. Twenty parts of lead are required ; oxide of tin is rapidly
formed, but the slag is not easily fusible. Large amounts of
borax are necessary, or still better, borax mixed with potash
which forms a fusible stannate with Sn0 2 .
Aluminium. Alloys containing this metal cannot be assayed

by scorification and cupellation. As soon as fusion takes place
in the muffle, aluminium floats to the top of the bath, being of
low density, and is rapidly oxidised, producing alumina which
forms an exceedingly infusible scoria not easily removed by
litharge. The production of the latter, moreover, is checked by
the scum.
Mr. Edward Matthey observes J that the removal of aluminium
by digestion in hydrochloric acid, and collection of the residual
gold, does not yield satisfactory results. The process he recom-
mends is as follows Accurately weighed portions of 50 grains
:
each of the alloys are fused with litharge, under a flux of potas-
sium carbonate and borax with a small proportion of powdered
charcoal, and the resulting slag re-fused with a further small
quantity of litharge and powdered charcoal. The lead buttons
containing all the gold (the aluminium having combined with
the fluxes employed) are cupelled, and the resulting gold
cupelled with silver and parted with nitric acid in the usual
*
Fremy, Ency. Chim., T. iii., L'or, p. 147.
tO/>. cit., p. 148.
Trans. Roy. Soc., 1892, p. 643.
manner. The assays must be worked with checks or standards

of fine gold and pure aluminium.
In the majority of the preceding cases it is better to
analyse
the alloys by wet methods, see p. 457.
B. Amalgams. The alloy is placed in a weighed porcelain
crucible and gradually heated so as to drive off the mercury.
After the greater part of the mercury has been driven off the
temperature is raised to a full red heat which is maintained for
half an hour. About Ol per cent, of mercury still remains in
the gold after this treatment and can only be completely removed
by cupellation and parting. Checks must be used, as the loss of
gold in the operation may amount to 1 per 1,000.
C. Platinum Group. Cupellation must be performed at a
higher temperature than usual, and the iridescent bands are seen
to remain longer, although they are less numerous.
(1) Platinum-Gold Alloys. The button obtained by cupellation
is dull and crystalline. If the alloy contains as much as 7 or 8
per cent, of platinum, the cupellation proceeds slowly, brighten-
ing is only obtained at a very high temperature and the button
appears flattened, and has a rough crystalline surface and a grey
colour. If more than 10 per cent, is present, brightening does
not occur at all, and the other features just mentioned are more
strikingly exhibited. On parting, the platinum is dissolved
with the silver if it does not form more than 10 per cent, of the
original alloy, but the assay-piece must be boiled in acid for a long
time, and the parting is less complete than in the simple case of
a gold-silver button, from 0-05 to 0-4 per cent, of platinum being
left in the cornets. This is completely removed by a second
parting. If a larger proportion of platinum is present pure gold
is added, or else the operations of inquartation and parting are
repeated two or three times, the loss of gold showing a corre-

sponding increase. Unless proofs are made up of similar com-
position the results are not satisfactory if more than 1 or 2
parts of platinum are present per 1,000 of alloy.
According to results recently obtained in the laboratory at the
Royal Mint by Mr. A. Stansfield, it is impossible to free buttons
containing platinum from lead at the ordinary temperatures
attainable in a muffle. It is necessary to employ the oxy-
hydrogeii blowpipe for the purpose, with the result that the
losses are irregular. In the cupellation of pure platinum,
buttons weighing from 0'002 to O'Ol gramme retain about 10
per cent, of their weight of lead, and those weighing from
OO4
to 2 grammes retain about 33 per cent.
If parting is effected by boiling in concentrated sulphuric acid
instead of in nitric acid, almost all the platinum remains undis-
solved with the gold.
Chaudet* recommends the following proportions of silver for
this purpose :
*
L'art de I'essayeur, Paris, 1835.
Ratio of Platinum to Gold.

cornet with the lower face outwards (p.

444), iridium occurs as
black sooty spots or streaks which are seen
by a lens to nil up
depressions in the surface of the gold.
Both rhodium and iridium are almost insoluble in
aqua regia.
If gold alloys containing both of them are
parted in the ordinary
way with nitric acid, only a small quantity of rhodium goes into
solution with the silver. The residue consisting of the gold, the
iridium and most of the rhodium may then be attacked
by
dilute aqua regia when the gold is dissolved
together with only
traces of the other metals. These may be separated by evaporat-
ing the solution to dryness and heating to dull redness, when
the reduced metals being no longer alloyed may be
completely
separated by dissolving the gold in aqua regia.
According to d'Heiinin,* iridium may be separated from gold
by fusing it with fluxes, the charge being made up as follows :
Iridicgold, 12 '5 grains.

Sodic arseniate, 3 ,,
Black flux, 18 ,,
Ordinary flux (consisting of borax, cream

of tartar, charcoal and litharge), 20 ,,
The iridium forms a speiss with the iron and the arsenic, and
the lead button formed at the bottom of the fused mass contains
all the gold.
D. Tellurium Compounds. These must be treated by wet
methods. An aqua regia solution containing both gold and
tellurium is evaporated with a large excess of hydrochloric
acid until no more chlorine is given off, when both gold and
tellurium are readily precipitated by a current of sulphur dioxide
gas. On attacking the precipitate with nitric acid the tellurium
is dissolved in the state of tellurous acid, and the gold residue
may be dried and weighed, and its purity ascertained by

inquartation and parting. Other methods for separating gold
and tellurium will present themselves on studying the properties-
of the latter.
Wet Methods of Assay of Gold Alloys, Compounds,
&c. Assays or complete analyses of gold bullion, natural
minerals, &c., can be made by the ordinary chemical methods
given by Fresenius, Crookes and others. From 1 to 5 grammes
of bullion are usually enough, but 10 grammes are necessary if
the alloy is nearly pure gold. In general the residue left after
prolonged action of nitric or sulphuric acid is not sufficiently
pure to weigh as gold, and complete solution in aqua regia is-
usually necessary, From the solution the gold may be precipi-
tated by (a) ferrous sulphate, (6) sulphurous acid, (c) oxalic acid,
(d) sulphuretted hydrogen, (e) ammonic sulphide, followed by the
addition of hydrochloric acid. The following remarks may be
*
Dingl. Polyt. Journ., vol. cxxxvii., p. 443.
<53 THE METALLURGY OF GOLD.
of value in aiding the chemist in his choice of a precipitant in

any particular case.
Nitric acid must always be expelled from the solution by
warming with successive additions of hydrochloric acid. The
acid solution must not be heated too strongly or loss of gold
ohloride by volatilisation occurs. Some other chlorides escape
more freely. Ferrous sulphate and sulphurous acid act well in
strongly acid (HC1) solutions ; oxalic acid, sulphuretted hydrogen
and ammonic sulphide best in presence of small quantities of
HC1.* The solution should be dilute (say 1 part of gold in
300 of water), so that other metals may not be carried down by
the gold. Sulphate of iron gives a very finely divided precipitate
which is difficult to wash by decantation without loss ; precipita-
tion is slow in cold solutions. Oxalic acid causes plates and
scales to form which are readily washed and are very pure it acts
;
best in boiling liquids, but a temperature of 80 for forty-eight

hours suffices ; in the cold or in the presence of much hydro-
chloric acid or alkaline chlorides the action is very slow and
partial ; a large excess of the precipitant must be present. Oxalic
acid is used for solutions containing metals of the platinum
.group, which are not precipitated by it. Alkaline oxalates act
better than the free acid.
Sulphurous acid is an excellent precipitant for solutions from
which all other metals but gold have been removed. It acts
rapidly and completely in the cold, but unfortunately precipitates
many other metals. Sulphuretted hydrogen is used in the
absence of all other metals whose sulphides are insoluble in
hydrochloric acid.
In all cases careful consideration must be given to the nature
of the base metals present, and the precipitant, which will not
render any of them insoluble, must be selected.
Other Methods of Bullion Assay. Among methods which
have been proposed at various times, and which may still be of
service occasionally in particular cases, may be mentioned :
{!) The by the touchstone (a method formerly more exten-

trial
sively used by jewellers than at present) the assay by means of

;
considerations as to (2) the colour, and (3) the density of alloys ;

(4) spectroscopic assay ; (5) assay by electrolysis, and (6) by the
Induction balance. A brief description of each of these methods
is appended.
Trial by the Touchstone. This is the oldest method
1.
of assay. It consists in rubbing the gold bullion to be tested on
a hard smooth stone, and comparing the appearance and colour
of the streak with those made by carefully prepared needles of
known composition. The effect of the action of nitric acid and
dilute aqua regia on these streaks is also noted. Touchstones
usually consist of Lydianstone or of silicified wood, and black or
H Rose.
dark green stones are best. Only alloys of

gold and copper or
of gold and silver can be thus tested. The trial is more sensitive
for alloys below 750 fine than for
higher standards. The amount
of gold in alloys between 700 and 800 fine can be determined
correct to 5 parts per 1,000.
2. Colour and Hardness of Alloys. These properties
form a guide to the composition of copper-gold alloys, an increase
of copper corresponding to a heightening of the colour and an
increase of the hardness as tested by shears or a knife. On
heating the alloy to redness in air, the degree of blackening of
the surface is a further indication of the percentage composition,
ifcompared with plates of known fineness.
3. Density of Gold-copper Alloys. The determination of
the fineness of these alloys by taking their densities was investi-
gated by Professor Roberts-Austen at the Royal Mint in 1876.*
He showed that the densities found by experiment were nearly
equal to those obtained by calculation on the assumption that
the union of the two metals was accompanied neither by con-
traction nor expansion. The alloys examined ranged from 860
to 1000 fine, and were made into discs which were all compressed
to the same extent. The conclusion arrived at was that the
fineness of large masses of gold can be deduced from their
densities correct to TO o OQ part. In the case of individual coins
the results are only approximate.
4. Assay by means of the Spectroscope. This method
of determining the composition of gold-copper alloys was investi-
gated by Lockyer <fe Roberts-Austen, f The spectrum of pure
gold was shown to be altered by successive additions of copper,
and near the English standard (916-6 fine) a difference of two
or three-tenths of a millieme in composition could be readily
detected, but the amount of metal volatilised is so small that it
cannot be made to represent with certainty the average com-
position of the mass, which is never perfectly homogeneous.
The detection of traces of gold in alloys or ores by means of
the spectroscope, though sometimes attempted, is not remarkable
for its delicacy or certainty. The method of procedure J is to
dissolve the auriferous material in aqua regia, evaporate off the
nitric acid, and pass induction sparks through the surface film of
liquid, when the spectrum shows some narrow
bands and some
nebulous bands. The latter only are seen if a drop of the
solution is placed in a Bunsen flame. The method may some-
times be useful when complex minerals are being examined.
Mr. Cape! has shown that TT^ny of a milligramme of gold will
show a spectrum if the spark be passed through a weak solution
* Seventh Annual Mint Report, 1876, p. 41.
t Phil. Trans. Roy. Soc., 164, part ii., p. 495, and Mint Report for 1874,
p. 38.
J Fremy, Ency. Chim., T. iii., L'or, p. 134.
of the pure metal. But when operating on a slip of alloy

formed of
Silver, 708
Copper, 254
Gold, 38
1000
the spectra of copper and silver alone were visible. In an alloy

of gold and copper containing from 200 to 250 parts in the
thousand of the precious metal, the gold spectrum is barely
visible. On the other hand, in an alloy of gold and copper
containing traces only (*01 per cent.) of the latter, the copper
spectrum was distinctly shown. Alloys of gold are rarely so
perfectly homogeneous that the particles of metal volatilised and
giving the spectrum represent the whole mass.
5. Assay by Electrolysis. Solutions of gold are readily
decomposed by low tension currents, a difference of potential of
one volt sufficing in the case of chloride of gold. Consequently,
gold can be separated thus in solutions of chlorides from copper,
lead, iron, &c. The process differs little from the electrolytic
assay of copper.* The special precautions to be observed are (a) to
replace the positive platinum pole by one of carbon (b) to cover ;
the platinum cone (negative pole) with a thin deposit of silver or

copper, so as to render the gold easily detachable ; (c) to avoid
an excess of HC1 or 3
HNO
while having a few drops of free
,
H SO
2 4 present. The precipitation is effected at 50
with one
Bunsen cell ; washed, detached from the cone by
the gold is
nitric acid, and weighed with the usual precautions.
6. Assay by the Induction Balance. Full descriptions of
the instrument employed may be found in the published papers
of Professors Hughes and Roberts- Austen, f but the principle on
which it depends may be briefly stated as follows The balance :
of two rapidly-intermittent induced currents of equal strength,

