Spyros E. Iakovidis - The Mycenaean Acropolis of Athens - 2006

THE MYCENAEAN ACROPOLIS
OF ATHENS
ISSN 1105-7785
ISBN 960-8145-52-X
© The Archaeological Society at Athens

Panepistimiou 22, Athens 106 72
Fax 210 3644996 - tel. 210 3625531 - archetai@otenet.gr - www.archetai.gr
Editing: Ilektra Andreadi

THE ARCHAEOLOGICAL SOCIETY AT ATHENS LIBRARY No 240
SPYROS E. IAKOVIDIS

OF ATHENS
ATHENS 2006
CONTENTS
Preface to the English edition 9-10

Preface 11-13
Introduction 15-16
PART ONE
THE MYCENAEAN ACROPOLIS IN THE ANCIENT SOURCES
AND IN THE BIBLIOGRAPHY
The ancient sources 19-24
Excavations and research 25-29
The Mycenaean Acropolis in the bibliography 30-39
Conclusions 40-41
PART TWO
THE MYCENAEAN ACROPOLIS ON THE BASIS OF THE FINDS
The rock 45-49
The Acropolis during Pre-Mycenaean times 50-55
The Mycenaean period 56-245
I. The Late Helladic remains on the Acropolis 56-68
II. The building phases 69-72
1. The first phase 73-75
2. The second phase 76-114
The area W of the terminus of the NE ascent 76. The area
SE of the end of the NE ascent 101. The NE ascent 105. The
Northwest descent to the caves 107. The second Late Helladic
phase as a whole 111. Chronology of the second phase 113
3. The third phase 115-231
The top of the rock 115-196
The west side 115. The north side up to the North Foun-
7
CONTENTS
tain 128. The North Fountain House 140. The north side from
the North Fountain to the NE ascent 144. The north side
from the NE ascent to the Belvedere 157. The east side 160.
The SE corner 163. The south side 171. The arrangement of
the west entrance 182. The palace 190
The Pelargikon 197-221
The sources and the bibliography 197. The remains of the
Pelargikon in situ 210
The slopes 222-224
The third phase as a whole 225-226
Chronology 227-231
The wall 227. The other buildings 228
III. Construction methods 232-245
The terraces 232. The fortification 234. The North Fountain
239. The houses 243. The palace 244. The stairways 244. The
graves 245
EPILOGUE
THE ACROPOLIS OF ATHENS IN RELATION
TO THE OTHER MYCENAEAN CITADELS 247-253
Appendix I: Ancient testimonia 255-272
Appendix II: The pottery sherds 273-283
List of Illustrations 285-288
Abbreviations - Bibliography 289-293
Index 295-301
8
PREFACE TO THE ENGLISH EDITION
The Greek edition of the Mycenaean Acropolis of Athens was published in

1962 and was favourably reviewed by 0. Broneer (Gnomon 35, 1963, 708-711,
Archaeology 17, 1964, 70-71), H. W Catting (JHS 86, 1966, 271-272) and F.
Schachermeyr (Anzeiger fur die AltertumswissenschaftX/X, 1966, Forschungs-
bericht). Since, however, it was written in Greek it remained largely unread by
non-Greek scholars. Those who dealt with the subject referred as a rule to a
couple of precis of the monograph, published much later, the first as a chapter
in the volume E 1 of Archaeologia Homerica, pp. 193-204 (Gottingen 1976)
and the other as part of my monograph Late Helladic Citadels on Mainland
Greece (Leiden 1983). Thus, the detailed description of the Mycenaean remains
of the Athenian Acropolis - either already known or identified as such by me after
careful investigation and small scale excavations which I had conducted on the
rock - went mainly unnoticed and so did the resulting close argumentation lead-
ing to the conclusions I had arrived at. As a result the Late Helladic citadel of
Athens is still much less known than the other similarly fortified sites of the period.
Time went by with much to do and the idea of a translation was shelved again
and again, until finally I realized that, if I wanted my work to reach a larger pub-
lic, it would have to be translated into the lingua franca of international scholar-
ship, namely English. For this purpose I was fortunate enough to secure the col-
laboration of Dr. Miriam Caskey, an archaeologist in her own right and an expe-
rienced translator of scholarly texts, who produced not a literal metaphrase but
an accurate and at the same time a free English version of the Greek original.
During the years between the publication of the book and its translation very
little has been added to the relevant bibliography. What was written concerned the
west retaining wall of Terrace Ill, the arrangement of the West Entrance and the
houses on the NE ascent. The views expressed about them, however, do not take,
in my opinion, account of the archaeological evidence. It became therefore necess-
ary for me to alter some of the footnotes of my original text and add a few
others with my own comments.
9
PREFACE TO THE ENGLISH EDITION
My thanks are due to the Board of the Athens Archaeological Society and its
Secretary General, Dr. B. Petrakos, who agreed to include this English transla-
tion of my book in the prestigious monograph series of the Society.
My thanks and congratulations go to Mrs Electra Andreadi for her invaluable
editorial assistance. During our close cooperation this book and its author prof-
ited greatly from her long and noted expertise and her meticulous attention to
detail.
SP. IAKOVIDIS
10
PREFACE
The fortified citadel, a characteristically Mycenaean creation, has always

