US1025932A - Means for producing light. - Google Patents

Description

c. P. STEINMETZQ HEANS FOR PBODUOING LIGHT. APPLICATION FILED APB. 11,l1902.

Patented May 7, 1912.

4 SHEETS-SHEET 1.

v lnoenbor: Charles PSLeinmeCz,

bk, MMM@ 2m' 04P. STEINMETZ. MEANS POR PBODUGING LIGHT.

APPLICATION FILED APB.. 11l 1902.

Patented May 7, 1912,

4 SHEETS--BHBET 2.

lnverlbor".

. by I C. P. STEINMETZ.

MEANS POR `PRODUGING LIGHT.

APPLIGATION FILED un. 11, 1902.

1,025,932. j mummy 7,1912.

5 4: SHEETS-SHEET 3. N 6 N E: .5

Charles lo. Steinmetz,

muy.

o. P. STEINMETZ. l MEANS ron PBonUcING LIGHT.

APPLICATION FILED IPR. 11, 1902.

, 4 Simms-sanza: 4.

Fig. 9.

(Ni/Chasses:

lnvenbor: Charles @.Sbeinme'z.,

jung.

atena )Stay 7, 1912.

UNITED sTATEs PATENT oiaEIcEf4 l.

CHARLES r. sTEINMETz, or scHENEcTADY, NEW YORK, AssIGNon To GENERAL Emeal i Tmc coMrANY, A conoEATI'oN or NEwvYonx.

MEANS FOR PRODUCINGLIGHT.

To all inkom t may concern Be it known that I, CHARLES P. S'rEiN- METZ, a citizen of the UnitedStates, residing at Schenectady, in the county of Schebration of the ultimate particles of vapor in i the are stream, which vibration is probably.

caused by the electric stress due to the potential gradient ofthe are. It is found that the color of the light, or, more -broadly speaking, the wave length ofthe radiant energy derived from an arc, is practically independent olf temperature, and dependent principally on the nature of the conducting vapor which forms the arc stream. The vibrations therefore appear to be, as it were, sympathetic, and may take place very considerably below the temperature of incandescence.

In general, the temperature of an arc is fixed by the nature ofthe electrode, just as the temperature of any; boiling liquid is fixed bythe nature of the most volatile component of that liquid, It is common in practice to use electric 'arcs bet-Ween carbon electrodes for commercial lighting. In carbon arcs the major portio'nof the-light comes from the craters, so that it is perhaps not strictly correct. to speak -ofthese devices as arc lamps.Y This incandescence is the result Iof the very high temperature of the carbon are, but the high temperature causes very heavy radiation losses.

The temperature of the carbon arc stream is Xed by the volatilization temperature of carbon, as is the temperature of any t-rue vapor, and this temperature is such that a very large proportion of the energy from a carbon arc is radlatedin the form cf dark,

Specification of Letters Patent.

Original application led `March 5, 1900,Serial No. 7,265. Divided and this'a'pplication led f w 1902. Serial No. 102,415.

Patented May '7, 1912" Aprn n',

heat Waves, so that the carbon arc becomes an eXtremely-ineiicient means of converting electric energy. into light. lI find it highly advantageous in my new method of producing light to avail'myself of substances vaporizing at very low temperatures, and by using such substances I am able to greatly reduce the radiation losses, and to correspondingly increase the efficiency.

It is obviously necessary'for the maintenance of the arc that the vapors from the electrode should be conductors of electricity, and it 'is therefore necessary to use as an electrode .a substance producing conducting vapors. This requisite is fulfilled by mercury, and further, mercury is a subst-ance vaporizing at very-low temperature. Hence it appears thatan are between electrodes, at leastv one. of which is formed of mercury,`

should be an extremely efiicient light-giving arrangement, and this is indeed the fact, but the mercury arc is, unfortunately, of an extremely disagreeable color., It gives a discontinuous spectrum containing the Fraunhnefer lines 4047; 4359; 5461; 5769 and 57 90, and some fainter intermediate lines. The sodium line, 5890, sometimes appears faintly,vbut this is probably due to the action of theglass or the presenceof some impurity.` It will be seen'thatall of these lines' are either violet, blue' or green, and that the spectrum is practicallydevoid of yellow, orange and red. As a result of this deficiency, red objects appear black or gray when viewed by light from a mercury are, and human faces acquire a peculiar and disagreeable pallor, so that the lightfrom the mercury arc seems to be utterly unsuitedy to commercial application.`

It is one object of my invention to so Amodify .the light from the mercury arc as to produce a soft, brilliant, white light with very low radiation losses, and with an efficiency rising as high as three to five candle-power per watt; whereas ordinary incandescent lamps, with which my improved lamp may properly be compared, y require about three watts per candle-power, while even the carbon arc, which is in many ways unsuited to interior illumination. requires in general about one watt per candle-power.