flowing in two coils connected by a wire, is disturbed by the
presence of metal within one of the coils, and the fact of this
disturbance is made evident by a telephone. The balance can
be restored and the telephone silenced by introducing into the
opposing coil an identical mass of metal, and it is found that
widely different effects are produced by equal volumes of different
metals and alloys. In testing discs of the gold copper alloys,
the balance is maintained by a tapered and graduated rod of
zinc, which is moved in and out of one coil ; this method,
however, is not adapted for examining alloys which differ but
slightly in composition, as is shown by the following table :
*
Crookes's Select Methods.
iProc. Roy. Soc., vol. xxix. (1879), p. 56; Phil. Mag. [5], vol. viii.
(1879), p. 50 ; and Proc. Phys. Soc., vol. iii. (1879), p. 81.
Standard by
Assay.
CHAPTER XXL
ECONOMIC CONSIDERATIONS.
Management of Gold Mills. The
cost of extracting gold
from its ores in the mill of interest to the producer
is chiefly
when it is combined with the cost of mining. Nevertheless, it
is convenient that the two items should be ascertained separately.
For this purpose, each truck load of ore delivered from the mine
to the mill, and the amount actually treated in the latter, should
be weighed carefully. Materials, implements, &c., should not be
served out to the mill from the stores without an exactly-worded
order. The cost of transport of the ore from mine to mill is
often ascertained separately, but if this is not done it is better
to include it in the cost of mining. The cost of superintendence
must usually be distributed between the various operations, and
this course must also be sometimes resorted to in the case of
power, lighting, and other items. When the total expenses of
milling are ascertained for any period, say for one month, it is
necessary to allow for depreciation of the plant each time. It is
also advisable not to neglect the question of interest on the
capital sunk in providing the mill, and of a sinking fund, as the
ore may come to an end before the machinery is discarded for
other reasons.
Similarly, exactness in ascertaining the percentage of gold
extracted is in the highest degree desirable. Each load of ore
on its way to the mill should be automatically sampled, and
frequent assays made on mixtures of these samples, the value of
the tailings being determined with equal care. From the results
thus obtained, not only is a watch kept on the relative success
of the treatment from day to day, but an additional check is
afforded on the amount of bullion produced in the mill. A
well-appointed assay office in connection with the works is
obviously necessary in order to carry out these tests, and
laboratory extraction trials should also be made at frequent
intervals in order to determine how far the efficiency of the mill
is being maintained.
Details concerning the cost of treating gold ores by the
various processes have already been given separately in the
chapters respectively devoted to them.
Cost of Production of Gold. In view of the fact that the
value of money is measured in almost all gold-producing coun-
tries by the metal itself, it would be of special interest to
ECONOMIC CONSIDERATIONS. . 463
estimate the average cost per ounce of its production. This was
done by Prof. Roberts-Austen in the case of silver in the year
1887,* and all subsequent computations have served to show
that a high degree of accuracy was attained by him. In the
case of gold, further difficulties are encountered in the en-
deavour to frame a trustworthy estimate, owing to the
differences in the mode of treating silver and gold ores. Thus
the greater part of the silver produced annually is derived from
the output of large mills or smelting establishments, where the
total cost of treatment is well known to the managers, even
though they withhold it from the public. On the other hand,.
a large amount of gold is even now extracted in small mills or
by individuals, particularly in the case of placer deposits, and
the exact cost is frequently a matter of doubt to the proprietors
themselves. Moreover, both silver and gold mining companies
are usually reticent as to their costs, although the advantages of
this course of action to the proprietors, as distinguished from
the managers, are not easy to understand. By certain large
companies, systematic accounts are published, and from these
the following results are extracted :
At the Alaska Treadwell Company's mill about 20,000 tons

of ore, containing 3 dwts. of gold per ton, are treated monthly
by crushing and amalgamation, followed by concentration and
chlorination of the sulphides. The cost (including both mining
and milling) per ounce of gold extracted was 2 Is. Id. in
1890-91, and 2 4s. 8d. in 1891-92. (The value of 1 ounce of
pure gold is 4 4s. lljd.) At the El Callao mine, Venezuela,
the cost per ounce was formerly as low as 25s., but rose to
3 10s. 9d. in 1891, and 3 Is. 8d. in 1892. The cost to the
Montana Company was 3 10s. 6d. per ounce in 1891, and as
much as 4 8s. in 1892. In some of the large hydraulicking
companies the cost was similar to that of the Alaska Treadwell
mine. Thus, at the La Grange Company's workings in 1874-76 r
the cost was 2 4s. 3d. per fine ounce of gold, and at the New
Bloomfield Company's Claim the cost was 3 Os. lOd. in 1875,
and 2 Is. in 1876 and in 1877. The above companies, how-
ever, were thriving, and the average cost per ounce in America,
even at the mines and mills which pay their way, is probably
not less than 3. The average cost of production in California
is estimated by A. G. Charleton f at about 2 2s. 6d. per ounce.
In South Africa, on the Witwatersrand, the cost of production
has been greatly reduced during the last few years. In 1892
the cost of production per fine ounce of gold at those mills
which were working more or less continuously seems to have
been about 3 10s., and in 1895, for the mines situated on the
estimate it at
outcrop of the Main Reef, Hatch and Chalmers
* p. 56.
Nineteenth Annual Report of the Royal Mint, 1888,
t Trans. Fed. Inst. Mng. Eng., 1893, p. 217.
about 2 14s. At
Barberton, the cost of gold obtained at the
Sheba mine is said to be
1 4s. per ounce. In Mysore, where
fuel is costly and the climate bad, the cost is said to be about
1 18s. ICd. per ounce of gold extracted.* At the placer mine
at Saint-Elie in French Guiana, which produces about 1,600
ounces of gold per month, the cost of production was 2 Is. 6d.
in 1887, and at the Siberian placers the cost is 1 17s. 2d. at
Berezovsk. 1 17s. 8d. at Nijni-Tassnil, and 2 15s. 8d. at

Tchernaia-Retchka.
Judging from the published results generally, it would appear
that the average cost of production of gold is somewhat less than
3 per ounce. It must be remembered, however, that the interest
on the initial cost of discovering and developing the mine, and
of the machinery used in both mining and milling is, in many
cases, not included in the so-called cost of production. It has,
moreover, been frequently pointed out, particularly by Mr. A.
Del Mar, that if the working costs of unsuccessful mines, and the
money spent in fruitless prospecting, and in the erection of mills
unsuitable to the ore to be treated, were to be included in the
cost of the production of the gold, it would doubtless appear
that a heavy loss is annually experienced by the gold-mining
industry. It has been well observed that, no matter what the
market prices of gold, silver, copper, &c., may be, mines and ores
will be worked which leave no profit, in the never-failing expec-
tation that the market will rise again, or that the ore will
become more abundant or richer.
The " market price " of gold at any one time must, of course,
be derived from the mean price in gold of all, or at any rate a
large number of, other commodities, when compared with this
mean price at some other time.
Annual Production of the Gold Mines of the World.
The production of gold in ancient times cannot be closely esti-
mated, but, judged from a modern standpoint, it was probably
very small. In the middle ages, however, between the fall of
Rome and the discovery of America, the production was far
smaller than before, and Jacob observes that in this period
" the
precious metals were sought not by exploring the bowels
of the earth, but by the more summary process of conquest,
tribute,and plunder." Even after the exploitation of the New
World began, the output of gold was for many years much too
small to satisfy the cupidity of the conquerors. The develop-
ment of the mining industry was prevented by the ruin and
destruction of the natives, and by the almost incessant irregular
warfare waged against the Spaniards in America in the 16th
century, first by the Dutch and later by the English. Fifty years
elapsed after Columbus discovered America, before the annual
production of gold reached 1,000,000, and even at the end of
*
Ibid., p. 218.
ECONOMIC CONSIDERATIONS. 465
the 17th century Soetbeer estimates that it was only 1,500,000.

The discovery and working of the rich Brazilian placers during
the next half century raised the annual product to over 3,500,000
in the period 1740-1760 (Soetbeer), but as these
deposits became
exhausted, the output again fell off, and in the period 1810-1820
had again sunk to about 1,500,000 per annum. The gradual
development of the Siberian placers was the main cause of the
subsequent steady increase in production up to an average of
7,500,000 per annum in the period 1841-1850, and this was
followed by a sudden rise consequent on the discoveries in
California and Australia. The maximum output from the rich
placers of these countries was reached in 1853, when the world's
production of gold is estimated by Sir Hector Hay to have been
38,000,000. After falling to 21,000,000 in 1867, the output
remained nearly stationary until about the year 1888, when,
from various causes mentioned below, the production again
began to increase, and in 1895 reached 41,000,000, the greatest
amount on record.
The doubling of the world's production which has taken place
in the course of the last eight years is due (1) to the discovery
and development of new districts, (2) to the progress in the
art of metallurgy, and (3) to the increased attention given to gold
mining consequent on the fall in the price of silver. This fall
administered a severe check to the silver mining industry and
set free a considerable amount of skilled labour and of capital
which were largely diverted to gold mining. The most important
new districts which have been developed are the Rand, Cripple
Creek in Colorado, and Coolgardie in Western Australia. The
output of gold at the Rand was valued at 81,000 in 1887, and
7,838,000 in 1895 ; Cripple Creek had not been discovered in
1887, and produced about 1,420,000 in 1895, while Western
Australia produced 20,000 in gold in 1887, and 880,000 in
1895. The total increase in annual output on these three fields
was therefore about 10,000,000 between 1887 and 1895.
The increase of production due to improved metallurgical
methods is difficult to estimate. Most of it must certainly be
which gold
put down to the account of the cyanide process by
valued at about 3,000,000 was produced in 1895, though
the greater part of this was produced on the Rand, and is
included in the amount already given. The increase of about
1,500,000 which took place in the Russian output
between
1888 and 1895 is partly due to improvements in the methods
of washing the gravels, and partly to the extension of the
Siberian railway and the greater attention now being paid
to vein mining. The third cause of the increase in output
of gold has been more felt in the United States and Mexico
than elsewhere. Of the increase in the output of the United
States from little more than 6,000,000 in 1887 to over
30
466 .THE METALLURGY OF GOLD.
.9,000,000 in 1895, about half is attributable to the output

from the new field at Cripple Creek as already noted, and the
remainder is mainly due to the changes in the conditions of
silver mining and their ultimate effects. The increase of about
1,000,000 in Mexico is also to be ascribed to the last-named
cause. The output of Australasia (excluding Western
Australia)
was 2,000,000 more in 1895 than in 1887, and although local
causes were largely responsible for this increase, as, for
example,
the encouragement given by the government of New South
Wales in 1894 and 1895 to "fossicking," nevertheless improved
methods of extraction were also of great effect.
Speaking generally, it is clear from the above considerations
,that of the increase in the annual output of
gold of some
20,000,000 which took place between 1887 and 1895, about
half is due to the discovery of new fields, and the other half
to the development of districts already known. The credit for
liaving caused the latter part of the increase may be about
equally divided between improvements in metallurgical pro-
cesses and the greater attention which has been given to gold
mining during the last few years.
As regards the future production, it is obviously impossible
to predict what will be the effect of the opening up of gold
fields as yet undiscovered. Even without these, however, it
seems almost certain that the production will continue to rise
rapidly for some years to come. The most conservative
estimates place the output of the Rand in five years time at
from 12,000,000 to 14,000,000 per annum, and when the
labour problem has been fully solved, even this huge output
might conceivably be doubled soon afterwards. The output
in Western Australia may similarly be expected to increase
largely when, if ever, the difficulties due to lack of water have
been overcome. Siberia, developed by French engineers with
French capital, may also contribute much more than at present
to the world's production, and the improvements in methods
of extraction which have been so marked of late years will
doubtless help to swell the total. It is, therefore, no exaggera-
tion to say that present indications point to an annual production
of gold of not less than 55,000,000 to 60,000,000 in the
year 1900.
The following table gives details of the production of gold in
various countries in the years 1894 and 1895 :
00-
II
ct -< us cb 04 e * ^MP+
eg
8
00000
CÔ OO
lO O ^X} O "^ O O O CO I"** ^ Oi QO l>* CO CO 'O CO
O'COlOOOJOrfCOiOQOOCOGCTfi lO' i^-liOOO

oooir
-'7OQOC
a G> <o <o in co c-*
i i
<MO--iOOiC5C5COC<JG<J
oocoooor
O CO O "* I
O O O O O O O O O O O O O CO O *O COtiO<N
CCOOCCOCCOCOO^fl^-OOOOCO'
OO
coco
"*(N '^rftf8Wtftftf
rfrfrfl *i* *< i * < * - *'
<MOCOCO(N<NCMr-H
'
e- g.S
The product of ea<

separate colonies. The African product in each year consists

of the returns for the Transvaal made by the Witwatersrand
Chamber of Mines, together with a rough estimate of the gold
produced in the rest of the Continent. The estimates of the
product of most of the other countries in 1895 are those compiled
by the Engineering and Mining Journal, and consist in some
cases of official returns for a part of the year, together with an
estimate of the remainder based on these returns. Thus, for
example, the Russian product consists of official returns for 10J
months in the year, and estimates of the remainder. The output
in Mexico consists of returns for the first 6 months, and an
estimate of the remainder, and the estimate for British Guiana
includes 1 1 months' returns. In other cases, such as most of the
South American States and China, the estimates are merely
based on the output of the previous year. The last two columns
giving the percentages have been added by the author.
Without attempting to make an exact estimate of the amount
of gold won by each metallurgical method, it is obvious from the
fact that almost all the Russian product, about one-fourth of that
from Australasia, and probably over one-fourth of that from the
United States, is derived from placer washings, that these
deposits yield little less than 30 per cent, of the world's pro-
duct. The yield of gold from quartz crushing is now nearly
double as much as that from placers, although, up to a few
years ago, it had always been less. Gold obtained by smelting
(chiefly in the United States) may be about 3 per cent, of
the total produce of the world, while that by the wet
methods (viz., treatment by chlorine and by cyanide of
potassium) is probably over 10 per cent., the reagent last
named being instrumental in obtaining about two -thirds of
this amount. It is probable that over nine- tenths of the gold
is won by methods best suited to the ores dealt with in each
case, taking into account the conditions prevailing in the

district. Alterations in these conditions, such as the improve-
ment of transport or changes in the cost of labour or materials,
may render some other method preferable to the one in use.
The great problem has long been to find a cheap method of
treating large quantities of low-grade pyritic ores
which do not
readily yield their gold to the action of mercury.
Such ores
can usually be treated by chlorination or by smelting, but the
expense of either method is in cases too great to afford a
many
profit. It seems quite possible that the cyanide process will
come into general use for these ores.
Consumption of Gold. Many attempts have been made at
various times to estimate the amount of gold used annually
for purposes other than that of additional coinage. In 1831,
" converted into
Jacob * estimated the annual amount of gold
*
History of the Precious Metals, vol. ii., p. 322.
utensils and ornaments" at about 4,500,000, while the pro-