drawn my interest, even in my student days. Of all these citadels the Acrop-
olis of Athens is by far the most difficult to comprehend, for its remains are
few and seemed somehow to demand investigation. The existing bibliography,
moreover, convinced me that there was sufficient material to justify a sys-
tematic study of the subject. Yet, the Acropolis rock itself had to be thor-
oughly explored in order to reach any valid conclusions.
My participation in the excavations of the citadel of Mycenae carried out
by Professor G. Mylonas for the Archaeological Society gave me first-hand
experience with the problems of such citadels and their peculiarities. It was
good training for any exploration that might prove necessary and feasible on
the Acropolis of Athens, where special conditions pertained through long and
continuous use of the place.
To collect and organise the existing material presupposed an examination
of all accessible remains thought to be Mycenaean, so I decided to carry out
minor excavations wherever the fill was not so deep as to cause problems.
Wherever possible, I carefully examined the rock itself, especially along the
north wall. In fact, by excavation, clearing or simple observation, I explored
the entire length of the north wall, the area in front of the Propylaia, all the
space between the Propylaia and the Erechtheion, a number of places east of
the Erechtheion, the rock behind the base of the statue of Athena Promachos
and the space between the Belvedere and the Museum. New measurements
were also needed, either to correct earlier ones or to include in the plans items
hitherto overlooked. A considerable amount of new material came to light in
this way. In some cases chronological information was provided by sherds
found in Mycenaean walls, or in later walls thought to be Mycenaean. The
results of this work are shown in a series of plans accompanying the text. They
are based mainly on Kawerau's plans, on sheet n° 55 of the Office of Real
Estate Registry of the Ministry of Transportation and Public Works (1929 sur-
11
PREFACE
vey), on N. D. Ioannitis' survey of the area around the caves in the northwest
part of the Acropolis rock, on J. Travlos' plans of the area of Klepsydra and
the northeast descent and on supplementary measurements made by myself.
The research was carried out during the summer of 1960 and winter of
1961. I was fortunate in having the valuable assistance of all the authorities.
Everyone I approached for an opinion or for assistance with the various ques-
tions that arose during the course of my work was equally helpful. The Direc-
torate of Antiquities gave me permission to carry out the necessary excava-
tions. The Directorate of Restoration provided the personnel required. The
Director of the Acropolis, J. Miliades and the Ephor G. Dontas, with their
warm interest, contributed in a fundamental way to the unimpeded progress
of the work. Moreover they permitted me to mention finds, as yet unpub-
lished, from excavations at various places on the South Slope. The director of
the Agora excavations, Professor H. Thompson, allowed me to consult the
notebooks of the excavation in the Klepsydra area and, with his colleague E.
Vanderpool, made it possible for me to examine the sherds found there, most
of them unpublished. J. Travlos, architect of the American excavations, placed
his plan of that area at my disposal together with all the relevant information
I requested. That veteran researcher and Acropolis expert, G. Stevens, called
my attention to the existence of a Mycenaean terrace behind the base of the
Athena Promachos, made many suggestions and gave me also much valuable
advice. Professor A. Orlandos provided me with information about the S side
of the tower of the west entrance which he had uncovered, and gave me per-
mission to make measured drawings of it. Professor G. Mylonas allowed me
to use various conclusions about details of the excavations at Mycenae, as yet
unpublished. Moreover, he kindly read the manuscript of my study, and found
time for discussion. I profited much from his valuable advice. To Professor 0.
Broneer I owe much significant information, mostly in connection with the
chronological evidence, not only from his own discoveries but also from the
finds of Kolbe and Balanos, of which he had personal knowledge. E. Fiandra
of the Italian Archaeological School, the discoverer of the Mycenaean well in
the area of the Stoa of Eumenes, allowed me to refer to her still unpublished
find. The Ministry of Transportation and Public Works provided me with a
copy of the topographical plan of Athens that includes the Acropolis and the
area around it. The General Staff of the Air Force gave me an aerial photo-
graph of the area. From the Archaeological Society I received a full set of the
12
PREFACE
plates that had been used for the publication of the Acropolis excavations,
including Kawerau's plans. J. Bandekas, topographer and civil engineer,
undertook and completed the chief measurements needed for the new survey
of the area. The German Archaeological Institute furnished me with pho-
tographs from their archive of the great excavation of the Acropolis. Profes-
sor P. Mylonas, of the Polytechnical School of Fine Arts, gave me a photo-
graph of the section of the Mycenaean wall south of the Propylaia, taken
before the installation of floodlights and wires that hide it now.
Professor Sp. Marinatos, who approved the choice of subject and followed
the course of the entire research with lively interest, introduced the thesis
when I presented it to the Philosophical School of the University of Athens
in June, 1961. His advice and suggestions were invaluable contributions to my
efforts. To all those mentioned above, I give my warmest thanks for the assist-
ance they so willingly provided.
SP. lAKOVIDIS
13
INTRODUCTION
The Acropolis of Athens is known chiefly from its Classical buildings and
from the works of art in general that were within its area in historical times.
Yet, long before, in an era that was mythical even to the ancient Athenians
themselves and remembered only fragmentarily through tradition, the Acrop-
olis was already. inhabited and it was the seat of rulers. Through the myths
and traditions that have come down to us, we discern the efforts of the kings
of Athens, first to hold on to their position in Athens itself and then in Attica
as well. At the same time there were struggles with Eleusis and difficulties
with Crete during her temporary hegemony over their land.
As we now know, a citadel stood on the rock in Late Helladic times. It was
clearly planned, with strong fortification walls. Here lived the ruler and some
of his subjects. At the end of Mycenaean times, the other Mycenaean citadels
and settlements were destroyed by fire and subsequently deserted. As a result
of conflagration and the fill that built up after their abandonment, much has
been preserved, so much, in fact, that we know more today than did the
Ancients themselves. Yet the Acropolis of Athens was untouched by such a
catastrophe and it continued to be used without a break. The claim of the
Athenians that the Dorians never took their land clearly is correct. From that
time on, the Mycenaean buildings on the rock fell victim to just that contin-
uous use, for they were torn down, rebuilt and continuously altered. The for-
tification wall alone (apart from a number of changes made during Archaic
times) was preserved in its entirety until 479 B.C., when the Persians razed
it. 1 After this, the last remaining traces of the Mycenaean buildings disap-
peared beneath artificial terracing and foundations in the general reconstruc-
tion of the Acropolis in Classical times. Successive occupants in turn corn-
1. Herod. IX 13.
15
INTRODUCTION
pleted the destruction, the last being the Turks. So it is that very little has sur-
vived, far less than in the other Mycenaean citadels.
The traces of Mycenaean remains preserved on the rock today are for the
most part inaccessible. Some can hardly be described as specific traces, but
rather as traces of traces. They comprise cuttings in the rock, probable foun-
dation trenches and hints given by the siting and disposition of later buildings
and constructions which, for one reason or another, had to be accommodated
to the Mycenaean remains. Moreover, many remains of Archaic, Mediaeval
and even Turkish times have been wrongly identified as Mycenaean. This has
only added to the confusion. A systematic study and sorting of the existing
material, and new research (of the sort and to the extent noted in the Pref-
ace) were therefore needed to find further evidence to illuminate and sup-
plement what was known already.
The subject has, indeed, attracted scholars from time to time and there is
a certain amount of relevant ancient information as well. With the publica-
tion of various excavations and studies on the Acropolis and the immediate
area, there is now a fairly extensive bibliography. The first half of the present
study is devoted to a commentary on this. The second half gives the results
of my research on the rock itself. On the basis of this research and relevant
excavation reports, I have attempted to compose a picture of the Mycenaean
Acropolis.
16
PART ONE
IN THE ANCIENT SOURCES
AND IN THE BIBLIOGRAPHY
THE ANCIENT SOURCES*
The ancient sources contain a fair number of references to the Mycenaean