Further, my invention is not necessarily restricted to the production of a pure White light, but I am able to so modify the light of the mercury-arc stream as 'to cause almost any desired wave length to strongly prepon-V from this lamp by passing the light. through ,red glass. It is well known that the action of such red'glass is simply to screen out the virtue of my improvements it is larger proportion of the violet, blue, green and yellow rays, and to allow only the red rays to pass. rlhis, obviously, takes away a verylarge percentage of the radiant energy generated, and greatly reduces the efficiency of an already inefficient lamp. By l possible to generate once for all red light, -so that a very large proportion of the radiant energy actually issues from the lamp in the form of red rays, and the same principle may be applied to the production of light of any desired color. Nor is my invention necessarily restricted to the use of mercury,

. as certain other substances fulfil fairly Well the conditions above set forth, though I have thus far found mercury best adapted for practical use. Y

My improvement in modifying the color of an electric arc consists, broadly speakin in introducing conducting vapors of sustances whose are spectra have the desired Wave length. This, however, is not as simple in practice as the above statement, for itl seems essential that electric arcs, particularly from mercury, should be inclosed in glass tubes or vessels. I will discuss below the starting of the arc` One method of v starting is to have an initial high voltage in combination with a lamp exhausted to a high vacuum. Obviously, for such a lamp as this, glass is the' only suitable material,

and even in lamps which are started much as is the ordinary carbon are, it is highly preferable to surround the arc by a glass-in- 50 closing chamber,.both lto prevent unsteadiness `and to prevent the poisonous vapor of mercury from escaping into the surrounding space. Now I find that all of the substances Which might naturally be introduced to the mercury arc to modify its color, and particularly lithium, sodium, plotassiuny and rubidium, violently attack t is glass envelop, or any other transparent ortranslucent envelop known to me, and within a -few minutes or at most a few hours destroy its ltraisparency by causing a black deposit or incrustation in the interior.

It is, therefore, another important feature of my invention to artificially maintain the translucency of the envelop in which that arc exists, and though I point out in this specification one particular means which in practice I have found most ad- \-'antageous for this purpose, and though I intend by the claims attached to this spccification to claim the means both broadly and specifically, nevertheless it should be understood that in many aspects my invention is not limited thereto, but I have dcvised other mea-ns for accomplishing this beneficial result, which means, though included under the broad claims of the present. application, are, nevertheless, to be specifically covered by other and later cases. The particular means of this case consists in adding-the modifying conducting vapors in a chemically inert condition, and, preferably, in what is known as the ionized state, that is to say, as free atoms co-existing with 'free atoms of some other substance combining with the conducting vapors at ordinary temperatures, but at eompleteor partial dissociation at the temperature of the mercury arc.

I find that the elements ofthe halogen class, that is to say: iodin, bromin, chlorin and tluorin, are useful in the order named; that is to say, iodin is the best, since the iodids dissociate at a comparatively low temperature. Thus, for example, if a ceri tain shade of red is desired, I introduce into the arc-stream a large quantity of lithium vapors in the ionized condition, by covering the surface of the mercury with an hydrous lithi/um iodid. If, again, I wish to produce pure White light I use a comlparativelyl small quantity of a mixture of the iodids of lithium and sodium or even of lithium, sodium4 and potassium.'

If I Wish to produce a bright blue light I use the iodid of indium; if I wish to produce a dark green light I use iodid of thallium; a yellow light is produced by iodid of sodium, and another shade of red is produced by iodid of potassium and rubidium.

When producing these highly-colored lights it seems that the mercurysarc -stream acts prmcipally as a carrier of the vapors of the substances introduced, while when yproducing white light the mercury arc will itself give the violet, blue and green, and the rays of longer wave leneth may bc reduced by the substances artificially intro uced.