duction was under 2,000,000, without counting, however,
imports from India and China, which were supposed to be
considerable. In 1881, Dr. Soetbeer estimated the annual
industrial consumption of gold in the world, after making
deductions for old material employed, as amounting to 84,000
kilogrammes or 11.500,000, and in 1885 he put it at 90,000
kilogrammes or 12,300,000, against a production of 21,000,000.
In 1891, the same authority gave the industrial consumption of
gold added to the amount hoarded and that exported to the
East as 120,000 kilogrammes or 16,400,000, against a pro-
duction of 24,000,000. In 1894, Ottomar Haupt estimated
the industrial consumption at only 280,000,000 francs or
11,200,000, but the exports to the East are not included in
this amount. In the same year the Director of the United States
Mint* estimated the industrial consumption at $50,177,300 or
about 10,300,000. Doubtless the industrial consumption of
gold is increasing in amount with the increase of population
and wealth, but nevertheless it is fair to assume that no such
rapid growth has taken place of late years in consumption as
has been already pointed out to have taken place in production.
Taking the annual consumption of gold, including that exported
to the East, as being, on the basis of the estimates given above,
from 10,000,000 to 15,000,000, it would appear that the
amount of gold added to the world's coinage, or retained as
reserves in the form of bullion, was from 5,000,000 to
10,000,000 per annum previous to the year 1890, and has
been gradually increasing ever since, so that in 1895 it was
probably over 25,000,000. Now, in 1894, it was estimated
that the total available stock of gold in the world was
$3,965,900,000 or about 800,000,000, so that the annual
addition to this stock is now probably at the rate of between
3 and 4 per cent, per annum, and may be expected to increase
gradually for some time to come. With regard to the effect
which this increase may have on prices and on commercial
prosperity, it is to be observed that in the years succeeding
1848, when the available stock of gold is estimated to have been
less than 300,000,000, an annual addition (not including the
industrial consumption) was made to the available stock of about
20,000,000, or 6-7 per cent, of the total amount. This addition
is usually supposed to have stimulated trade and
inaugurated a
long period of commercial prosperity. Jevons estimated that
the purchasing power of gold was depreciated by from 9 to 15
per cent., and that it was only kept from falling lower by the
absorption of vast quantities of gold for use in currency by
several countries which had previously mainly employed silver
or paper. Similar results in the near future are now looked for
by some economists.
*
Production of the Precious Metals, 1893, p 53.
BIBLIOGRAPHY.
IT has been found impracticable to enumerate the articles and paragraphs

relating to the metallurgy of gold, which have from time to time appeared
in the various periodicals, as very little search results in the accumulation
of thousands of such references. In general, therefore, only the names'
of some of the publications which contain important or interesting matter
on the subject are given ; but a few exact references on special points
have been added, and many others occur in the footnotes to the text.
PERIODICAL LITERATURE.
Anales de la Mineria Mexicana, 6 sea : Revista de Minas. Mexico. From
1861.
Anales de Minas. Madrid. From 1841.
A nnalen der Berg-und Hutlenkunde. Salzburg, 1802-5.
Annales des Mines. Paris. From 1816.
Annuaire du Journal des Mines de Ruxsie. St. Petersburg. First published
in 1840.
Annuaire des Mines et de la Metallurgie Francaises. Paris. First published
in 1876.
Annual Reports of the Bnllarat School of Mines. From 1882.
Annual Renorts of the Californian State Mineralogist. Sacramento. From
1881.
Annual Reports of the Deputy-Master of the Royal Mint. London. From
1870.
Annual Reports of the Director of the United States Mint. Washington.
Annual Reports on Gold Mining. Victoria, British Columbia. From 1875.
Berg-und Huttenmam. itches Jahrbuch. Vienna. From 1866.
Bercj-und Huttenmdnxische Zeitung. Freiberg. From 1842.
Biennial Reports of the Nevada State Mineralogist. From 1871.
Boletin oficial de minns Madrid, 1844-5.
Bulletin de r Association amicale des anciens Sieves de VEcole des Mines.
Paris. From 1869.
Dingler's Polytechnischfs Journal. From 1815.
Engineering and Mining Journal. New York. From 1866.
Jahrbuch fur Berg-uud Hilttenwesen. Freiberg. From 1827.
Journal des Mines de Freiberg. Koehler. Freiberg, 1788-1793.
Journal des Mines de Ruwie. 1832 to 1835.
La Mineria. Mexico, 1843.
Mining and Scientific Press. San Francisco. From 1864.
Mining and Smelting Magazine. London, 1862-65.
Mining Journal. London. From 1836.
Mining Review. Denver, Colorado, 1873-76.
Mining World. London. From 1871.
Nouveau Journal des Mines de Freiberg. Koehler & Hoffmann. Freiberg,
1795-1804.
472 BIBLIOGRAPHY.
Oesterreich Zeitschriftfur Berg-und Huttenwesen. Otto Freihern. Vienna.

From 1853.
Precious Metals of the United States, Annual Reports on. Washington.
From 1880.
Reports of the Mining Commissioner of Neio Zealand. Wellington, New
Zealand. From 1871.
Revista Minera y Metalurgica. Madrid. From 1850.
School of Mines Quarterly. Columbia, (J.S. From 1879.
Scientific American. New York. From 1846.
Silliman's American Journal of Science and the Arts. New Haven and New
York. From 1816.
Transactions of the American Institute of Mining Engineers. Philadelphia.
From 1871.
GENERAL METALLURGY OF GOLD.

Geber, the Works of. Translated by R. Russell. London, 1686.
Biringuccio. De la Pirotechnia. Venice, 1540. French translation,
Rouen, 1627.
Agricola (G-eorgius). De re metallica. Bale, 1556.
Michaelis (Johannis). De Oro. Leipzig, 1630.
Barba (Alphonzo). Arte de los metales. Madrid, 1639. French trans-
lations, Paris, 1751, and La Haye, 1782.
Schluter. Principles of Metallurgy and Assaying. Brunswick, 1738.
Vargas (Perez de). Traite singulier de metallurgie. Translated from
the Spanish. Paris, 1743.
Lewis (Win.) Commercium Philosophico-Technicum. London, 1763.
Valerius. Grundriss der metallurgie. Ulm, 1768.
Cramer. Principes de metallurgie et docimasie. Blankenburg, 1774,
Jars. Voyages metallurgiques. Paris, 1774-81.
Karsten. System der metallurgie. Breslau, 1818.
Kiessling. Die metallurgie. Dresden, 1841.
Ansted(D. T.) The Gold-Seekers' Manual. London, 1849.
Landrin (H.) Traite' de 1'or. Paris, 1850 and 1863.
Bammelsberg. Lehrbuch der chemischen metallurgie. 1850.
Phillips (J. A.) Gold Mining aud Assaying. London, 1852.
Phillips (J. A.) Encyclopaedia, metropolitana. Article on Metallurgy.
London, 1854.
Eivot (L. ) Principes generaux du traitement des minerals metalliques.
Paris, 1859.
Crookes (Wm. and Rohrig (E.) Treatise on Practical Metallurgy, trans-
)
lated from the German. London, 1860.

Kerl(B.) Die Rammelsberger Hiittenprozesse. Clausthal, 1861.
Kustel (G.) Processes of Gold and Silver Extraction. San Francisco,
1863.
Kerl (B.) Handbuch der metallurgischen Hiittenkunde. Clausthal, 1865.
Phillips (J. A.) The Mining and Metallurgy of Gold and Silver. London,
1867.
Overman (F.) A
Treatise on Metallurgy. York, 1868. New
Blake (W. P.) Report on the Precious Metals. Washington, 1869.
Raymond (R. W.) Mineral Resources West of the Rocky Mountains.
7 vols. Washington, 1869-74.
Raymond (R. W.) Mines, Mills, and Furnaces of the Pacific States. New
York, 1871.
Schiern (F.) Sur 1'origine de la tradition des fourmis qui ramassent Tor.
Copenhagen, 1873.
Greenwood (W. H.) Manual of Metallurgy, vol. ii. London, 1875.
Cox (S. H. F.) Treatment of Gold Ores. Mining Journal, vol. xlvii.,
p. 1388, 1877.
BIBLIOGRAPHY. 473
Attwood (G.). The Batea, the Milling of Auriferous Veinstones, and ths
Mineralisation of Gold. Articles in Alta California. San Francisco
Sept., 1878.
Encyclopaedia Britannica. 9th Edition. Article on "Gold." London,
1879.
Kerl(B.) Grundriss der Metallhiittenkunde. Leipzig, 1880.
Simonin (L. ) L'Or et 1' Argent Paris, 1880.
Industrial Progress in Gold Mining. Philadelphia, 1880.
Percy (John). Metallurgy of Silver and Gold, vol. i. London, 1880.
Ryan (J.) Gold Mining in India. London, 1881.
Lock (A. G.) Gold: Its Occurrence and Extraction. Wii ha Bibliography
London, 1882.
Egleston (T.) The Progress of the Metallurgy of Gold and Silver in the
United States. New York, 1882.
Balch (W. R.) Mines, Miners, and Mining Interests of the United States
in 1882. Philadelphia, 1882.
Restrepo. Estudio sobre las minas de oro y Plata de Colombia. Bogota,
1884.
Gore (G.) Art of Electro-Metallurgy. New York, 1884.
Zoppeti. L'electrolisi in metallurgica. Milan, 1885.
Phillips (J. A. ) and Bauerman. The Elements of Metallurgy. London,
1887.
Egleston (Thos.) Metallurgy of Silver, Gold, and Mercury in the United
States. 2 vols. London, 1887-90.
Balling (C.) Grundriss der electrometallurgie. Stuttgard, 1888.
Fremy. L'Or dans le laboratoire. Encyclopcedie Chimique, vol. iii.
Cahier 16e. Paris, 1888.
Watt. Electro-Metallurgy Practically Considered. London, 1889.
Lock (G. W.) Practical Gold Mining. London, 1889.
Eissler (M.) The Metallurgy of Gold. London, 1891.
Fremy. L'Or dans les centres de travail et de 1'industrie. Encyclo. Chim. ,
vol. v. Cumenge and Fuchs. Paris, 1891.
Raymond (R. W.) Gold and Silver: Report of the llth Census of the
U.S. New York, 1892.
Hatch and Chalmers. Gold Mines of the Band. London, 189o.
De la Coux. L'or. Gites auriferes Extraction de Tor. Paris, 1893.
CHAPTERS I.AND II. PROPERTIES OF GOLD,

ITS ALLOYS AND COMPOUNDS.
Budelius (R.) De monetis. Colon iae A grip, 1591.
Savot. Discours sur les Medalles Antiques. Paris, 1627.
Potier (M.) Philosophica Chemica. Francfort, 1648.
Borrichius. Hermetes Êgyptiorum et Chemicorum Sapientia. Copen-
hagen, 1674.
Gobet. Les anciens mineralogistes du royaume de France. Paris, 1679.
Gellert (C. E.) Metallurgic Chemistry. Translation, London, 1796.
Hatchett (J.) Wear of Coins. Phil. Trans. Hoy. Soc., 1803, p. 43.
Hatchett. Experiences et observations sur 1'or. Paris, 1804.
D'Arcet. L'art de dorer le bronze. Paris, 1818.
Jacob (Wm.) A
History of the Precious Metals. 2 vols. London, 1831.
Schmieder. Geschichte der alchemic. Halle, 1832.
Boue (P.) Traite" d'orfevrerie, bijouterie, et jouaillerie. Paris, 1832.
Levol (C.) Liquation. Ann. de Chimie et de Phys., vol. xxxix., p. 163.
1853.
Ansell (G. F.) A Treatise on Coining. London, 1862.
Rossignol (J. P.) Les metaux dans 1'antiquite". Paris, 1863.
174 BIBLIOGRAPHY.
Watt's Dictionary of Chemistry and Supplements. Articles on "Gold."

London, 1864, 1872, 1875, and 1881.
Wilm (E.) Dictionnaire de chimie, Wurtz. Article on "L'Or." Paris,
1868.
Ronchaud (L. de). Dictionnaire des antiquites grecques et roraaines.
Article, "aurum."
Mommsen. Histoire de la monnaie romaine. Paris, 1868.
Jevons (S.) Wear of Coin. Jour n. of Statistical Society. London, 18G8.
Report on European Mints. London, 1870.
Skey Wm.
( Gold and Platina in Solutions of Alkaline Sulphates. Trans.
)
N.Z. Inst., vol. iv., p. 313. Wellington, 1872.

Skey (Wm.) On the Mode of Producing Alloys by Wet Processes.
Trans. N. Z. List., vol. v., 1873.
Skey (Wm. ) On the Oxidation of Gold in the Presence of Water. Trans.
N. Z. ln*t., vol. viii., p. 339, 1876.
Peligot. Alloys used for Coinage. Comptes Rendus, vol. Ixxvi. (1873),
p. 1441.
Booth (J. C.) and Wm. E. Dubois. Condemnation of Ingot Melts.
Washington, 1875.
Lepsius. Les metaux chez les Egypbiens. Paris, 1877.
Wright (C. R. A.) Metals. London, 1878.
Lenormant. La Monnaie dans 1'antiquite. 3 vols. Paris, 1878.
<3-ee (G-. E.) The Goldsmith's Handbook. London, 1879.
Pollen ( J. H. ) Ancient and Modern Gold and Silversmith's Work. London,
1879.
Noback Miinz-, Maass- und Gewichtsbuch. Leipzig, 1879.
(Fr.)
Cripps. Old English Plate. London, 1878, 1891.
Douan. Inoxydation, dorure, et platinage des metaux. Paris, 1880.
Kriiss (G.) Untersuchungen viber das Atomgewicht des Gold. Munich,
1880.
Brandis. Das Munz-mass und Gewichtwesen in Vorder-Asien bis auf
Alexander den Grossen. Berlin.
Del Mar (Alex.) History of the Precious Metals. London, 1880.
Wagner (A.) Gold, Silber, und Edelslema. Vienna, Leipzig, and Pesth.
1881.
Wheatley and Delamotte. Art Work in Gold and Silver, London, 1881.
Bloxam (C. L.) Metals their Properties and Treatment. London,
: 1882.
Martin (W.) Wear of Coins. Journ. Inst. Bankers. London, June, 1882.
Achiardi (T.) Metalli. Milan, 1883.
Kenyon (R. L. ) The Gold Coins of England. London, 1884.
Roberts-Austen (W. C. ) Alloys used for Coinage. Cantor Lectures, Soc.
of Arts. London, 1884.
Blake (W. P. ) Crystalline Forms of Gold. Precious Metals of the United
States. Washington, 1884.
Berthelot (M.) Origines de 1'alchemie. Paris, 1885.
Streeter (E. W.) Gold the Standards of all Countries. London, 1885.
:
Kopp (H.) Die Alchemic in alterer und neuerer Zeit. Heidelberg, 1886.
Rochas (A. de). L'or Alchemique. Article in La Nature. Paris, 1886.
Schaefer (H. W.) Die Alchemic, &c. Flensborg, 1887.
Roberts- Austen (W. C.) Crystallisation in Gold. Phil. Trans. Roy. Soc.,
vol. clxxix. (1888), p. 339.
Peligot. Liquation. Bull, de la Soc. d' encouragement, vol. iv. (1889),
p. 171.
Riche (A.) Monnaie, Medailles et Bijoux. Paris, 1889.
Brannt (W. T.) Metallic Alloys. London, 1889.
G-uettier (A.) Practical Guide for the Manufacture of Metallic Alloys,
1865. Translated from the French by A. A. Fesquet. New York, 1890.
Hiorns (A. H.) Mixed Metals. London, 1891.
Wagner (A.) Gold, Silber und Edelsteine. Handbuch fur Gold-, Silber-,
Bronze- Arbeiter und Juweliere. Leipzig, 1895.
BIBLIOGRAPHY. 475
CHAPTER III. MODE OF OCCURRENCE AND DISTRIBUTION

OF GOLD. PRODUCTION OF GOLD.
Holzchul. Remarques sur Tor des mines de Saxe. Penig, 1805.
Atkinson (S.) Discoverie and Historie of the Gold Mines in Scotland
Edinburgh, 1825.
Miers Travels in Chile and La Plata. London, 1828.
(J.)
Humboldt. Fluctuations in the Supplies of Gold. London, 1839.
Dupont (S. C. De la Production des Metaux precieux au Mexique:
)
considered dans ses rapports avec la metallurgie, &c. Paris, 1843.