fortifications of the Acropolis,2 yet the concrete information they give is
limited. Most of the references - and certainly the earliest - are indirect. They
refer to the fortifications incidentally and always in relation to something else.
Direct references, in fact, are practically non-existent. They are brief, for the
most part much later, and they are chiefly by lexicographers, whose depend-
ency on older indirect sources is evident enough.
All the ancient sources attribute the fortification walls to the Pelasgians
(or Tyrrhenians ), 3 a people that tradition says were driven out by the Athe-
nians, first from the city area and then out of Attica. 4 According to Cleidemos,
the Pelasgians levelled the top of the Acropolis rock and surrounded it with
a fortification wall that had nine gates. 5 This particular Acropolis fortification,
together with the area it enclosed, is sometimes called the Pelasgikon, with
cr-,6 sometimes the Pelargikon, with p. 7 In one case, the two spellings are used
* See infra, Appendix I. 4. Herod. VI 137 (Hekataios ).

2. The relevant ancient sources have 5. Bekker,Anecd. Gr. Ip. 419, 1. 27, Soui-
been collected by Jahn - Michaelis, Ar.x Ath. da s.v. aTie8a and TJTIE8t~ov. The i)Tie8t~ov
pp. 1-23, and by Judeich, Top. p. 52, n. 5. of Cleidemos is generally taken to mean the
In particular, the sources referring to the levelling of the rock by trimming it.
Pelargikon are quoted by Jahn - Michaelis, 6. Kratinos, uncertain fragments , Ed-
Ar.x Ath. p. 79, n. to I. 14, and Wide, Auso- monds, n° 321, corrected by Wilamowitz,
nia 1912, pp. 195-197. Hermes XIV, p. 183, Par. Chron. 60, Schol.
3. Herod. VI 137 (Hekataios), Aristoph. Aristoph. Lys. 1150, Luc. Pesc. 42, 47, and Bis
Av. 1139 (by allusion), Paus. I 28, 3, Schol. Acc. 9, Philostr. Vit. Soph. II, a, V, Pollux
Aristoph. Av. 832, Bekker, Anecd. Gr. I p. Onom. VIII 101, Strabo IX 401, Schol. Thuc.
299, ll. 16-18, Hesych. s.v. nei\acrTtKOV and II 17, Hesych. s.v. nei\apytK6v (where he
nei\apytKOV, Phot. Lexikon s.v. nei\apyt- corrects it to er).
KOV . See also Berard, Stud. Rob. p. 151 f. 7. Herod. V 64, Thuc. II 17, Ditten-
19
THE ANCIENT SOURCES
indiscriminately,8 and Hesychios adds the form nEAOCJTlKOV. 9