It is best in any case to cause the powdered iodids, or speaking more generally the color-modifying substance or compound, to cover completely the mercury electrode or electrodes, in order to prevent the arc from striking uncovered mercury, which would cause its color to vary. But I find that it sometimes happens that an electrode covered in this way gives a light in which the spectrum lines of the color-modifying substances are too prominent,

It is another object of my invention to overcome this difthe color-modifying substance by adding thereto a relatively' inert substance, Vand vby preference one which does not `dissociate at the temperature used. Calcium fluorid works well, as it is inert` in the presence of Vthe halogen salts used as color-modifiers,

laa

and so stable that. it does not-dissociate at the temperature of the mercury arc.

By varying the proportions of active coloring substancesl andI inactive diluting substances any shade may be produced, and

candle light, incandescent electric light,

etc., may successively be imitated.

I have, then, in compliance wit-h the statutes, briefly indicated the natureJand object of my invention. I will now proceed to describe the apparatus which. I use for carrying it out in practice, with reference to the accompanying drawings, and wi1l,in the claims attached to this specification, indicate the scope of my invention in its several aspects.

In the drawings attached to this specification, Figure l is a'view of a lamp constructed in accordance with my present 1n` vention, connected in a series system; Fig. 2 is a View of a modified form lof lamp; Fig. 3 is an enlarged detail'in cross section of the form shown in Fig. 2; Fig. 4 shows an improved method of connecting into the ci r cuit the lamp of Fig. l; Fig. 5 shows a still further modification; Fig. 6 shows a form` ox" lamp devised to take the place of the ordinary inclosed arc lamp of commerce; Fig. 7 is a diagram of connections; Fig. En is a view in section of a modification; Fig. 9 shows a stilldifferent form; and F l0 is a diagram of circuit for Fig. 9.

In Fig. l, A is a generator, which may be supposed to typify any source of constant electric current, whether alternating or direct, and B, B', etc., are lamps connected in circuit in series. C is a switch shunting the lamp B, and similar switches will be used to shunt each lamp of the system. The lamp. B is shown as consisting of Van inverted U tube with mercury electrodes D D and platinum terminals E E. The surface. of the mercury is covered with a thin layer, F. ofv iodid of some material or materials giving conducting vapors, in accordance wit-h the general principle above set forth, and in the particular lamp shown we maysupposc that this layer is composed of a mixture of the iodids of lithium and sodium with or without t-he addition of'dilut-ant substances such as fiuorid of calcium. These iodids are introduced into the tube, and care must be taken to have them perfectly anhydrous, sincethe presence of moisture is extremely harmful. The tube is then exhausted in the manner usual 'with ordinary incandescent lamps, but I find it advantageous to replace the residual air by hydrogen, since an atmosphere of rarefied hydrogen requires a very low initial voltage as compared with the atmosphere of rarefied air. 'Y

With helium or argon, where either of these gases ca be obtained, the starting voltage may 'be still further reduced. If now the tube besealed o' and the switch. C be thrown to the position shown, and particularly if the mercury electrodes D D be artificiillly heated, a luminous arc or discharge will strike from one electrode to the other, and the whole tube will seem to be'filled With a white, soft light of considerable i11- tensity. When the light is once fairly started the voltage across the lamp falls con` siderably. The action. appears to be somewhat as follows: A true mercury arc exists between t-he terminals Dv D, that is to say, the current flows from one terminal to the other through a path of mercury vapor, and this mercury vapor, under the influence of the electric current, gives a brilliantA light containing violet, blue and green rays, at a temperature far below the temperature of incandescence, and with an eiiiciency much greater than can, be obtained from any known incandescentlight. The mercury is evaporated from moment to moment, and during the operation of the lamp condenses in the condensing chambervG, whence Ait trickles back to the electrodes. The arc stream is thus maintained as a continuous body of mercury vapor. The arc, which is more definite at its ends than at any other point, plays back and forth over the surface of the mercury electrodes, and continuously vaporizes small quantities of the halogen salts of alkali metals. I have in this particular case assumed thesesalts to be iodids of lithium and sodium. These salts are evaporated as above stated, and are carried into the arc stream. Atthis temperature they partially dissociate, so that freeatoms of lithium and sodium exist in the arc stream, at least as far as their' light-giving function is concerned, and the arc shows thc characteristic spectra of lithium and sodium. These substances, however, are not chemically free, but are in very intimate relation with the iodin atoms froml which they have l'ously combining, dissociating and recombinthat it is adapted for a series circuit, and,v