Papers relating to the Discovery of Gold in Australia. 2 vols. London
1852-57.
Reports on the Gold Returns of Victoria of 1859-62. Melbourne.
Ansted (D. T.) Gold in Wales. Min Journ., vol. xxviii., p. 241, 1858.
Clarke (W. B.) Researches in the Southern Gold Fields of New South
Wales. Sydney, 1860.
Readwin (T. A.) The Gold Discoveries in Merionethshire; and a Mode
for its Economic Extraction. Manchester, 1860.
Davison (S.) Geognosy of Gold Deposits in Australia. London, 1861.
Resales (H.) Essay on the Origin and Distribution of Gold in Quartz
Veins. Melbourne, 1861.
Dubois (Win. E.) The Dissemination of Gold. Trans. Am. Phil. Soc.
Philadelphia, June, 1861.
Clarke ( W. B. ) Auriferous and Non -auriferous Quartz Reefs of Australia.
Geol. Mag., vol. iii., p. 561, 1866.
Brown (J. R.) Mineral Resources of the United States. Washington,
1S67.
Smyth Brough). Gold Fields of Victoria. Melbourne, 1867.
(R.
Lovell (J.)Gold Fields of Nova Scotia. Montreal, 1868.
Forbes (D.) Gold from Clogau. Geol. Mag., vol. v., p. 224, 1868.
Bankart (H.) Gold Fields of Uruguay. Journ. Roy. Geog. Soc., vol.
xxxix., p. 339. London, 1869.
Mackay (J.) Report on the Thames Gold Fields. Wellington, N.Z.,
1869.
Cotta (B. von). Treatise on Ore Deposits. Translated. New York,
1870.
Bateman(A. W.) South African Gold Fields. The Times, Sept. 28, 1874.
Domeyko (J.) Ensayo sobre los Depositos Metaliferos de Chile. Santiago,
1876.
Bain (A. G.) Gold Regions of S.E. Africa. London and Cape Colony,
1877.
Ott (A.) Nature and Distribution of Gold in Metallic Sulphides. John
Franklin Institute Journal. 3rd series, vol. IviL, pp. 129-132.
Phillips (J. A.) Ore Deposits. London, 1878.
Daintree (R. ) Modes of Occurrence of Gold in Australia. Journ. Geol
Soc., vol. xxxiv., p. 431, 1878.
Jenney (W. P.) Mineral Resources of the Black Hills of Dakota. Wash-
ington, 1880.
Ball (Prof. B. ) Diamonds, Coal, and Gold in India their Occurrence and
:
Distribution. London, 1881.

Blake (W. P.) Geology and Mineralogy of California. Sacramento, 1881.
Jervis (G.) Dell'Oro in Natura. Rome, 1881.
Burton (R. F.) Gold on the Gold Coast. Journ. Soc. of Arts, vol. xxx.,
p. 785, 1882.
Ratte (A. F.) Descriptive Catalogue (with notes) of the General Collection
of Minerals in the Australian Museum. Sydney, 1885.
Emmons (S. F.) and Becker (G. F.) Precious Metals: being vol. xm.
of U.S. Census Reports of 1880. Washington, 1885.
476 BIBLIOGRAPHY.
Handbook of New
Zealand Mines. Wellington, 1887.
Liversedge Minerals of New South Wales. London, 1888.
(J.)
Mathers (E. P.) Gold Fields of South Africa revisited. London, 1889.
Anderson (J. W.) The Prospector's Handbook a guide for the Prospector
;
and Traveller in search of metal-bearing and other valuable minerals.

London, 1889.
CHAPTERS IV. AND V. PLACER MINING.

Moneeram. Native Account of Washing for Gold in Assam. Journ. Asiat.
Soc. Bengal, vol. vii., p. 621, 1838.
Abbott (Capt. J.) Account of the process employed for obtaining gold from
the sand of the river Bey ass with a short account of the gold mines
:
of Siberia. Journ. Roy. Asiat, Soc. Bengal, vol. xvi., pp. 266-272, 1847.
Delesse. Gisement et exploitation de 1'or en Australie. Paris, 1853.
Blake (W. P.) Hydraulic Mining in Georgia. Am. Journ. Scl and Art.
2nd series, vol. xxvi., p. 278, 1858.
Report of the Royal Commission appointed to inquire into the best methods
of removing sludge from the gold tields. Melbourne, 1859.
Sauvage (Edw.) On Hydraulic Gold Mining in California. Proc. Inst.
C.E., vol. xlv., p. 321, 1859.
Radde (Gustav). Reisen im Siiden von Ost-Sibirien in den Jahren, 1855-59.
St. Petersburg, 1863.
Debombourg (G.) Gallia aurifera. Êtudes sur les alluvious auriferes de
la France. Lyons, 1868.
Wilkinson (C.) Formation of Gold Nuggets in Drift. Trans. Roy. Soc.
Victoria, vol. viii., p. 11, 1872.
Newbery (J. Cosmo). Formation of Nuggets in Auriferous Deposits.
Trans. Roy. Soc. of Victoria, vol. ix., pp. 52-60, 1873.
Skey (Wm.) Formation of Gold Nuggets in Drift. Trans. N.Z. Inst.
vol. v., p. 377, 1873.
Christy (S. B.) Ocean Placers of San Francisco. Proc. Cal. Acad. Sci.,
August, 1878.
Egleston (T. ) Hydraulic Mining in California. London, 1878.
Goodyear (W. A.) Auriferous Gravels of California. Proc. Cal. Acad.
Scl San Francisco, 1879.
Whitney (J. D.) Auriferous Gravels of the Sierra Nevada of California.
Cambridge, U.S., 1880.
Egleston (T.) Formation of Nuggets and Placer Deposits. Trans. Am.
Inst. Mug. Eng., vol. ix., p. 633, 1881.
Hammond (J. H.) Auriferous Gravels of California and the Methods of
Drift Mining. Prod, of Gold and Silver in U.S. for 1881.
Washington,
Newbery (J. Cosmo). Genesis and Distribution of Gold. The School of

Mines Quarterly, vol. iii., 1882.
Randall (P. M.) The Miner's Inch. Prod, of Gold and Silver in U.S. for
1884- Washington, 1885.
Bowie, Jr. (A. J. ) Practical Treatise on Hydraulic Mining in California.
New York, 1885.
Bowie, Jr. (A. J.) Mining Debris in Californian Rivers. San Francisco,
1887.
Gould (E. S.) Practical Hydraulic Formulas for the Distribution of Water
through long Pipes. New York, 1891.
Kirkpatrick (T. S. G.) Hydraulic Gold Miner's Manual. New York,
Wagenen (T. F. von). Manual of Hydraulic Gold Mining. New York,
1891.
BIBLIOGRAPHY. 477
CHAPTERS VI., VII., AND VIII.-QUARTZ CRUSHING AND

AMALGAMATION.
De Born. Amalgamation des minerals d'or et d 'argent. Vienna, 1786.
English translation, 1791.
Sonneschmied. L'Ainalgame espagnol. Leipzig, 1811.
Sonneschmied. Traite sur 1'amalgamation. Ronneburg, 1811.
Kivot. Nouveau precede de traitement des minerais d'or et
d'argent (a
manuscript in the archives of the Ecole des Mines de Paris). Paris,
1818.
Ortmann. Kurze Geschichte der Amalgamation in Sachsen. Freiberg, n.d.
Lawson (G.) Improvements in Amalgamation. Trans. Nova Scotia Insl.
1866.
Hague (J[.
D. )
Gold Mining in Colorado. Report on the Fortieth Parallel,
vol. Washington, 1870.
iii.
Keith (N. S.) Amalgamated Copper Plates. Enrt. and Mng. Journ., vol.
xi.,p. 270, 1871.
Blake (W. P.) Mining Machinery. New Haven, 1871.
Fonseca. Memoire sur 1'amalgamation Chilienne. Paris, 1872.
Bergmann (E. von). Die Anfange des Geldes in Êgypten. Vienna, 1872.
Thompson (H. A.) Extraction of Gold. Trans. Boy. Soc. Viet., vol. viii.,
pp. 15-26, 1872.
Skey (Wm.) Electromotive Power of Certain Metals in Cyanide of
Potassium with reference to Gold Milling. Trans. N.Z. Inst., vol.
viii., p. 334, 1876.
Attwood (G.) Chile Vein Gold Works, South America. Proc. Inst. C.E.,
vollvi., p. 244, 1879.
Cumenge and Fuchs. Effect of Antimony and Arsenic on Amalgamation.
Comptes fiendus, March 17, 1879.
Egleston (T.) Californian Stamp Mills. London, 1880.
Randall (P. M.) Quartz Operators' Handbook. New York, 1880.
Habermann (I.) The Heberle Mill. Oesterr. Ztschr. fur Berg, und
Htnwesen, 1880.
Egleston (T.) Treatment of Gold Quartz in California. London, 1881.
Egleston (T.) Losses in Amalgamation. Trans. Am. Inst. Mug. Eng.,
vol. ix., p. 633, 1881.
Egleston (T.) Causes of Rustiness in Gold. Trans. Am. Inst. Mng. Eng.,
vol. ix., p. 646. New York, 1881.
Yale (Charles G.) Mining Machinery. Prod, of Gold and Silver in the
U.S., 1881.
Richards (J. ) Quartz Crushing Machinery. Prod, of Gold and Silver in
U.S. Washington, 1881.
Attwood (G.) Milling of Gold Quartz. Prod, of Gold and Silver in U.S.,
1881.
Reed (S. A.) Ore Sampling. School of Mines Quarterly, vol. iii., p. 253.,
1881.
M'Dermott and Duffield. Gold Amalgamation and Concentration. London
and New York, 1890.
Lock (G. Warnford). Gold Amalgamation. Proc. Inst. Mng. and Met.,
Session ii., 3rd meeting. London, Dec., 1892.
Curtis (A. Harper). Gold Quartz Reduction. Proc. Inst. C.E., vol. cviii.
(1892), part ii.
Charleton (A. G.) Coarse and Fine Crushing. Trans. Fed. Inst. Mng.
Eng., 1892-3.Four Papers.
Louis (H.) Handbook of Gold-Milling. London, 1894.
Rickard (T. A.) Variations in Gold Milling. New York, 1895.
478 BIBLIOGRAPHY.
CHAPTER DC-CONCENTRATION OF GOLD ORES.

Bittinger (P. von). Lehrbuch der Aufbereitungskunde. Berlin, 1867,
1870, and 1873.
Smyth (Sir W. W.) Dressing, or the Mechanical Preparation of Gold
Ores. Lectures on Gold. Mng. Journ., 1873.
Full descriptions of the different kinds of round buddies are given in the
following papers :
Teague (W., Jun.) "On Dressing Tin Ores."

Proc. Mining Inst.
Cornwall, vol. i., No. 3 (Truro, 1877).
Ferguson (H. T.) "On the Mechanical Appliances used for Dress-
ing Tin and Copper Ores in Cornwall." Proc. Inst. Alech. Eng.,
1873.
Schmidt (A. W. )
Der Schlamfiinger auch Kornfanger genannt. Dillen-
burg, 1877.
Reytt (C. von). Comparison of the Hand Buddie and the Rotary and
Percussion Tables. Berg-u. Huttenmann. Jahrbuch, vol. xxii., 1877.
Habermann (J. ) Comparison of the Salzburg Table, the Rittinger Table,
and the Hand Buddie. Oesterr. Zeitschr. fur Berg-und Htittenwesen,
1879, No. 8.
Cazin (F. M. F.)Dynamical Metallurgy or Mechanical Ore Concentra-
tion.Mining Record, 1881-2.
Richards (R. H. A new Hydraulic Separator to prepare Ores for Jigging
)
and Table Work. Trans. Am. Inst. Mng. Eng., 1883.