The distinction between the two given forms, Pelasgi,kon and Pelargikon,
the determination of one as correct and the way in which one is derived from
the other, lies in the realm of critical examination of the texts and is quite
outside the limits of the present study. I therefore confine myself to a few
remarks.
Recent authors use one form or the other, usually indiscriminately, only a
few giving reasons for their choice. Harrison 10 considers Pelargikon to be the
correct form, since she believes the name to be derived from the storks re-
presented on the geison of the Archaic Hekatompedon. None have supported
this peculiar view. Berard, on the other hand, 11 considers the earlier, and
therefore more correct, form to be Pelasgikon, observing that this is the form
given by Herodotos and Hekataios. He notes that the p appears only after
the Peloponnesian War. Yet this is not actually the case. Where Herodotos
refers to the information given by Hekataios,12 the discussion is about the
Pelasgians and the wall they built. The term Pelasgikon or Pelargikon is not
mentioned at all. To the contrary, in his references elsewhere Herodotos uses
the term Pelargikon. Picard 13 considers Pelargikon to be correct, as this is the
berger, Sylloge3 n° 83, Aristoph. A v. 832, viewpoint, this form is to be connected with
Aristot. A th. Pol. XIX 5-6, Cleidemos (in the Epirote toponym naA.aicrTT), which
Bekker, A necd. Gr. I p. 419, 1. 27, and Kretschmer believes (Ciotta 30, 1943, pp.
Souida), Bekker, Anecd. Gr. Ip. 299, 11. 16- 152-154) to be the probable origin of the
18, Didymus in Schol. Aristoph. A v. 836, Philistines. These and others have been
Dion. Hal. Antiq. Rom. A XXVIII, 4, Schol. compared with the Pelasgians (Ed. Meyer,
Aristoph. A v. 1139, and 832, Eust. Thess. Gesch. d. Altertums II, 1, p. 562). Berard
Schol. Dion. Per. 347. too has drawn this parallel between Pelas-
8. Schol. Luc. Bis Acc. 9. gians and Philistines on the basis of the
9. The word neAOO"TlKOV: even though Athenian toponym (Stud. Rob. p. 148). On
this is probably a spelling error, it should the subject as a whole, see G . L. Huxley,
be noted that the writing of n EAOO"TlKOS Early Sparta (London 1962) p. 98.
rather than neA.acrytKOS occurs in a num- 10. PA. pp. 25-29.
ber of Homeric manuscripts (Codd. B, L, 11. Stud. Rob. pp. 144-150.
T, Iliad 16, 233), where the reference is to 12. Herod. VI 137.
the Dodonean Zeus. According to one 13. L'Acropole Ip. 11.
20
THE ANCIENT SOURCES
term used in the 5th century inscriptions. Miller too employed the same argu-
ment basing it on other sources as well. 14
An examination of the sources shows that apart from the fragment of
Cratinos (528-423 B.C.), all the others using the form Pelasgikon are signifi-
cantly later. The earliest of all is the Parian Chronicle (264/3 B.C.). The others
are datable to the time of Christ (Strabo, 67 B.C.-A.D. 23) and later. The
Cratinos fragment, as it has reached us, is corrupt, 15 so it may well be that
the term Pelasgikon is not in its original form. In any case, all the fifth cen-
tury B.C. writers use the form Pelargikon. Texts are always liable to uncon-
trolled alteration but the inscription of 423/2 B.C. 16 provides concrete evi-
dence that the Athenians of Classical times used the form Pelargikon. This,
therefore, is more likely to be correct.
Thus, Pelargikon; but what was the exact meaning of the term according
to the Ancients? Rather than simply mentioning the word, the following
sources give some actual information.
Herodotos 17 refers to it as the fortification wall surrounding the Acropolis.
The same author 18 narrates that when the Athenian democrats, aided by
Cleomenes, the king of Sparta, made a stand against the tyranny of the sons
of Peisistratos, the tyrants took refuge within the Pelargikon, where they were
besieged. He adds that the stronghold proved impregnable because the defend-
ants had provided themselves with food and water. The same information is
to be found in the Parian Chronicle .19 Aristophanes alludes to it, 20 while Aris-
totle21 adds that Hippias, besieged within the Pelargikon and obliged in the
end to surrender and to leave Athens, handed over the Acropolis to the Athe-
nians. According to these sources, therefore, Acropolis and Pelargikon were
synonymous. A parallel notice is to be found in Thucydides' 22 account that
when Kylon and his companions plotted to seize power in Athens, they forti-
fied themselves and were besieged within the Acropolis. If the Peisistratids
14. AJA 1893, p. 482. 19. I. 60 (ed. Jacoby).

15. Wilamowitz, Hermes XIV, p. 183. 20. Lys. 1150.
16. Dittenberger, Sylloge 3 n° 83. 21. Ath. Pol. XIX 5-6.
17. VI 137. 22. I 126.
18. v 64.
21
THE ANCIENT SOURCES
used the Pelargikon for fortified protection in 510 B.C., the same holds all
the more for Kylon in 632 B.C.
The fortification itself continued to stand in its entirety down to the Per-
sian Wars. Herodotos23 says that the few Athenians who had remained in the
city when the Persians made their incursion in 480 B.C., secured the Acrop-
olis by closing the entrances with doors and beams (6up11cri TE Kai ~u:Ao1cr1).
He goes on to say that while the enemy, shooting firebrands from the Areo-
pagus, burned the wooden barricades, this small band successfully held their
own until some Persians succeeded in scaling the rock near the sanctuary of
Aglauros. Since the rock at that point was precipitous it had been left unpro-
tected; it was EµTipocr6e ... Tipo TflS aKpoTI6A1os, 0TI1cr6e 8E TWV TivAec.vv
Kai TflS av68ou· (in front of the Acropolis, behind the gates and the ascent).
In this way the attackers managed to take the Acropolis, killing the defend-
ers and burning the sanctuaries. The following year, the destruction was com-
pleted by Mardonios who, on leaving Athens, systematically knocked down,
burned and buried whatever of the sanctuaries, buildings and walls were still
standing. 24
Thus the fortification of the Acropolis, the Pelargikon, which encircled the
rock,25 had its gates and approach at the west26 and was a functioning strong-
hold until the time of the Persian Wars. Before this, Peisistratos had used it
as his headquarters during his tyranny. 27 It was systematically and totally
destroyed by the troops of Mardonios in 479 B.C.
After this catastrophe, the Acropolis fortification walls were built anew by
Themistocles and Kirnon along the line preserved today. What was left of the
old fortification was buried beneath the fill of Classical times. Even so the
Pelargikon continued to be mentioned, the term now specifying an established
area below the Acropolis, an uninhabited, open and forbidden space. 28 That
part of the fortification together with the area it included retained the name,
at least down to Roman Imperial times, while the Acropolis itself ceased to
23. VIII 51-53. 26. Herod. VIII 51-53.