also that it requires to be started by bringing the mercury nearly to the point of boils ing by artificial means. This particular lamp, however, is typical of my main inf. vention, though other types of lamps, to be hereinafter described, avoidin a. greater or less degree the difficulties abovev mentioned. I have found that 1t 1s not necessary that both. electrodes should be formed of mercury or other substance vaporizable at low temperatures, for the reason that if one of the electrodes is of such a` substance the evaporation from this electrode will tend to reduce the temperature of the are to an etlieient point. I, therefore, am able to form one electrode of iron, or other conductor, as carbon, etc., though when direct currentis used it is preferable to make the refractory terminal the cooler and the mercury or volatile terminal therhotter electrode, in order to ins-ure the maintenance of the proper vtemperature 'in the globe. The negative electrode is the cooler -at atmospheric pressure and at' low vacua, but this relation may sometimes reverse at higher vacua. v

In Fig. 2 I show another form of lamp which I have found to operate well. This lamp is composed of a single vertical tube B', expanding at its upper portion to a. globe B2. The lower electrode D is of mercury as before, and should be provided with color-modifying substances as above eX- plained. This electrode ,is connected through the terminal E to one side of the system, or to one terminal of the generator. The upper electrode D is formed of metal, as iron, and is conduct-ively supported Afrom the negative terminal E. I have found in t-he use of lamps of this type that the arc has a. tendency at high vacua to run up the side of the electrode D', and to fuse the glass at or near the point where the terminal passes through. I attribute this phenomenon to the'fact that the transition resistance between the vacuum and the electrode is larger than the resistance of the vacuum. Whatever the cause may be, the phenomenon is highly objectionable, for the reason that it tends to melt and discolor the glass. I ind in practice that the trouble can be overcome by surrounding the electrode D', for a part of its length, by a glass tube H, which is fused directly to, or forms part of, the lamp itself. It is difficult for the arc to exist in the narrow space bet-Ween the sides of the tube and the electrode, so that the are naturally tends to strike the electrode at or near thelower portion of the tube, at a point where the electrode is out of contact with the glass.

Fig. 4 shows the lamp of Fig. 1 with an improved device for connecting it in cir` cuit. In this figure, A is a constant-current generator. B is the lamp and C', is a switch. The resistance R is shunted around the lamp and the amount of this resistance in circuit is controlled by the resistance of the switch' C. In the position shown in the drawings current passes freely from t-he generator through the switch around the lamp. When the switch is moved to the left one point, a.

small portion of the. resistance It is .connected in .series in the constant-current circuit and the lamp is shuntedaround the drop of potential ycaused by thisresistance.

For example, in a ten-ampere circuit, if the Athe voltage across the lamp until finally an are is struck and that in the meantime the constant-current circuit will not be interrupted, provided that the switch arm C is constructed as such switches usually are constructed, with a terminal wider than the distance between contact segments. When the lamp is heated into condition of stabilit the arm C can be thrown still farther to the' left, which will open-.circuit the resistancewR and cause .all of the current to pass through the lamp. I prefer to place lightning-arrester spark gaps or film cutouts J around each lamp in order that if for any reason the are breaks, the constant-current'cireuit may not be interrupted.. In

' practice I find that the arc tends to flicker unless the mercury is maintained atits boiling point, and with some designs of tube it is not possible to accomplish this by the heat developed by the arc itself. I, therefore, iind it advisable in some cases rto artifcially generate heat at each electrode. The arrangement for this purpose is shown at Fig. 5, in which the lower ends of the tube B are surrounded by soft iron caps K K, and around each of these caps is coiled a number of turns of the Wire which carries the current to thelamp.` If the current is alternating it will tend to set up eddy currents in the iron caps and these caps will thus become heated and tend to maintain the mercury at the proper-temperature. If the current is direct it is necessary to use more Wire and to place it in heat-conductive relation to the cap. The heating' coils may be placed in series with the lamps, as shown in Fig. 5, or they may be placed in multiple as shown in Fig. 4 at K. The advantage of those skilled in the art that it may readily be altered to work on constant-current circuits as well, or to consume direct current.