Reytt (C. von). Salzburg Percussion Table. Berg-und Hilttenmanisches
Jahrbuch, xxxiii., p. 3, 1S83.
Gallon (J.) Lectures on Mining. Translated by Le Neve Foster and
W. Galloway, vol. iii. London, 1886.
The Settling of Solid Particles in Liquids. Bulletin No. 36. United Staffs
Geological Survey. Washington, 1886.
Ore Dressing in California. Sixth Report of the Col. State Min., 1886.
This is a full account of the machines actually at work.
Reytt (C. von). The Linkenbach Table and Hand-Buddie compared.
Oesterr. Ztch., Oct. 6, 1888.
Lock (C. G. W.) Mining and Ore Dressing Machinery. London, 1891.
Kunhardt (W. B. ) The Practice of Ore Dressing in Europe. York, New
1891.
Scheidel (A.) New Process of Dressing the Gold Ores of the Thames
Valley. New Zealand Mng. Comm. Rept.. 1891, p. 28.
Meier (J. W.) Concentration at Pribram, Bohemia. Eng. and Mng.
Journ., vol. liv. (1892), p. 665, vol. Iv. (1893), pp. 5, 28, and 52.
Commans. Concentration and Sizing of Crushed Ore. Proc. Inst. Civil
Eng., 1894.
Resales (H. ) Report on the Loss of Gold in the reduction of Auriferous
Veinstone in Victoria. Melbourne, 1895.
CHAPTER XL-ROASTING OF GOLD ORES.

Plattner (C. F.) Die metallurgische Rostprozesse.
Freiburg, 1856.
Dixon (Wm.) of extracting Gold, Silver, and other Metals from
Methods
Pyrites. Chemical News, vol. xxxviii., pp. 281, 293, and 301; and
vol. xxxix., p. 7. 1879.
Kiistel (G-. ) Roasting of Gold and Silver Ores. San Francisco, 1880.
Cumenge (E. ) Note relative a 1'emploi de la vapeur d'eau dans certaines
operations metallurgiques. Annettes des mines, vol. i., p. 1852. Paris,
1882.
Stetefeldt (C. A.) On Salt Roasting. Trans. Am. Inst. Mng. Eng. y
vol. xiii. New
Haven, 1885.
BIBLIOGRAPHY. 479
Christy (S. B.) Losses of Gold in Roasting. Tran*. Am. Inst. Mnn. Ena
1888.
Producer Gas for Roasting. Tenth Col. State Min. Report, p 897 San
Francisco, 1890.
Adams (W. H.) Pyrites Practical Methods for Extraction of Gold, &c.
:
New York, 1892.

Stetefeldt (C. A.) Taylor's Gas Producer for Roasting. Enq. and Mna.
Journ., vol. Ivi. (1893), p. 124.
CHAPTERS XII., XIII. , AND XIV. CHLORINATION.

The earliest researches on the chlorination of gold were made by
(1) Duflos (Dr.) Die schles. gesell. Uebersicht. Breslau, 1848.
(2) Richter (Theo.) Journ. far Prak. Chem., vol. li. (1S49), p. 151.
(3) Lange Karsten's Archiv, vol. xxiv., pp. 396-429.
(Herr). 1852.
(4) Plattner (C. F.)
Probirkunst, p. 570. 1853.
(5) Percy (J.) Phil. Mag., vol. xxxvi. (1853), pp. 1-8.
Whelpley and Storer. Method of separating Metals from Sulphurets.
Boston, 1866.
Kiistsl (G.) Concentration and Chlorination. San Francisco, 1868.
Bleasdale (J. J.) On Chlorine as a Solvent for Gold. Trans. Roy. Soc.
Victoria, vol. vi., pp. 47-52. 1870.
Pyrites Report of the Board appointed to report on the methods of treat-
:
ing Pyrites and Pyritous Veinstutf, as practised on the Gold Fields.

Melbourne, 1874.
Attwood (G-. M.), and Aaron (C. H.) Prod, of Gold and Silver in the
United States. Washington, 1881.
France (Ch. de). Extraction par voiehumide du cuivre, de 1'argent, et de-
1'or. Brussels, 1882.
Egleston (T. ) Leaching Gold and Silver Ores in the West. New York,
1883.
Egleston (T.) Leaching Gold Ores containing Silver. London, 1886.
Stetefeldt (C. A.) Lixiviation of Silver Ores. New York, 1888. Thi
work gives many details applicable to all wet processes.
O'Driscoll (F. ) Notes on the Treatment of Gold Ores. London, 1889.
The Pollok Process of Chlorination. Eng. and Mng. Journ., vol. xlix.
(1890), Feb. 15.
Burfeind (J. H.) Chlorination of Gold Ores. Eng. and Mng. Journ., voL
li. (1891), p. 446.
Precipitation of Gold from Chloride Solution. Letters and Articles by John
E. Rothwell, W. Langguth, C. H. Aaron, L. D. Godshall, J. T. Blom-
field, and T. K. Rose in Eng. and Mng. Journ.,
vol. li. (1891), pp. 74,
112, 165, 204, 229, 282, 347, 373, 465; vol. lii. (1891), p. 211; and
Mining Journal, March, 18, 1893.
Vautin (C.) Decomposition of Auric Chloride. Proc. Inst. Mining and
Met., Session ii., 5th meeting. London, Feb. 15, 1892.
Langguth (W.) Chlorination of Gold. Trans. Am. Inst. Mng. Eng.
June, 1892.
Barrel Chlorination. Eng. and Mng. Journ., vol. Iv. (1893), pp. 244 and
269.
Rothwell (J. E.) Recent improvements in Chlorination. Mineral Industry
for 1892, p. 236. New York, 1893.
Godshall (L. D.) Modern Chlorination. Eng. and Mng. Journ., vol.
Ivii. (1894), Jan. 6 and 13.
480 BIBLIOGRAPHY.
CHAPTERS XV. AND XVI.-CYANIDE PROCESS.

Scheidel(A.) The Cyanide Process. Sacramento, 1894.
Janin, Jr. (L.) Article in Mineral Industry for 1892, pp. 239-270.
Reunert(T.) Diamonds and Gold in South Africa. London, 1894.
Butters (C.) and Smart (E.) Plant for the extraction of Gold by the
Cyanide Process. Proc. Inst. Civil Eng., vol. cxx. 1895.
Eissler (M.) The Cyanide Process. London, 1895.
CHAPTER XVII. PYRITIC SMELTING.

Austin (W. L.) Matting Dry Auriferous Sulphides. Trans. Am. Inst.
Mng. Eng., 1889.
The Austin System of Pyritic Smelting. Denver, Colo., 1892.
Pyritic Smelting. Articles and letters on the subject have appeared in the
Engineering and Mining Journal under the following dates Vol. 1. :
(1890), Aug. 2 and Nov. 15; vol. li., p. 229 (Feb. 21, 1891) vol. Hi., ;
pp. 174, 471, 721 (Aug. 8, Oct. 24, and Dec. 26, 1891); vol. lv.,
pp. 28, 99. 244, 292, 339, and 364 (Jan. 14, Feb. 4, Mar. 18, April 1,
April 15. April 22, 1893).
l,ang(H.) Matte Smelting. New York, 1896.
CHAPTER XVIII. -REFINING AND PARTING OF GOLD

BULLION.
Goddard (Jonathan). Experiments on Refining Gold with Antimony.
Phil. Trans. Roy. Soc., 1676.
Leibius (A. ) Separation of Gold from Silver Chloride in Miller's Process.
Trans. Roy. Soc. of Aew South Wales, 1872, p. 67.
Egleston (T.) Parting Gold and Silver in California. New
York, 1877.
Egleston (T.) Parting Gold and Silver by means of Iron at Lautenthal.
New York, 1885.
Êgleston (T.) The Separation of Silver and Gold from Copper at Oker.
Washington, 1885.
Hugon. Etude sur le raffinage electrolytique du cuivre noir. Paris, 1885.
Egleston (T.) Treatment of Gold and Silver at the United States Mint.
London, 1886.
Skidmore (W.) Parting Gold and Silver in the United States. Ninth
Report of the Cal. State Min., 1889, pp. 67-90. This is a complete
account of the methods in use in the United States.
CHAPTERS XIX. AND XX. ASSAYING.

Carranza (A. ) El Ainstamieto i Proporcion de las Monedas de Oro, Plata
i Cobre, i la reduccion distos Metales a su Debida estimaccion. Madrid,
1629.
Badrock (Wm.) A new Touchstone for Gold and Silver Wares. London,
1651. 2nd edition, 1679.
Le Febure (N. R.) Compleat body of chymistry. 1670.
Petty (Sir John). Laws of Nature in Assaying Metals. Translated in part
from the German of L. Erckern. London, 1683.
Cramer (J. A.) Elementa artis docimasticae. Lugduni Batavorum
(Leyden), 1744.
Symonds (W.) Essay on the Weighing of Gold, &c. London, 1756.
Cramer, M.D. (J. A.) Elements of the Art of Assaying Metals. London,
1764.
BIBLIOGRAPHY. 4$J
Pouchet. Le nouveau titre des matieres d'or et cl 'argent. Rouen, 1798.

Becquerel. Gold and Electricity. Ann. de Chimie ei de Physique, vol.
xxiv. (1823) also Article by Oersted, d vol. xxxix. (1828),
;
p. 274.
Chaudet. L'Art de 1'essayeur. Paris, 1835.
Bodemann (Th. Anleitung zur Berg-und Hiittenmannischen Probirkunst.
)
Clausthal, 1845.
Berthier. Traite des essais par la voie se"che. Paris, 1847.
Pettenkofer. Bergiverksfreund, vol. xii. (1849). Article on Gold Bullion
Assaying.
Watherston (J. H.) The Gold Valuer. London, 1852.
Plattner (C. F.) Probirkunst. Freiberg, 1853.
Bodemann and Kerl. Treatise on Assaying. Translated by W. A. Good-
year. New York, 1868.
Domeyko (J. )
Tratado de Ensayes, tanto por la via seca comopor la via
humeda. Chile, 1873.
Foord (G.) Mechanical Assay of Quartz. Trans. Roy. Soc., Victoria,
vol. x.,pp. 139-147. 1874.
Broch (Dr. O.) Assay of Gold by means of its Density. Norwegian Nyt.
Mag. fur Xatnrvsk. Christiania, 1876.
Ricketts ) Notes on Assaying. New York, 1876.
(P.
Kerl(B.) Metallurgische Probirkunst. 1880.
Attwood (G.) Practical Blowpipe Assaying. London, 1880.
Chapman (E. J. ) Assay Notes. Practical instructions for the determin-
ation by furnace assay of Gold and Silver in rocks and ores. Toronto,
1881.
Mitchell (W.) Manual of Practical Assaying. Edited by Wm. Crookes.
London, 1881.
Balling (C.) L'Art de 1'essayeur. Paris, 1881.
Rossler (H. ) Article on Gold Bullion Assaying. Dinyler's Polyt. Journ.,
vol. ccvi. (1884).
Black (J. Chemistry of the Gold Fields. Dunedin, N.Z., 1885.
G.)
Aaron (C. Manual of Assaying. San Francisco, 1885.
H.)
Hiorns (A. H. ) Practical Metallurgy and Assaying. London, 1888.
Fremy. L'or dans le Laboratoire. Cumenge and Fuchs. Ency. Chirn.,
vol. iii., c. 16e. Paris, 1888.
Ross (W. A.) Blowpipe Analysis. London, 1889.
Brown and Griffiths. Manual of Assaying of Gold, Silver, &c. London,
1890.
Beringer (J. J. & C. ) Manual of Assaying. London, 1890.
Lieber (O. M.) Assayer's Guide. New York.
Plattner. Blowpipe Analysis. Enlarged by Richter (Th. ) Translated by
H. B. Cornwall. New York, 1890.
Riche (A.) L'Art de 1'essayeur. Paris, 1892.
Furman. Practical Assaying. New York, 1894.
31
483
INDEX.
Aaron, C. H., 127, 274, 278, 412. Antimony alloys, Assay of, 453.
Africa, Gold
in, 40, 207, 468. ,, in roasting furnace, 231,
Agitation in cyanide process, 312. 232.
Agricola, 1, 89, 93, 371. , , used for parting, 370.
Alaska Treadwell Mine, 239, 286, 463. Apron plates, 119.
Albertus Magnus, 371. Arborescent gold, 8, 34.
Alchemy, 1. Arnold, J. 0., 19.
Allen, A. H., 448. Arrastra, 90.
Allotropic forms of gold, 11. ,, for prospecting, 93.
Alloys, Assay of, 453. Arsenic alloys* Assay of, 453.
Crystalline, 13, 14. ,, in roasting furnace, 231.
Gold, 12. Artificial crystals of gold, 9.
Gold and copper, 16. Asbestos filtering cloth, 292, 295.
Gold and silver, 15. Assay by amalgamation, 428.
Scorification of, 453. blowpipe, 407.
Aluminium Alloys, Assay of, 454. bromine, 429.
Amalgam, Cleaning of, 131. cadmium, 452.
,, Composition of, 133. chlorination, 428.
,, Retorting of, 133. colour and hardness, 459.
Amalgamated plates, see Plates, crucible method, 410.
Amalgamated. density, 459.
Amalgamation, 122. electrolysis, 460.
, 5 assay, 428. induction balance. 460.
,, Causes of prevention of, scorification, 424, 453.
144. spectroscope, 459.
,, Designolle process of, 129. ,, touchstone, 458.
, ,
Effect of chemicals on, 124. Fluxes used in, 413.
,, ,, hammering on, furnaces, 410, 436.
140. General charges in, 411.
,, temperature on, materials, Examination of, 423.
116. of base ores, 417.
Methods of causing, 141. bleaching powder, 270.
,, pans, 158. complex materials, 427.
Amalgams, 14. cupel, 423.
,, Assay of, 455. cyanide of potassium, 345.
Ammonium carbonate used in treat- gold bullion, 431.
ing gold ores, 307. ,, Accuracy of, 451.
Ancient rivers of California, 63, 67. ,, Lead used in, 435.
Annealing cornets, 446. ,, Losses of goldin. 447.
,, crucibles, 360. ,, Use of proofs in, 450.
fillets,444. gold eras, -107.
Annual production of gold, 464, 468. Cleaning slag in, 417.
484 -
INDEX.
Assay of gold ores, Cupellation in, Base bullion, 356.

417 ,
Assay of, 430.