24. Herod. IX 13. 27. Herod. I 59.
25. Herod. VI 137. 28. Thuc. II 17, Schol. Thuc. II 17.
22
THE ANCIENT SOURCES
be known as the Pelargikon. 29 During that time the restricted Pelargikon was
protected both by oracle30 and by legislation31 against all destruction or ex-
ploitation. Yet its importance waned through time, so that by Lucian's day it
appears no longer to have received any special attention.
It is clear that part of the older fortification system of the Acropolis
remained outside the line of the Classical wall. It was this space, together with
what was left of its progressively disintegrating walls, that retained the old
name of Pelargikon which had once signified the entire rock. Keramopoullos32
too accepts this interpretation and there can be no doubt that it is correct.
The entire question of the extent and boundaries of the Pelargikon in its
restricted meaning, has been the subject of much serious disagreement in the
bibliography. Most scholars have been interested in it as a problem in itself.
For this reason it is examined further in a separate chapter. To avoid confu-
sion in the present discussion about the Mycenaean wall of the Acropolis, the
wall that existed before the Persians is referred to as the Mycenaean fortifi-
cation wall or the Cyclopea'n wall. The term Pelargikon is here applied solely
to the section below the Acropolis that remained outside the Classical fortifi-
cation circuit.
In addition to the above sources that refer to the fortification of the Acrop-
olis, there are testimonia referring to the existence of other very ancient con-
structions within the area itself. We are told of the tomb of Kekrops, 33 at the
SW side of the Erechtheion beside the Porch of the Maidens. 34 Mentioned
too is a palace of Erechtheus, 35 as well as the most ancient mythical "tokens"
29. Luc. Pesc. 42-47, Bis Acc. 9, Philostr. 33. Clem. Al. Protr. III 45, Theodoretus,
Vit. Soph. II, a, V. 'Ell fJEe. na8. n Eei r:ij~ r:wv 'µaer:vewv' n-
30. Thuc. II 17. µij~ H , 30, Arnobius, Adv. Nat. VI 6.
31. Dittenberger, Sylloge 3 n° 83, Pollux 34. JG I 2, 372, 1, 11. 9, 59, 63, 84
Onom. VIII 101. (Erechtheum p. 268 f.). See also Erechtheum
32. Keramopoullos, PraktAkAth 1932, pp. 132, 136, 362. On the subject as a whole
pp. 111-112, and Ephemeris 1934/1935, p. see G. Mylonas, Ot ~amA.txol i:a<pm pp.
89. See also Frazer, Paus. II 356, and Dorp- 415-422.
feld, AM 1911, p. 72. 35. Hom. Od. 7, 80-81.
23
THE ANCIENT SOURCES
in the area of the Erechtheion: 36 the olive tree, the sea of Erechtheus and the
marks left by Poseidon's trident, which caused water to gush forth from the
rock (or, according to another version, the traces of the lightening strike that
killed Erechtheus ).
To summarize, we may conclude from the sources that the ancient Athe-
nians believed that during the time associated by tradition with the period we
now call Mycenaean, the Acropolis had fortification walls that had been built
by the Pelasgians. Within these walls, on the levelled top of the rock, stood
the palace of one of the mythical kings, Erechtheus, the tomb of another,
Kekrops, and tokens of the presence of the gods and their rivalry for patron-
age of Athens: the olive tree, the sea and marks left by the trident or the thun-
derbolt. All these things were located in the area later occupied by the
Erechtheion. Outside the fortification walls, and lower down, there was a
walled space that had retained the name Pelargikon from Classical to Roman
times.
36. Herod. VIII 55, Pa us. I 26, 5; I 27, Hesych. s. v. TiayKucpos and acrTfi EAaia.
2, Apoll. Bibi. III 178, Strabo IX 396,
24
EXCAVATIONS AND RESEARCH
Over the years, various excavations and studies have been carried out on
the Acropolis. Listed below are those which have resulted in the discovery and
identification of prehistoric finds in general and, in particular, Mycenaean.
1835-1836 Ross, with the help of Schaubert and Hansen, carries out excava-
tions on the krepis of the Parthenon, reaching the Mycenaean
level, which he observes, without, however, recognising it as such;
he characterises it as "schwarze feste Erdart." (L. Ross, Archiio-
logische Aufsiitze I, 1855, p. 89.)
1852-1853 Beule explores the entrance to the Acropolis. He mentions the
part of the wall that is S of the Propylaia, attributing it to the Pe-
largikon, and he finds the traces of the Mycenaean ascent west of
the bastion of Athena Nike. (E. Beule, L'Acropole d'Athenes I,
Paris 1853, pp. 83, 85.)
1864-1867 The Acropolis Museum is built. Recovered in the excavations
made in order to set its foundations are Mycenaean terracotta fig-
urines and pottery decorated with bands and tentacles. (W. Hel-
big, Bulletino dell'Instituto di Corrispondenza Archeologica 1875, p.
137.)
1876-1877 Excavations by the Archaeological Society on the S slope of the
Acropolis. A prehistoric (?) tomb is found W of the Asklepieion.
(Praktika 1876-1877, pp. 31-32.)37
1880 Bohn, in his studies of the Propylaia, refers to the section of the
wall to the S of it, and attributes it to the Pelargikon. He connects
it with the part of the Mycenaean bastion beneath the temple of
37. See also Skias, Ephemeris 1902, p. 123, n. 1.
25
Athena Nike that is visible through the opening left in the north
side of the Classical enveloping wall, and the section of the
Archaic crowning of the bastion further east, near the steps. He
notes also the traces of the ascent uncovered by Beule. 38 (R. Bohn,
Die Propyliien der Akropolis zu Athen, Berlin-Stuttgart 1882, pp. 15,
16, 29.)
1885-1890 The Archaeological Society, represented by Kavvadias and Ka-
werau, carries out systematic excavations on the Acropolis. They
discover practically all the Mycenaean remains. (P. Kavvadias - G.
Kawerau, 'H avaauacp~ rfj~ 'Aueo.noA.t:w~ a.no rov 1885 µiXQl rov
1890, Athens 1906.)39
1897 Kavvadias excavates the area of the caves on the NW part of the
Acropolis rock. He identifies the caves of Pan and Apollo and dis-
covers the steps of the NW descent, some remains of the sup-
porting walls connected with these, and a section of the N wall
without either recognising or referring to it. He discovers also the
beginning of the descent to the North Fountain. (P. Kavvadias,
Ephemeris 1897, pp. 1-32, pl. 1.)
1902 Excavation by Skias of a Middle Helladic grave mound on the S
slope. (A. Skias, Ephemeris 1902, pp. 123-130.)
1909 Exploration by Koster in the area of the caves. He notes a num-
ber of traces which he attributes to the Pelargikon. (A. Koster,
Das Pelargikon, Strassburg 1909.)
1915 Pelekides publishes some prehistoric and Mycenaean pottery from
the Acropolis. (E. Pelekides, Deltion 1915, suppl. pp. 35-37.)
1905-1920 Exploration of the Erechtheion by the American School of Clas-
sical Studies. Discovery of Mycenaean remains in and around the
foundations. (G. P. Stevens, J. M. Paton, L. D. Caskey, H. N.
38. Beule, L'Acropole p. 85. BCH 1888, pp. 244-245, Montelius, VHAM
39. See also AM 1886, pp. 162-169; 1889, pp. 46-60.
1887, pp. 141-142; 1888, pp. 107-108, 228,
26
Fowler, The Erechtheum, Cambridge, Mass. 1927, pp. 13-14, 122,