Referring to Figs. 6 and 7 which show vthis lamp, it will be seen that'current enters by the terminal E2 and thence. passes to a regulating device L. As this particular lamp is intended .f for alternating current, this regulating device may well take the forml oi? closed magnetic circuit, such as is ordinarily employed in alternating arc lamps. The usual taps or connections may be provided for regulating the amount of this reactance in circuit. From this reactive coil, current passes by the wireM to the mercury elec' trode D3, and thence through the arc in the tube B to the upper terminal D3, which is preferably formed of iron. From this terminal itpasses by the collecting spring N through the wire M2 to the solenoid coils N3 N4, and thence to the terminal E3. It will be seen that these are the ordinary connections of an are lamp to be operated in multiple are on a singlegphase alternating circuit. To describe the structure more in detail, it may be said that the support N3 is screwed to a cap N3, and that from this-cap depend a plurality of rods N7 N7, fastened at their lower ends by screws as shown, or otherwise, to the lower plate of the lamp N3. These rods also carry, by the clips N9 N3, the reactive coil L. A U-shaped core O of laminated iron is fastened to the lower plate N3, and the solenoids N3 N 4, above referred to, are mounted respectively on the limbs of this core, and are so connected that they have a tendency to polarize the magnetic circuit formed by this core in the same direction. A repulsion device O, which may consist of a solid casting of copper, alumi num, or other suitable material, is pierced with two holes in order that it may tit over the two limbs of the laminated piece O, and to this repulsion device are attached lfour guides O2 O2, two only of which are shown in the drawing, which work through suitable openings in the upper flange N10 of the spools of the solenoid cores N3 N4. The upper tlange is expanded for this purpose, Iwhile the lower flange is of smaller dialneter. The upper electrode D3 is fastened to the repulsion device O, and is .so adjusted that the weight of the parts is sufficient, when no current is passing through the lamp, to compress the coil spring O, and to cause the lower end of the electrode to dip into the mercury at D2. Z is a dash-pot, serving to steady the lamp and to prevent pumping. It will be seen that a screw-ring P is atta'ched to the lowerplate of the lamp, and that into this ring is screwed the glass con- (lensing chamber P2. A suitable washer'P3, of felt or asbestos, is interposed between this condensing chamber and the lower plate. A second screw-ring P4 is fastened to the lower lpass to the condensing chamber.

a self-induction coil with a nearly plate of the lamp -by the screw-clam P3, and into this ring ,is screwed the tube with a second washer of felt or asbestos P3. Considerable clearance. is allowed between the electrode and the ed e of the opening, in order to make it possible for the -vapors to ln order to carry current to the lower terminal of the lamp, and also in order to provide for the return to the lower terminal of the mercury vapors which may escape into' the chamber P2, I provide a glass tube'P7, which. is connected by asbestos washers P3 with an eXtension from the chamber P2, and .in a similar way is connected with an extensionfrom the lower end of the tube B. Through this tube I pass the wire M', which at its lower portion is coveredwith an enamel insulation in order to protect it from the mercury vapors. I further prefer to coil this wire into a heating coil M3, and to uncover the insulation only at the extremity M4, whereby current will pass through the heating coil M3 and will tend to maintain the mercury at D2 at a sufficiently high temperature. Another way of accomplishing this same result is shown in Fig. 8, in which a metal cap K is screwed on to the lower end of the tube, and a second cap K3 is screwed on the cap K', leaving between the two an annular space in which may be coiled the resistance wire M3. The parts K K2 may bey formed of iron in which case the Foucault currents I generated will tend to heat described in Fig. 5.

The operation of the lamp shown in Fi 6 is as follows: The current enters at the terminal E2 and passes through the selfinduction coil M, in the path above described through the heating coil M3 to the mercury electrode D2. If we suppose the lamp to be just starting into action, it is clear that the weight of the repulsion device O will have compressed the spring O4, so that the electrode D3 will be incontact with the mercury D2. Currentl will therefore pass through the solenoids N3 N4 and out to the terminal E3. The solenoids will the mercury, as

set up lines of force in the core O, and byv This may be repeated several timesbut` finallythe mercury will be heated to a sutliciently high temperature, andthe electrode D3 will rise to the position shown in the drawing. The length of the resultant arc will obviously be regulated by the design of the' solenoids N3 N", the design of the reactance coil, the voltage of the circuit, the

diameter of the tube B', etc. An increase in the resistance of the arc will weaken the current, weaken the repulsive action, and

cause the yarc to be shortened. The tendency is toward constant current in the branch in- "mentioned in the portion D2.

cluding the lamp, as in most commercialv are lamps. In this-position the tube B. will be filled with a brilliant glow of light, and

the color of this light will be regulated by introducing the various substances ,above The vapors will tend to escape from the tube B, but

.they are collected in the condensing cham ber P2, andpass back through vthe tube P7.