,, ,, Mint sweep, 431. , ores, Assay of, 417.
,, purple of Cassius, 431. Bassick mine, 37.
pyrites, 428, 430. Batea, 44.
,, rich sulphides, 425. Battery, Stamp, see Stamp battery.
,, tellurides, 426. Bayly, F. W., 435.
Preliminary, 452. Bazin's centrifugal amalgamator, 129.
Roasting in, '>i6. Beach mining, 61.
Special methods of, 427. Beckmann, 93, 371.
Wet methods of, 457. Becquerel, 3.
Whitehead's method of, 430. Berdan pan, 205
Assay-ton, 411. Berthelot, 335.
Attwood, G. M., 146. Berzelius, 2.
Atwood's amalgamator, 190. Bettel, W., 328, 346.
Aurates, 29. Bibliography, 471.
Auric bromide, 27. Biringuccio, 89, 371.
,, chloride, 20. Black concentrates, 48.
,, ,, Decomposition of ,
23. Blackhawk, Stamp mills at, 195.
,, ,, Volatilisation of, 21. Blake crusher, 96, 222.
oxide, 28. W. P., 9, 35.
Auricyanide of potassium, 28. Blake-Marsden crusher, 97, 208.
Auriferous quartz, Formation of, 32, Blanching, 17.
41. Blanket strakes, 172, 201.
of potassium, 27. Bleaching powder, 269.
Aurocyanide
Aurosilicates, 32. Assay of, 270.
Aurosulphites, 29. Blomfield, J. T., 278, 295.
Aurous chloride, 20. Blowpipe assay, 407.
cyanide, 27. Blue gravels, 65.
oxide, 28.
,, , ,
lead theory, 65.
Austin, W. L., 353. ,, Spur Company, 77, 84.
process, 354.
,, Bone-ash for cupels, 418.
Australia, Battery practice in, 199. ,, to cover bullion, 361.
Gold in, 40, 468. ,, to protect muffle, 418.
Automatic feeding machines, 113. Booming, 53.
Available chlorine, 269. Booth, J., 18.
,, cyanide, 350. Borax for refining bullion, 361.
in crucible assay, 413.
,,
Bagration, 333. ,, in scorification, 424, 454.

Balances for bullion assay, 433. 434, Boss, M. P., 159.
447. system of pan-amalgamation,
Ball, E. J., 415. 162.
mills, 156, 158. Boxes, Puddling, 47.
,, stamp, 151. ,, Sluicing, 49.
Balling, C., 429, 452. Boyle, 4.
Banket, 207. British Isles, Gold in, 38.

Bar mining, Deep, 57. Brittle gold, 18, 363, 364, 403.
Barba, 89. Bromide of gold, 27.
Barrel for cleaning up, 131, 202. Bromination of ores, 299.
,, process of chlorination, 262, Bromine, Action of, on gold, 255.
287. Assay by, 429.
,, ,, Advantages of, 264. Brough,B. H., 94, 326.
,, ,, Amount of chemicals Brown and Griffiths, 412.
used in, 270. Brown, R. E., 67.
, ,
Amount of water u sec Bruckner furnace, 233.
in, 268. Biichner, 2.
History of, 262.
,, Buckboard, 409.
Basalt overlying auriferous deposits Bucyrus steam shovel, 62.
63. Buddie, Cornish, 173.
INDEX. 485
Buddie, German, 171. Chloride of gold, Volatility of, 21,

Buisson, 2. 236.
Bullion, Gold, Assay of, see Assay of ,, of silver, see Silver, chloride
gold bullion. of.
Casting of, 364. Chlorination at Alaska - Treadwell
Composition of, 357, Mine, 286.
381,387,400,402,403. ,, at Deloro, 287.
Definition of, 356. at Golden Reward Mill, 291.
Granulation of, 371. at Mt. Morgan, 305.
Losses of, 366. at Rapid City, 299.
Melting of, 360. at Robinson Mine, 326.
Parting of, 369. ,, Barrel process, 283.
357, 361.
Refining of, Cassel' s process, 283.
Skimming 362.of, Greenwood process, 283.
Toughening of, 363. in California, 284.
Bunker Hill mine, '240, 265. in Carolina, 290.
Butters, C., 257, 259, 260, 284, 311, Julian process, 283.
318, 320, 339, 343. Munktell process, 281.
Newbery- Vautin
" process.
Cadmium, Assay by, 452. 280.
Calaverite, 37. ,, Pollok process, 2SO.
Caledonia Mill, 206. Vat process, 249, 284.
California, Battery practice in, 190. Chlorine, Action of, on gold, 254.
,, Chlorination in, 284. ,, on gold alloys, 255.
,, Deep placers in, 63. ,, ,, on organic matter,
Cam-pulley, 108. 256.
,, -shaft, 106. ,, ,, on oxides of metals,
Cams, J06. 256.
Canada, Gold in, 40. ,, on sulphates, 256.
Capel, 459. ,, ,, on sulphides, 253,
Carnelley, 3. 255.
Carnot, A., 427. ,, Amount required for ores,
Carolina, Chlorination in, 290. 257.
Cassel process, 283. ,, Generation of, 252.
Cassius, Purple of, 2, 26, 33, 34. Parting by, 386.
,, Assay of, 431. Refining by, 403.
Casting ingots, 364. Christy, S. B., 235, 236, 238, 389.
Cement gravel, 54, 65. Chrome-steel, 105.
,, ,, treated by arrastra, 92. Clarkson, T., 188.
,, mill, 82. Classification of ores, 169.
,, pan, 83. ,, of parting processes,
Cementation, 370. 369.
Challenge feeder, 115, 208. Claveus, Gasto, 4.
Chalmers, see Hatch. Clay crucibles, 360.

Charcoal, Precipitating gold by, 277. Clean-up barrel, 131, 202.
Charging scoop for bullion, 360. in cyanide process, 318.
Charleton, A. G., 463. in hydraulicking, 79.
Chaudet, 435, 445, 455, 456. in sluicing, 52.
Checks in bullion assay, 450. of stamp battery, 130.
Chemicals for stamp battery, 124. pan, 132.
Chemistry of cyanide process, 347. Cleaning slags, 417, 425.
,, of oxidising roasting, 229. Clennel, 318, 320, 339.
Chester, 9. Cobalt alloys, Assay of, 454.
Chilian Mill, 90. Cohesion of gold, 3.
China, Gold in, 39. Coinage, Alloys used in, 16.
Chlor-auric acid, 26. Collins, H. F., 354.
Chloride of gold, 20. Colorado, Battery practice in, 1 95.
,, ,, Decomposition of, 23. Colour of alloys, Assay by, 459.
of gold, 1.
,, ,, Melting point of, 23.
486 INDEX.
Composition of bullion, 357, 381, 337. Cost of working shallow placers, 85.
,, ,, from chlorinat ion Coyoting, 43.
process, 357. Cradle, 45.
from cyanide process, 320. Crawford Mill, 156.
,, ,, placers, 358. Creuzbourg, 2.
,, ,, stamp battery, 358. Crinoline hose, 73.
of gold slimes, 310.
,, Cripple creek, 37, 465.
of native gold, 38.
,, Croesus Mine, 209.
Compounds of gold, 19. Crookes, W., 137, 460.
,, ,, Natural, 37. Crosse, A. F., 348.
Concentrates, Amalgamation of, 163, Crown Mine, 214.
190, 437. Crucibles for assaying, 411.
Black, 48. melting bullion, 360.
,, Chlorination of, 284, Size of, 365.
286. Crushing before chlorination, 218,
Cyaniding of, 327. 302.
Grey, 48, 59. ,, ,, cyanide process, 308.
Concentration, 163, 166. ,, in stamp battery, 88.
Concentrator, Centrifugal, 174. Crystalline alloys of gold, 13, 14.
Clarkson & Stanfield's, 189. Crystallisation of copper sulphate,
Duncan, 174. 379.
Embrey, .183. ,, gold, 8, 34.
Frue vanner, 176. , ,
silver sulphate.
Gilpin County, 175. 383, 384.
Golden Gate, 193. Cumenge, 32, 427, 435.
Hendy, 174. Cupellation in assay of bullion, 436.
Liihrig vanner, 184. ,, ,, ores, 417.
Raising Gate, 173. ,, Influence of base metals
Triumph, 184. on, 420.
Conductivity of gold, 4. ,, ,, temperature
Consumption of cyanide, 32 J, 340, on, 440.
347. Cupels, 418, 438.
,, of gold, 469. ,, Assay of, 423.
Copper amalgamated plates, 117. Curtis, A. H., 152. 219, 223.
,, ,, discoloration of, 123. Cyanide of gold, 27.
,, ,, for cement gravel, 82. ,, of mercury, 349, 350.
,, ,, in sluicing, 52, 61. ,, of potassium, action on gold
,, Oxide of, in toughening and other metals,
bullion, 364. 333.
,', sulphate, Crystallisation of, ,, ,, action on salts and
379. minerals, 340.
Cornets, 444. ,, ,, action on sulphides,
,, Silver in, 449. 147, 341.
,, the, 447.
Weighing , , , , Assay of, 345.
Corrosive sublimate, in toughening! ,, ,, Commercial, 316.
gold, 363. ,, ,, Consumption of,
Cost of barrel chlorination, 291, 298, 324, 347.
302. ,, ,, Decomposition of,
cyanide process, 214, 325, 324, 338, 340, 343.
331. ,, in stamp battery,
Munktell process, 282, 283. 124, 325.
parting, 375, 376, 380, 381, ,, ,, Selective action of,
386, 401, 407. 307, 341.
,, Plattner process, 261, 287. ,, ,, Solubility of gold
,, production of gold, 462. in, 336.
,, stamp amalgamation process, ,, ,, Solubility of silver
214. in, 338.
,, working placers, 88. ,, ,, Solubility of vari-
,, ,, deep placers, 87. ous metals in, 335.
INDEX. 487
Cyanide of zinc, 339. Diodorus Siculus, 89.

,, process, 306. Dip sample, 433.
,, ,, Chemistry of, 333. Dissemination of gold, 35.
>, ,, Direct treatment in, Distribution of gold, 34, 38.
325. >, ,, in gravels, 66.
,, ,, Disposal of tailings Ditches in hydraulicking, 72.
in, 316. Dixon, 348.
,, ,, Double treatment in, Dodge crusher, 97.
311, 328. Double discharge in stamp battery,
,, ,, Mac Arthur-Forrest,
see Mac Arthur-For- Double treatment in cyanide process,
rest process.
,, ,, Results of, 331. Dredging for gold, 57.
,, ,, Siemens-Halske,329. Drift mining, 81.
i ,, Sulman-Teed, 351. Drop of stamps, 110.
Cyanogen bromide, 351. Drops in sluice, 51.
Cyclops mill, 158. Dry crushing, 218.
diggings, 54.
Daintree, R., 68, 146. Drying ore, 224.
Dakota, Battery practice in, 206. Dubois, W. E., 35.
,, Bromination in, 299. Ductility of gold, 2.
,,Chlorination in, 291. Dutfield, 143, 183.
Pyritic smelting in, 354.
,, Duflos, 215, 262.
Roasting in, 244.
Dams, in river mining, 55. Egleston, T., 68, 140, 367.
D'Arcet, C., 375, 435. Eissler, 284, 353.
Davis, W. M., 277. El Callao Mine, 463.
De Lacy, 262. Electricity in amalgamation, 128.
De Morveau, Guyton, 2. Electrolysis, Assay by, 460.
Dead-leaf gold, 15. ,, Parting by, 404.
Debray, 20, 34. Electrum, 15.
Decomposition in zinc boxes, 339. Elephant stamp, 151.
of gold chloride, 23. Elevator, Hydraulic, 58, 83.
,, of potassium cyanide, Elkington, 333.
324, 338, 340, 343. Eisner, 333.
Deep bar mining, 57. Embrey concentrator, 183.
,, placer deposits, 62. Emmons, 35.
,, ,, Method of working, Endlich, F. M., 307.
70. Equal falling particles, 167.
Deetken, G. F., 190, 194, 216, 250, Esson, 24.
255, 262. Eureka rubber, 190.
Del Mar, A., 42. Europe, Gold in, 38.
Deloro, Mears process at, 287. Expansion of gold, 4.
Dendritic gold, 8, 34. Extra solution, Russell's, 30.
Dennes, D., 222. 273, 293, 299, 301.
Density, Assay by, 459. Faraday, 2, 333.
of gold, 3. Farrar, S. H., 207.
,, of alloys of gold, 16. Feather river, 56, 80.
Designolle process, 129. Feeding in stamp battery, 113.
Desmarest, 2. Feix, J., 367.
Detection of gold in alluvium, 44. Feldtmann, 340.
ores, 426. Ferry, N. A., 260.
solution, 26, 427. Figuier, 2.
Deville, 4. Filter beds for leaching, 249, 251,
D'Hennin, 457. 292, 295, 310.
Dibromide of gold, 27. press, 320, 323.
Dies, 105. Fineness of bullion produced by
Diffusion of gold, 14. Miller's process, 402.
Dingier, 430. Finkener, 346.
488 INDEX.
Flashing, 419, 439, 440. Gilpin County, Battery practice at,

Float gold, 142, 143. 195.
Flouring of Mercury, 136. ,, ,, concentrator, 175.
Flumes for hydraulicking, 72. Godshall, L. D., 302.
Flu viatile theory of Calif ornian placer Gold, Action of chlorine on, 255.
deposits, 62. ,, ,, of potassium cyanide
Fluxes in assaying, 412, 413. on, 336.
,, in pyritic smelting, 354. ,, bullion, see Bullion, Gold.
,, in refining bullion, 362. Detection of, see Detection of
Fly catchers, 55. gold.
Foliated tellurium, 37. ,, Losses of, see Loss of gold.
Formation of auriferous quartz, 32. ,, ores, Assay of, 407.
, ,
of nuggets, 68. ,, Properties of, 1.
Forms of gold found in nature, 34. ,, residues, Treatment of, 373,
Forrest, W., 307. 278, 383, 385.
,, R. W., 307. ,, slimes, 319.
Foster,Le Neve, 116. ,, Treatment of, 320.
Foundations of stamp battery, 103. ,, Solubility of, 9, 31, 255, 336.
Framework of 103.
stamp battery, Golden Gate concentrator, 193.
Freezing point of gold, 3, 12. Reward Mill, 291.
Fremy, 427, 430, 435, 448, 454, 459. Gooseneck in Mears process, 263.
Fresenius, 342, 346. Gore, G., 334, 335, 404.
Fromm, 13. Grade of plates, 125.
Fruevanner, 176. sluices, 50, 59, 78.
Riffled belt for, 183. Granulation of bullion, 371, 375.
Fuchs, 427, 435. Graphic tellurium, 37.
Fulminating gold, 29. Graphite crucibles, 360.
Furnaces, Assay, 410, 436. Gravel, Cemented, 54.
Bruckner, 243. ,, Method of washing, 44.
Chlorination, 393. Gravels, Blue and red, 65.
Gas Assay, 437. Gravitation stamps, 102.
Hofmann, 244. Green gold, 15.
Mechanical, 238. Greenwood process, 283.
Melting, 359. Grey concentrates, 48, 59.
Muffle, 436. Griffiths, H. D., 185.
O'Hara, 238. Grizzly, 51.
Pearce Turret, 240. Grommestetter, 93.
Refining, 359. Ground sluicing, 53.
Reverberatory, 226. Groves, J., 438.
Revolving cylindrical, 243. Griiner, 226.
Rotating-bed, 240. Guides in stamp battery, 109.
Shaft, 238. Giiptner, 452.
Spence, 239. Gutters in ancient rivers, 66.
White, 246. Gutzkow, F., 381, 383, 386.
White-Howell, 247. ,, process, 381.
Fusibility of gold, 3. ,, ,, New, 383.
Future production of gold, 466.
Haile Mine, 129, 290.
Haindl, 456.
Gas, Chlorine, method of production, Hall-marking, 15.
252. Halogen compounds of gold, 20.
,, ,, used in Plattner pro- Hammond, J. H., 191.
cess, 252, 254. Hankey, J., 144.
Roasting by, 248. Hanks, 362.
Gates crusher, 98, 208. Harcourt, 24.
Geber, 350. Hardness of alloys, 459.
Geographical distribution of gold, 38. cfgold, 2.
Geological 35. Hare, R., 4.
Gernet, von, 330. Harris, 91.
INDEX. 489
Harscher, C., 93. Kandelhardt, 435.