126, 138-142, 424-429, 580-581.)
1922 Excavations of the Italian School of Archaeology on the S slope,
with Neolithic, Early Helladic and Middle Helladic finds. (D. Levi,
Abitazioni preistoriche sulle pendici meridionali dell'Acropoli,
ASAtene 13/14, 1930-1931, pp. 411-498.) 40
1923 On the occasion of restoration work being carried out on the
Erechtheion and exploration in that area, Holland re-excavates
the Mycenaean foundation discovered by Kavvadias beneath the
paving N of the Erechtheion, and studies the other remains in the
area. (L. B. Holland, AJA 1924, pp. 1-23, 142-169, 402-434.)
1925 The prehistoric pottery from the Acropolis is published. (Graef-
Langlotz [Wolters], Die Antiken Vasen von der Akropolis zu Athen
I, Berlin 1925.)
1927-1928 Keramopoullos explores the N, W and S lower slopes of the
Acropolis. He identifies the cave of Zeus Olympios. (A. Kera-
mopoullos, Deltion 1929, pp. 73-101, PraktAkAth 1932, pp. 110-
124, Ephemeris 1934/1935, pp. 85-116, pls 1-3.)
1931 Excavation of the N slope is initiated by Broneer. The sanctuary
of Eros and Aphrodite is discovered and identified. (0. Broneer,
Hesperia I, 1932, pp. 31-55.)
1931-1932 Broneer finds the continuation of the NE ascent outside the Clas-
sical wall, covered over in places by floors of Mycenaean houses.
(0. Broneer, Hesperia II, 1933, pp. 329-417, pl. XI.)
1933-1934 Revelation of the NE ascent is completed as far as the Peripatos
and more houses are found. (0. Broneer, Hesperia IV, 1935, pp.
109-188, pl. I.)
40. See also Bd.A 1922-1923, p. 278 f. , 1924-1925, p. 88 f.
27
1936 Stevens publishes the results of his work on the terrace by the base
of the statue of Athena Promachos. (G. P. Stevens, Hesperia V,
1936, pp. 499-519.)
1936 Broneer explores the E cave. (0. Broneer - M. Z. Pease, Hespe-
ria V, 1936, pp. 247-272.)
1936-1939 Balanos, while restoring the bastion of Athena Nike, discovers the
Mycenaean bastion inside it. Welter also takes part in the work.
After Balanos' retirement, the work is continued by Orlandos, who
brings to light the rest of the S side of the bastion. (N. Balanos,
Ephemeris 1937, r [Athens 1956], pp. 785-791, 795-800, pl. 1, G.
Welter, AA 1939, pp. 1-22.)
1937 Prehistoric pottery is collected during the course of excavations on
the N slope (H. Hansen, Hesperia VI, 1937, pp. 539-570.)
1937-1938 Broneer discovers and excavates the North Fountain. (0. Broneer,
AJA 1938, pp. 445-450, Hesperia VII, 1938, pp. 168-170; and VIII,
1939, pp. 317-433.)
1937-1939 Excavations of the American School of Classical Studies in the
Klepsydra area, resulting in the discovery of Neolithic, Early Hel-
ladic, Middle Helladic and Late Helladic wells. (T. L. Shear, Hes-
peria VII, 1938, pp. 335-338; VIII, 1939, p. 221; IX, 1940, pp. 297-
298, figs 38-39.)
1938 Kolbe explores the Mycenaean wall by the SW corner of the
Parthenon and E of the Museum and collects material that dates
it. (W. Kolbe, AA 1939, pp. 227-236, FuF 1939, pp. 393-394 and
427-429, Bericht uber den VI Intemationalen Kongress far Archiiolo-
gie, Berlin 1940, pp. 344-346, pl. 27a-b, Research and Progress 1940,
pp. 253-259.)
1939 Exploration of the N slope continues. (0. Broneer, AJA 1940, pp.
252-256.)
1940 The American School of Classical Studies excavates the Klepsy-
dra, recovering a few Mycenaean finds. (T. L. Shear, Hesperia X,
1941, p. 7, A. W. Parsons, Hesperia XII, 1943, pp. 191-267.)
28
1946 Stevens publishes the results of his research in the area of the
entrance to the Acropolis. (G. P. Stevens, Hesperia XV, 1946, pp.
73-79, 102.)
There are thus two important landmarks in the exploration of the Myce-
naean Acropolis: first, the general excavation of the rock during the years
1885-1890, when most of the preserved remains were brought to light; second,
the period of 1932-1939, when, through the work of Broneer, Kolbe, the
American School of Classical Studies and Balanos, the picture was filled in
with the addition of finds from the N slope, the continuation of the NE ascent,
the North Fountain and the tower inside the bastion of Athena Nike. With