I find it much better to pass the condensed I sure, but it is best not to. allow mercury fumes to escape from the lamp. I am able to accomplish these objects by the tube T, which is inserted in an opening inthe lower plate Nfand registers with a round or crescent-shaped opening in the washer P3, and with a corresponding opening or recess in the upper wall of the chamber P2. This tube extends vertically upward, being bent around the lamp mechanism. It allows air esctto escape from or to renter the chamber P2, but the mercury vapors, which are very heavy, will not' rise to thetop, but will be condensed on the sides and will ltrickle back to the condensing chamber,. and thence through the tube P7 to the lower electrode D2.

As an illustration of lanother manner in `which the principles of my invention may be applied, I will no w describe the lamp shown in Figs. 9 and 10. This lamp is, or

.may be, identical with the lamp shown in Fig. 6 with res ect tall portions below the lower plate N2, so that further illustration or description of these parts is deemed unnecessary. Referring more particularly to Fig. 10, which shows the circuit of the lamp, current =enters, for example, at the terminal E2 and passes through the selfinduction coil L, thence through the lower4 electrode D2, to the upper electrode D3, through the series coils N1? and N12 and the switch V to the terminal E3. It will be seen that in this lamp the repulsion armature O is replaced by a U-shaped core of laminated iron O6, which tends tobe pulled up into the solenoids N11 N12. "FurtherpI provide two additional solenoids N18 N, which are shunted around the arc. `The presence of these shunt solenoids enables me to weaken the series coils, while obtaining the same side rods NT and the lower plate N2 are substantially identical with Athe corresponding parts in Fig. 6, land the portions of the lamp below the lower 'plate N8 are also substantially identical with the correspond ing parts in Fig. 6 The series solenoidsfN11 N12 and the'shunt solenoids N1a N are connected as shown in Fig. *10, and the upper electrode D3 is vfastened 4rigidly to the U'- shaped core O, which is preferably composed of iron wire bent to the form shown. This U-shaped core O6 'is suspended by a spiral spring O" from the cap-plate of the lamp in order to assist the coils 1n springing `the very long. arc which is preferable in lamps constructed inaccordance with my present invention.

I have said that the object of the cumulative shunt solenoids was to reduce the rapidity with which the separation of the elec i trdes took place. I am able to st-ill further reduce this rapidity of separation by the use of the dash-pot W. vThis dash-pot is provided with a piston W', which is mounted on the upper electrode D3, and the upper portion of the dash-pot is ared to a funnel shape, as indicated at W2, in order to facilitate the entrance of the piston into the dashpot as the electrode descends. The sides of the piston are grooved, as shown at -W3, to

increase the dash-pot action, and the piston is pierced with one or more openings W, controlled by valves W5. The valve shown is'a sim le piece of thin sheet asbestos, or 'other so material, arranged to o en as the dashpot descends but to close as 1t ascends.

The operation of this lamp will be apparent from the description already" given. When the switch V is closed current passes through t-he lamp in the path above traced, vand the series coils tend to lift the core 'O6 and' the electrode D3.

This action is, however, weaker than the correspondmg action in the lamp shown in Fig. 6, since neither the' series nor the shunt coils alone -are required to be strong enough to sustain the 'moving .system in its upmost position. The action is still further resisted by the dash-pot W, so that some litt-1e time elapses before any considerable separation of the electrodes takes place, and during this time a heavy current is owing throu h the lamp, suiiicient to heatthe mercury Iorming the lower electrode D2, shown in Fig. (i. lThe parts are so adjusted that as soon as the mercury is fairly well heated the ldash-pot piston W8 passes beyond the narrow'portion of the dash-pot WV, and the series coils are then left Vfree to raise the core O and strike the arca As the arc lengthens the shunt coils also become energized, and these coils, togeth'el` with the spring O7, assist in extending the arc to a proper length. If, however, it should happen that the arc became long before the mercury was properly heated, the circuit would be broken and the upper elect-rode would fall into they mercury cup, when the action would be repeated, 1f necessary several times, until the mercury attained the proper temperature. V