Hartz jigs, 187. Kasentseif, 14.
Hatch and Chalmers, 210, 212, 213, Kedzie, G. E., 340.
311, 325, 328, 463. Keith, N. S., 348.
Heat of formation of chlorides, 20 Kennel, California, Chlorination
,, of cyanides, 335.
,, works at, 284.
Hendy's challenge feeder, 115. Kerl, B., 216, 448.
Heycock and Neville, 12, 13. Kerr, Prof., 35.
Hickock, 244. Knox pan, 132, 190.
Hofmann furnace, 244. Kohlrausch, 11.
Holleman, 11. Kraift, 9.
Homburg, 4. Krom, S. R., 188, 219.

Homestake Mill, 206. rolls, 219.
Hood, J. J., 350. Kriiss, 2, 4, 24, 25, 28, 33.
process, 349. Kumara tail sluice, 54.
Horn spoons, 45. Kunckel, 375.
Howe, H. M., 232. Ktistel, G., 226, 229, 235, 255.
Howell- White furnace, 246.
Huggins, Dr., 3.
Hughes, Prof., 460. Lang, H., 353, 355, 356.
Huntington Mill, 152. Lange, 215.
Husband's Stamp, 148. Langguth, VV., 275.
Hydrates of gold, 28. Langlaagte estate, 317.
Hydraulic elevator, 58, 83. Latent heat of gold, 3.
,, sizing boxes, 170. Latta, 146.
Hydraulicking, 56, 70. Laurie, 4, 13.
Hydrocyanic acid, 346. Lava above auriferous gravels, 63.
Hydrogen amalgamator, 164. Law, R., 395.
Hydrolytic decomposition of potas- Le Chatelier pyrometer, 441.
sium cyanide, 338. Leaching by decantation, 271.
Hyposulphites of gold, 30. pressure, 273, 294, 300.
vacuum, 272, 280.
Inch, Miners', 73. Difficulties of, 272.
Independence Mill, 293, 295. in chlorination barrel, 293.
India, Gold in, 39. ,, cyanide process, 309.
Induction balance, 460. ,, with agita-
Ingots, Casting, 364. tion, 314.
Inquartation, 372, 421. in Plattner process, 258.
Iodide of gold, 27. vats in chlorination pro-
Iridium alloys, Assay of, 456, 457. cess, 249.
Iridosmine, 69. ., ,, cyanide process,
Iron alloys, Assay of, 454. 309.
,, pipes in hydraulicking, 72. Lead, amount used in bullion assay,
,, riffle bars, 77. 435, 436.
Sulphate of, used to precipitate ,, ,, reduced in ore assay,
gold, 259, 274, 286, 287. 413.
,, vessels for parting, 374, 381. ,, lining for chlorination barrel,
268.
Jacob, 464, 469. Leibius, A., 5, 389.
Janin, Jun., L., 18, 129, 146, 147, Lever or jack in stamp battery, 109.
307, 337, 342. Levol, 15, 31, 33, 419.
Japan, Gold in, 39. Lime used in cyanide process, 344.
Jevona, 470. Lindet, 20.
Jigs, Hartz, 187. Liquation of gold alloys, 17.
,, Pneumatic, 188. Liversedge, 69.
Johnson's filter press, 320, 323. Lockyer, 459.
Jordan's amalgamator, 165. Long Tom, 46.
Julian, 283, 310. Loss of gold by volatilisation, 367.
,, process, 283. ,, ,, in amalgamation, 138.
490 INDEX.
Loss of gold in assaying, 447. Mercury cyanide, Use of, in cyanide

,, ,, ,, refining, 366. process, 348, 350.
,, ,, roasting, 235. ,, Flouring of, in stamp bat-
Loss of mercury, 203. teries, 136.
Lossen, 10. ,, Loss of, in stamp battery,
Louis, H., 3, 89. 137.
Lowe, 23. ,, Sickening of, 136.
Liihrig vanner, 184, 208. Use of, 14, 79, 89.
,, ,, in sluicing, 52.
MaeArthur, J. S., 307, 308, 316, ,, wells in stamp batteries,
320, 336, 341, 342, 347. 128, 142, 201, 206.
MacArthur-Forrest process, 307-328. Merrick, J. M., 428.
At the George and May, 324. Merrineld Mine, 237.
At the New Primrose, 328. Metallics, 409.
At the Sylvia mine, 326. Methylamine, 339.
Composition of bullion from, 320. Mexico, Gold in, 40.
Consumption of cyanide in, 329, Miers, 89.
347. Mill, Site for, 119.
Cost of, 331. Miller, F. B., 387.
Direct treatment of ore in, 324. ,, process, 386.
in S. Africa, 331. Minerals in placer deposits, 69.
Leaching in, 311. ,, occurring with gold, 36.
Ores suitable for, 352. Miner's inch, 73.
Plant required in, 322. ,, pan, 44.
Production of bullion in, 319. Mint, Method of assaying in, 432.
Results of, 331. ,, sweep, Assay of, 431.
Strength of solution for, 315, 337. Miocene Ditch Company, 72.
Treatment of concentrates by, 326. Mitchell, 411, 420, 428.
M'Cutcheon, J., 133, 401. Modderfontein Mine, 209.
M'Dermott, 143, 169, 182, 183. Mode of occurrence of gold, 34.
M'Dougal, G., 142. Moebius, B., 404.
Maclaurin, J. S., 333, 337. ,, process, 404.
Mactear, 145. Moldenhauer, 334.
Magnetism of gold, 3. Molloy's amalgamator, 164.
Makins, G. H., 5, 447. ,, method of precipitating gold
Maldonite, 38. cyanide, 321.
Malleability of gold, 2. Monitor, 74.
Mallet, 4. Monochloride of gold, 20.
Maltitz, Count von, 93. Mortar for crushing gold ores, 89.
Management of gold mills, 462. ,, stamp battery, 103.
Manganese alloys, Assay of, 454. Morton, Prof., 147.
Maray, 89. Mould for ingots, 364.
Materials for crushing surfaces, 100. Mt. Morgan Mine, 38, 305.
Mathison, 375. ,, Precipitation of gold
,,
Matthey, E., 7, 19, 454, 456. at, 277.

Matthiessen, 13. ,, Roasting at, 145,272.
Mattison, Edw., 70. Muffle furnace, 436.
Maynard, G., 405. ,, Temperature of, 441.
Mears, Dr. H. 262, 263.
, Miihlenberg, 307.
,, process, 263, 287. Miiller, 24/30.
Mechanical furnaces, 238. Munktell process, 281.
Medina, 89. Muntz metal plates, 126.
Mein, 311. Mylius, 13.
Meinecke, 113.
Melting gold bullion, 360. Nagyagite, 37.
Mercur Gold Mine, 308, 312. Napier, 4, 5.
Mercury, 14, 125. Native crystals of gold, 8.
Chloride of, used to toughen gold, 38.
gold, 363. Natural compounds of gold, 37, 38.
INDEX. 491
Natural forms of gold, 34.
Parting by Gutzkow process, 381. '
Neville, see Heycock. new 383.

New Primrose Mine, Cyanide pro- ,, ,, nitric acid, 371.
cess at, 328.
,, ,, sulphide of antimony,
New South Wales, Gold in, 41. 370.
New Zealand, Battery practice in, ,, ,, sulphur, 370.
199. ,, ,, sulphuric acid, 375.
Gold in, 40. ,, flask, 422.
Newbery, J. Cosmo, 136, 186, 262, ,, in bullion assay, 444.
280. ,, ,, ore assay, 421.
Newbery- Vautin process, 280. ,, ,, platinum tray, 445.
Nickel alloys, Assay of, 454. ,, ,, porcelain crucible, 423.
Nickles, 10. ,, ,, test tube, 423.
Nitric acid, Action of, on metals, 373. ,, processes, Classification of,
,, Parti ug by, 371. 369.
Solubility of gold in, 447. Patio process, 90.
North Bloomfield Mine, 71, 79. Pearce, 14, 15, 17.
Nuggets, 36, 69. Peel, 446.
,, Formation of, 68. Peligot, 17.
Pelouze, 435, 448.
Occurrence of gold in nature, 34. Pepys, 432.
Percussion tables, 175.
O'Driscoll, F., 262.
O'Hara furnace, 238. Percy, J., 215, 277, 370, 376, 381,
Ore bins, 121. 401, 412, 417, 425, 449.
Ores, Assay of, 407. collection, 9, 14, 35, 277.
Pestarena Mine, 92.
Origin of deep placer deposits, 62.
Pettenkofer, 435.
,, gold in gravels, 67.
in ores, 41. Petzite, 37.
,, ,,
Osmiridium, Separation of, from gold, Philadelphia Mint, Liquation at, 18.
368. ,, ,, Parting at, 380.
Oxides of gold, 28. Philippe of Valois, 432.
Phoenix Mine, Chlorination at, 290.
Oxidised pyrites, Action of potassium
Pi DOS Altos, Treatment of bullion at,
cyanide on, 343. 406.
,, ,, Amalgamation of,
Placer deposits, Deep, 62.
146.
Shallow, 42.
Oxidising roasting, 229.
gold, 36, 66, 68.
Oxygen, action on dissolution of gold
,, mining, Cost of, 41.
by potassium cyanide, 333, 337, 117.
348. Plates, Amalgamated,
Copper, see Copper
plal
plates.
Palladium alloys, Assay of, 456. Corrugated, 128.
Pan-amalgamation, 159. Grade of, 125.
Pan, Boss, 162. in sluicing, 52, 61.
,, Knox, 132. in Transvaal, 210.
,, Patton, 161. Muntz metal, 126.
,, Siberian, 60. Preparation of, 117.
Washing by the, 42. Shaking, 127.
Paracelsus, 1. Treatment
of, 122,
Parting, 369. 133, 211.
assay, 431. Platinum alloys, Assay of, 455.
by sulphuric acid, , , Tray for parting, 445.
451. Plattner, C. F., 215, 235, 407, 428.
,, double, 452. ,, process, 215.
History of, 432. ,, Amount of water used
by cementation, 370. in, 251.
,, chlorine, 386. ,, at Reichensteip, 216.
,, combined process, 380. ,, ,, at various mills, 284-
,, electrolysis, 404. 287, 326.
492 INDEX.
Plattner process, Cost of, 261, 287. I

Proportion of gold and silver for part-
,, ,, Method of working,249. ing, 372.
,, ,, Reactions in, 255. ,, ,, to gaugue in ores,
,, ,, Vats used in, 249. 37.
Pliny, 89, 370. Prospecting trough, 45.
Pliocene gravels of California, 63. Protobromide of gold, 26.
Plymouth Consolidated Mill, Chlor- Proust, 2.
ination at, 285. Prussian blue in cyanide process, 347.
Pneumatic jig, 188. Puddling tub, 46.
Poggendorf, 334. Pare gold, Preparation of, 10.
Pointed boxes, 170, Purple alloy, 13.
Pollok chlorination process, 280. ,, gold, 2.
Poussee process, 362. ,, of Cassius, 2, 26, 33.
Power, F. R., 392. ,, ,, Assay of, 431.
Prat, 20. Pyrites, Assay of, 430.
Precipitation of gold chloride, 24, ,, Condition of gold in, 145.
273. Pyritic smelting, 353.
,, by charcoal, 277. Austin process, 353, 354.
, ,
, by iron sulphate, 259, 274, Pyrometer, Le Chatelier, 441.