the material that was carefully collected and published, it was now possible
for the first time to date the construction of the Mycenaean fortification. It is
clear that these two landmarks in time must be seriously considered in any
judgement of past attempts to reconstruct the Mycenaean Acropolis.
29
IN THE BIBLIOGRAPHY
Theories about the form of the Acropolis in Prehistoric times were circu-
lating even before the Acropolis had been excavated. Concrete evidence lack-
ing, these theories were based chiefly on literary sources and on the few access-
ible remains that were thought to be prehistoric. The various theories are
indeed far from the actual state of affairs as known today, 41 but it should be
remembered that the Mycenaean civilisation was not yet really understood.
Two axioms were generally accepted at that time and on these all the the-
ories were based. The first was that the fortification, equated with the Pe-
largikon and, according to Cleidemos, with nine gates, was below the Acrop-
olis an~ included also the cave of Pan. The second was that the Acropolis
fortification · took in the city as well, that is, the inhabited area of Athens.
The first axiom was based on the ancient sources; the second was inferred by
analogy with historical times, but in ignorance of Mycenaean practice, which
was to leave the settlements themselves unfortified. Thus, three hypotheses
were possible: 1) the inhabited area was confined to the Acropolis, 2) the
Pelargikon extended to include the known area of ancient Athens, or else, 3)
a solution between these two, that the centre was the Acropolis and that there
was an area around it or next to it, the position and boundaries of which were
defined in accordance with various ideas including the location each scholar
might choose for the Eleusinion. Leake came out in favour of the first hypoth-
esis, according to which the Pelargikon comprised the Acropolis in itself and
the NW area of the slopes where the cave of Pan was located. 42 Also in favour
of the first hypothesis were K. 0. Muller, who accepted the Pelargikon as the
41. Nearly all are presented by Judeich, 42. Topography pp. 309-315.
Top. pp. 113-114.
30
THE MYCENAEAN ACROPOLIS IN THE BIBLIOGRAPHY
fortification of the NW end of the rock, the most vulnerable part of the Acrop-
olis, 43 Ross, who located the Pelargikon in front of the Acropolis, 44 Beule,
who considered the Pelargikon to be the fortification of the W side of the
rock with nine consecutive gates,45 Bursian, who had approximately the same
conception as Beule, 46 and Wachsmuth, likewise agreeing with this arrange-
ment.47 In support of the second argument were Welcker, who came out with
the idea of two separate Pelargikons, one on the Pnyx and one on the Acrop-
olis, 48 Gottling, who located the Pelargikon on the Pnyx,49 and Wilamowitz-
Mollendorff, who believed that the Pelargikon was the old city wall, and that
it protected the area froin the Areopagus to Hadrian's Gate. 50 The third the-
ory was accepted by Unger, who distinguished the Pelargikon as an inhabited
area on the SW slopes from the Pelargikon as a fortification on the E part of
the rock, 51 Davidson,52 E. Curtius, 53 and A. Botticher,54 all of whom considered
the Pelargikon to be a fortification taking in the lower slopes of the rock.
The excavations of 1885-1890, during which most of the Mycenaean
remains of the Acropolis were found, put an end to these hypotheses and specu-
lations. There was now enough specific evidence and it could not be disre-
garded. The sections of the fortification wall that had survived were enough
to give a general idea of the overall arrangement, and in particular to show
precisely where the citadel of Mycenaean times should be sought. The vari-
ous walls, graves and the ceramic finds added an indisputable assemblage of
material to the records of the ancient authors. So, if they presented fresh
problems to scholars, they provided at the same time a new basis on which to
found their theories.
43. Ersch u. Gruber 1, Sect. VI, 229 f. , A then.