It should be understood that the 'series coils are so powerful, with reference to the.l

shunt coils, that if the current becomes weakened by reason of 'an excessive lengthening of the arc, or otherwise, the Jseries coils allow the upper electrode to fall sufficiently to maintain the current taken by the lamp at approximately its proper predetermined value;

It will be evident that the dash-pot W is entirely independent from the dashepot Z shown in Fig. 6. The dash-pot 'Z is intended to take the place of -the ordinary dash-pot of the ordinary arc lamp to prevent the lamp from pumping, 'while the dash-pot W actsonly for a few seconds and is intended merely to facilitate Jthe heating of the mercury terminal. Both dash-pots may be used in either of the forms sh'own, or by an obvious modification they may be combined in one. y

It will, of course, be understood that in both of the forms of Figs. 8 and 9, I cover the mercury elect-rode by some suitable colcrmodifying substance, as above fully set forth, and that I adjust the operation of the color-modifying substance or` substances by proper proportioning and if necessary by the addition of a halogen salt stable at the temperature cf the mercury arc, as calcium iuorid, or'by other suitable inactive dilutent substance.

1tiVhat I claim as new and desire to secure by Letters Patent of the United States, is-

1. rli`he combination with an arc drawn from a mercury terminal, of means for continuously furnishing to the arc diluted salts of alkali metals.

2.. The combinationof a glass container, a mercury electrode therein, means for mailit-aining an arc from said electrode and means for continuously furnishing to lthe arc metallic'compounds in a chemically inert condition, said compounds being ioniz able at the temperature of the arc.

3. The combination with an arc drawn from a mercury terminal, of means for continuously furnishing to the are halogen metallic salts. i

4:. Thecombination with/an arc drawn from a mercury terminal, of means for con-y tlnuously furnishing to the arc iodids of iodids ofalkali met-als furnishing red rays. A

6. The combinationjwith a terminal, of means for drawing an arc therefrom" at a temperature below the temperature of incandescence, a surrounding globe or inclos; ure formed of means for mo ifyin the color'of the arc comprising an anhy rous powder of iodids of one or more alkali metals. f

7. The combination with 'a terminal, of

glass or sinnlar material, and l means for drawing. an arc therefrom at a temperature below thetemperature f'of incandescence, a surrounding globe or inclosure formed of glass or similar material, and means for modifying the color` of the arc comprising an anhydrous powder of Y iodids of one or morev alkali metals mixed with ouiid of calcium.

8. The combination in an are lamp, of a' mercury electrode with a layer of coloring material `mixed with dilutent material.

9. In an arc lamp, a mercury electrode covered with a mixture of. stable and unstable halogen salts of alkali metals.

10. In an' electric arc lamp, a terminal volatile at omparatively low temperatures, with amixture of halogen salts of alkali metals in operative relation`with said terminal, some of said salts being of such a nature as to dissociate below and some above the boiling point of the material forming the electrode. i

11. The combination in ,an arc lamp, of a mercury terminal, a movable terminal, a heating coil for the mercury terminal, and a regulating magnet coil in series with -the arc.

12. The combination of an inclosing globe, a mercury terminal, a movable terminal, means controlled by the current for operating said movable terminal, and means for continuously adding luminous conducting vapors to the arc, whereby its color may be modified.

13. In an electric arc lamp, an electrode of mercury, an electrode of more refractory material dipping in said mercury electrode, means controlled by the current for separating the two electrodes and for regulating the amount of separation, a heating coil for the mercury electrode, and artificial means for retarding the separation of the electrodes.`

14E. The combination in an arc lamp, of a volatile electrode, means for heating the electrode, and means for subsequently striking an arc from said electrode.

8 i y l I 1,025,932

` 15. In an arc lamp, an electrode covered `Wit-h a color modifying covered with4 metallic salts which will dissociete at the temperature ofthe mercury arc.

17. The combination of a mercury elecio trede, a. (1o-acting electrode, means for drawing an -arc between said electrodes and means forfurnishin the arc metallic compounds which will dissociate at the temperature of the arc.

In witness whereof, I have hereunto set 15 my handthis fifth day of April, 1902.

CHARLES P. STEINMETZ. Witnesses:

' BENJAMIN B. HULL,

HELEN ORFoRD.

Images