286, 287.
,, by metallic sulphides, 278. Quartz crushing in stamp battery,
by metals, 279. 88.
,, by organic substances, Quicksilver, see Mercury.
275.
,, by sulphuretted hydrogen, Rae, J. H., 306.
275, 289, 276. Randall, P. M., 73, 117.
,, by sulphurous acid, 11, 296. Rapid City Mill, 222, 299.
,, Cost 303.
of, Raymond, 172.
,, ,, ,, Use of mo- Reactions in chlorination vat, 255.
lasses in,260. Recovery of gold lost in refining, 366.
,, of gold alloys, 13. Reduction of silver chloride by zinc,
,, of gold cyanide by sodium, 374, 391.
322. by iron, 392.
,, by electricity, Reduction of silver sulphate by char-
330. coal, 385.
,, ,, ,, by zinc, 3 17,340. by copper, 379.
,, ,, ,, Influence by iron, 383.
of temperature by sulphate of iron, 383.
on, 340. Refining of gold bullion, 357, 361.
,, ,, in leaching vats, 345. by borax, 361.
,, , ,
from solution, 24, 68. by chlorine, 386, 403.
,, of silver by salt, 374. by nitre, 361.
,, ,, by copper, 379. by sulphur, 368.
Pressure leaching, 273, 294, 300. Regnault, 3.
Price, T., 402. Reichenstein, Plattner process at,
Primitive methods of crushing quartz, 216.
88. Reservoirs for hydraulicking opera-
Prinsep, James, 441. tions, 72.
Production of gold, Causes of increase Retorting, 133.
of, 465. ,, furnace, 134.
Cost of, 462. Reverberatory furnace, 226.
in the world, 464. Reynolds, J., 432.
331.
by cyanide, Rhodio-platinum couple, 441.
in the future, 466. Rhodium alloys, Assay of, 456.
,, past, 464. Richards, J., 224.
in various Rickard, T. A., 198, 200.
countries, 467. Ricketts, 417, 424, 426.
Proofs in bullion assay, 450. Riffle bars, 49, 59, 77.
Properties of gold, 1. blocks, 76.
INDEX. 403
Riffled belt in Frue vanner, 183. Scheidel, A., 326.

sluice, 173. Schonfeld, 371.
Rim-rock, 66. Schranz crusher, 99.
Riotte, E. N., 273. Schwartz's assay of pyrites, 438.
River mining, 55. Scorification, Assay of alloys by, 453.
Rivot, 415, 419. 5J ,, ores by, 424.
Roasting at Deloro, 284. Screens, 110.
at Kennel, California, 284. Selective action of cyanide, 307,
at Treadwell Mine, 286. 341.
by producer gas, 248. Selwyn, A. C., 68.
Cost cf, 291, 298, 302. Settling boxes, 168.
furnaces, see Furnaces, Shaft mining, 81.
roasting, Shaking copper plates, 127.
ores, 225. Shallow placer deposits 42.
,,for assay, 416. Sheba Mine, 211.
oxidising, Chemistry of, 229. Shovel, Steam, for placer gravels,
with salt, 233. 61,
Roberts - Austen, 3, 10, 13, 14, 17, Siberia, Gold in, 39.
226, 229, 403, 436, 440, 449, 459, ,, Method of washing in, 47,
460. 58.
Robinson Mine, 214, 312, 317, 326. Siberian pan, 60.
Rock-breakers, 96. ,, sluice, 59.
Rock-cut sluices, 78. ,, trommel, 60.
Rock pavements for sluices, 76. ,, trough, 47.
Rocker for washing gold, 45. Sickening of mercury, 136.
Roller feeder, 116. Siemens-Halske process, 329.
Rolls for crushing, 219. Silicate of gold, 32.
,, compared with stamps, 223. Silver chloride, Reduction of, 374, 391.
Resales, H., 41, 169. ,, Compounds of, dissolved by
Rose, G. 3, 38., potassium cyanide, 342.
Rose, H., 346, 456, 458. ,, dissolved by nitric acid, 373.
Rose, T. K., 4, 5, 18, 23, 26, 441, potassium cyanide,
449, 450. 338.
Rossler, 448, 449. ,, ,, sulphuric acid,
Rotary drying furnace, 225. 377, 381, 384.
Rothschild's refinery, 375. filter,385.
Rothwell, J. E., 221, 247, 288, 291, for gold bullion assay, 434.
298. Gold separated from chloride
Royal College of Science, 324, 415. of, 389.
Royal Mint, Bullion assay at, 432. in cornets, 449.
Cupels at, 438. ,, gold coins, 15.
Ruby gold, 2. Simmer & Jack, cyanide works, 312.

Russell process, 31. Simpson, J. W., 306.
Russia, Gold in, 39. Sizing boxes, 170.
,, Placer mining in, 47, 58. Skey, Wm., 68, 144, 349, 426.
Rusty gold, 144. Skimming bullion, 362.
Slime presses, 295.
St. John del Key Mine, 172. ,, separator, 311.
Sal ammoniac used to toughen Slimed ore, Treatment of, by cyanide,
bullion, 362, 363. 323.
Salamander crucibles, 360. Slimes, Gold, 319.
Salt used in ore assay, 414, 416. Sluice plates, 119.
,, roasting, 233. Sluicing, 49, 76.
Sampling cone, 408. Smith, E. A., 36, 421.
of bullion, 432. Soda, Caustic, use in amalgamation,
ores, 408. 125.
,,
,, tin, 408. ,, cyanide process,

344.
Sandberger, F. von, 41.
Savot, 372, 432, 435. Sodium amalgam, 137.
494 INDEX.
Sodium carbonate, used to separate Stock of gold, 470.

gold from silver chloride, Stone-breakers, 96.
389. Strength of cyanide solution, Method
,, hyposulphite, Action on gold of testing, 345.
of, 31. Strength of cyanide solution required
,, used to precipitate gold for ores, 347.
cyanide, 321. Sulman, H. L., 351.
Soetbeer, 465, 470. ,, -Teed process, 351.
Solubility of gold, 9, 31, 349, 447. Sulphate of copper, Crystallisation
,, ,, in potassium cyanide, of, 379.
333, 336. Sulphide of gold, 30, 31, 32, 33.
,, of silver chloride, 11. Sulphides, Action of potassium
sulphate, 377. cyanide on, 147, 341.
sulphates, 378. ,, Amalgamation of, 145.
,, ,, various metals in po- Assay of 425.
tassium cyanide, 335, Gold in, 145.
336, 338. Oxidation of, 343.
Solutions of gold, Test for, 25. Roasting of, 227.
Solvents of gold, 9, 31, 447. Smelting of, 353.
Sonstadt, 35. used as fuel, 354.
South America, Gold in, 40. Sulphites of gold, 29.
Clunes United Mill, 201 Sulphur as fuel, 354.
Spanish Mine, 155. in parting, 370.
,,
Specific gravity of gold, 3. ,, in refining, 368.

,, heat 3. Sulphuretted hydrogen as precipitant
Spectroscope, Assay by, 459. for gold, 275, 296.
Spectrum of gold, 3. Sulphuric acid, Action on metals of
Spence furnace, 239. 372.
Splash box, 105. Assay by, 455.
Spring, 10. ,, Parting by, 375.
Stamp battery, 82, 88, 93, 102. Sulphurous acid used to destroy
,, ,, Chemicals used in, 124. chloride, 275.
Foundations of, 103. Surcharge, 447.
Framework of, 103. ,, Effect of base metals on,
,, General arrangement 453.
of, 119. ,, Influence of temperature
,, practice, 190. on, 443, 449.
Calif ornian, 102, 108. ,, Variations in, 448.
Elephant, 151. Sweating copper plates, 133.
German, 94. Swedish chlorination process, 281.
Husband's 148. Sweep, Mint, Assay of, 431.
Steam, 149. Swinging amalgamated plates, 119.
of parts of, 109, 210.
Weight Sydney Mint, Parting at, 387.
Stamps, Order offall of, 113. Sylvanite, 37.
, , Weight of, in California, 191. Sylvia Mine, Cyanide process at, 326.
Standard gold of various countries,
16. Tail race, 53.
Stanfield, 188. ,, sluices, 54.
Stanford's self feeder, 114. Tailings, Examination of, 143.
Stansfield, A., 455. ,, from hydraulicking, 80.
Stapff's method of assaying pyrites, Tappet, 106.
430. Teed, Dr., 351.
Steam shovel for placer gravel, 61. Tellurides of gold, 37.
,, stamp, 149. ,, Assay of, 426, 457.
Stelzner, A., 41. Tenacity of gold, 2.
Stetefeldt, C. A., 164, 225, 235, 315, Test for gold in solution, 25.
320. Testing ores by chlorine, 429.
,, drying kiln, 225, 249. ,, ,, cyanide, 323.
furnace, 238. ,, solutions of gold, 259.
INDEX. 495
Testing toughness of bullion, 363. I

Volatilisation of gold, Effect of tem-
Thames Valley, N.Z., Battery prac- perature on, 5.
tice in, 203. Volatility of gold chloride, 21, 236.
Thies, A., 262, 265. Voltaic order of metals in potassium
process, 265, 280. cyanide, 334.
,, ,, Cost of, 291.
Thiosulphates of gold, 30.
Thompson, L., 386. Wales, Gold mining in, 38.
Thomsen, 20, 335. Washing gravel by the pan, 44.
Thorpe, 4. ,, ,, in sluices, 50.
Tin alloys, Assay of, 454. Water supply for hydraulicking, 71-
Tools used in assaying, 414, 418, 438. ,, used in barrel chlorination,
Touchstone, 458. 268.
Transvaal, Cyanide process in, 331. ,, ,, cyanide process, 322.
,, Production of gold in, ,, ,, sluicing, 50.
40, 214, 468. ,, stamp battery, 116 r
,, Stamp amalgamation in, 210.
207. ,, vat chlorination, 251.
Trapiche, 90. Watson-Denny pan, 205.
Tribromide of gold, 27. Watts, 346.
Trichloride ,, 20. Wear of screens, 112.
Triumph concentrator, 184. ,, shoes and dies, 105.
Trommel, Siberian, 60. Weight of parts of stamps, 109, 210.
Trough, Prospecting, 45. Weights, Assay-ton, 411.
Siberian, 47. , ,
for bullion assaying, 433.
Tub, Puddling, 46. Weinberg, 305.
Tulloch feeder, 116. Wells, J. S. C., 340.
Tunnels in drift mining, 81, 82. Welman dredge, 57.
Turret furnace, 240. Whisk brooms for copper plates, 124.
White furnace, 246.
Undercurrents, 51. White-Howell furnace, 247, 292.
United States, Gold in, 40. Whitney, 63, 69.
Ural Mountains, Gold mining in, 39. Wiebe, 4.
Wilm, 12.
Vacuum pump for dredging, 57. Wingdams, 55.
Van Riemsdijk, 420, 440. Witwatersrand, see Transvaal.
Vanner, Frue, 176. Worcester mill, 329.
,, Liihrig, 184. World's production of gold, 464.
Van't Hoff, 23.
Vat process of chlorination, 249.
Vats for leaching processes, 249, 309. Zinc, Action of potassium cyanide
Vautin, C., 262, 274. on, 339.
Victoria, Gold Mining in, 41. ,, alloys,Assay of, 454.
Violle, 3, 5. ,, Assay by, 452.
Volatilisation of alloys of gold, 5, 7. , Precipitation of gold cyanide
of gold, 4, 367, 447. by, 317, 340.
Effect of base metals Reduction of silver chloride
on, 5, 7. by, 374, 391.
BELL AND BAIN, LIMITED, PRINTERS, GLASGOW.

RETURN CIRCULATION DEPARTMENT
TO--* 202
LOAN PERIOD
HOME USE
YC 63559
V i
V,
m

The Metallurgy of Gold (IA Metallurgygold00roserich)

Uploaded by

Copyright:

Available Formats

The Metallurgy of Gold (IA Metallurgygold00roserich)

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Metallurgy of Gold (IA Metallurgygold00roserich)

Uploaded by

Copyright:

Available Formats

THE LIBRARY

W. C. ROBERTS-AUSTEN, C.B., F.R.S.,

In Large Svo, Handsome Cloth. With Illustrations.

1. INTRODUCTION to the STUDY of METALLURGY.

2. GOLD (The Metallurgy of). By T. KIRKE ROSE, D.Sc.,

3. COPPER (The Metallurgy of). By THOMAS GIBB, Assoc.

4. IRON and STEEL (The Metallurgy of). By THOMAS

particularly rich in chemical details. ...

5. METALLURGICAL MACHINERY : the application of

6. ALLOYS. By the Editor.

%* Other Volumes in Preparation.

T. KIRKE BOSE, D.Sc.,

BEINO ONE OF A SERIES OF TREATISES ON METALLURGY,

ROYAL SCHOOL OF MINES.

fcrot TO. (L 1Roberts*Busten, <r.., JMR.S.

WITH NUMEROUS ILLUSTRATIONS.

[All Rights Reserved.]

IN preparing this edition, the whole book has been care-

fully revised, due attention being paid to some useful

suggestions contained in the various reviews of the first

edition. Eight new . illustrations have been added, in-

cluding the frontispiece, which is from a photograph kindly

lent by Messrs. Fraser & Chalmers. The rapid progress

necessary to make considerable alterations in, and additions

to, the chapters devoted to that subject, and a new chapter

on Economic Considerations has also been added. The

author desires to express his thanks to the old students

of the Koyal School of Mines and others, who have kindly

As this is the first of a series of treatises devoted to indi-

vidual metals, which is being prepared under ray guidance

it may be well to offer a few introductory remarks.

general account of the Metallurgy of Gold and adequately

refining of bullion and to assaying. A large amount of

incorporated in the proceedings of various learned

Any attempt, however, to study the subject as

these sources of information would be hopeless for those

Mr. Rose gained his practical experience of gold and

written a volume which should prove to be very useful,

ROYAL MINT, March, 1894.

IN the present volume an effort has been made to supply

metallurgy of gold, for the use of students and others

precious metals. Ithas been said that mill-managers as

be so, they will still need to keep themselves acquainted

Although, in this book, brief accounts are given of some

The practice in particular extraction works has been

given. Some recently devised methods of great importance

cyanide process, the new barrel chlorination process, and

paid to the assay of gold bullion, the system in use at

the highest attainable accuracy being described. Little

space is devoted to the geographical and geological dis-

smelting, leaching, and pan-amalgamation used in the

as belonging to the metallurgy of silver.

advice throughout 'the progress of the work. It was

ishoped that it will contribute to the realisation of his

CHAPTER I. THE PROPERTIES OF GOLD AND ITS ALLOYS.

CHAPTER III. MODE OF OCCURRENCE AND DISTRIBUTION OF GOLD.

CHAPTER IV. PLACER MINING SHALLOW DEPOSITS.

52 New method of working shallow

CHAPTER V. DEEP PLACER DEPOSITS.