and appendix to the German translation of 50. Burg u. Stadt pp. 97-172.
Leake (Halle 1829) p. 466. 51. Sitzungsber. Akad. Munchen 1874,
44. Die Pnyx und das Pelargikon (1853). pp. 263-351.
45. L 'Acropole pp. 80-84. 52. The Parthenon Frieze and Other
46. Philol. 1854, pp. 643 f. Essays (1882) p. 147 f.
47. Die Stadt Athen pp. 387-392. 53. SBBerlin 1884, pp. 499-512, and Ge-
48. AbhBerlin 1852, p. 267 f., RhM 1856, sammelte Abhandlungen I (1894) pp. 435-450.
p. 30 f. , 581 f. 54. Akropolis pp. 56-61.
49. Das Pelargikon und die Pnyx in
31
The existence of a fortification wall similar to other Mycenaean fortifica-

tion walls, which were already becoming well known, had been adequately
demonstrated. It remained to determine its precise extent and arrangement.
To begin with, there had been the idea that the Acropolis was fortified only
in those places where remains of the wall were preserved and wherever the
rock was easily accessible. Elsewhere, the rock was thought to have been quar-
ried in order to make it inaccessible. This idea was supported by Lolling, 55
Harrison,56 and E. Curtius, 57 who demonstrated also the existence of a palace
in the area of the Erechtheion, with houses for the courtiers to the W of it.
It was soon understood, however, that this perception of the fortification sys-
tem was mistaken, and that the wall, just as Herodotos said, 58 went around
the entire brow of the rock. The sections found were now seen as parts of an
entire circuit, which had not survived, but had been destroyed during the
building of the Classical fortification, which, especially along the N side, fol-
lowed the course of the Mycenaean wall. This opinion was developed by
Miller,59 who attributed the quarrying of the rock in the area of the Askle-
pieion to an attempt by the builders of the wall to prevent scaling by attack-
ers. He located the palace in the area of the Erechtheion, with courtiers'
houses to the west. Of this opinion also were Belger, 60 Dorpfeld, 61 Tsountas, 62
and Harrison. 63 D'Oodge64 too was in complete agreement with these new
ideas. The above studies, however, were confined to general observations,
unsupported and without giving details.
The first attempt to determine the line of the Mycenaean wall and to date
its construction was made by Koster (Fig. 1). Koster believed that it was dat-
able in the 2nd millennium B.C. and that it ran around the entire surface of
the rock. On the W side he proposed a straight line continuing along from
the section preserved S of the Propylaia. He did not accept the existence here
55. Topographie p. 337. 61. RhM 1896, p. 131.

56. M. and M. p. 536. 62. 'H 'AXQ0.7WAl~ TWV 'A817vwv pp. 5-
57. Stadtgesch. p. 45. 11.
58. VI 137. 63. PA. pp. 13-15.
59. AJA 1893, pp. 476-484. 64. Acropolis pp. 21-23.
60. BerlPhilolWoch 1894, p. 16 f.
32
•
Fig. 1. The Mycenaean Acropolis according to Koster (Pelargikon pl. Wb ).
of an entrance, and he categorically rejected the possibility that there was a

bastion in this place. According to Koster, there was an entrance at the NW,
where in Classical times a stairway led to the caves and communicated with
the Klepsydra. He deduced the shape of this entrance from the section of the
Prehistoric wall preserved next to the Classical stairway. With the descent
leading to the caves he connected traces of walls, which he believed to be parts
of terrace walls supporting the last part of a ramp. The main entrance he
believed was the one on the NE next to the Erechtheion, which in the begin-
ning, as he says, continues along the rock toward the W, later turning toward
the S. In addition he claimed that toward the end of the second millennium
B.C., the fortification was extended westwards in order to include the Pelar-
gikon in the pre-existing fortification. The NE gate was then abandoned and
another one was opened to the west. 65 Kawadias agreed with these conclu-
sions.66
65. Pelargikon, especially pp. 5-16, 27 66. flQoiaro(}txi/ 'AQxawA.oyia p. 300.

and pl. IVb.
33
Koster's observations are in general sound but inadequate. His conclusions

are arbitrary and occasionally contrary to fact. Specifically, they run counter
to the rock formation. Thus, the line he suggests for the N leg of the
Pelargikon67 runs right through the cave of Pan; the change in orientation of
the NE ascent toward the S is impossible because the rock is precipitous in
that place, and also because, as we shall see, that passage is blocked by the
Mycenaean wall. The point in his study that aroused the most opposition, how-
ever, was the form of the W side of the fortification. Heberdey68 disagreed on
the basis of Kawerau's observations on the interior of the Pinakotheke and
on the fill of the Archaic cistern to the north of it. He showed that the W wall
formed a curve taking in the area of the Pinakotheke and that there was an
entrance that had been hidden by the Classical Propylaia. Pfuhl 69 had similar
reservations.
Even so, the work of Koster was a serious contribution to the study of the
Mycenaean Acropolis and it was generally accepted that the fortification wall
circumscribed the top of the rock along the line indicated by the sections of
the wall preserved. 70 After its publication, discussion turned to details, espe-
cially to the plan of the west entrance.
The next and perhaps most important contribution to the entire subject
was L. B. Holland's study of the Mycenaean remains in the area of the Erech-
theion. Availing himself of the opportunity provided by the Americans' study
of the Erechtheion, he excavated anew the wall covered over by the slab
paving to the north of that building. He dated it, and at the same time exam-
ined the other prehistoric buildings in the area. 71 With acute observation,
architectural for the most part, he divided these walls into three consecutive
phases, attributing them not to the palace itself, but to terraces on which the
palace had been built and to the remains of a bastion protecting the gate to
which the approach led. Judging by the construction of the wall that closed it
67. Pelargikon pl. IVb. and Schede, Die Burg von Athen p. 10.
68. 6Jh 1910, pp. 1-4, fig. 1. 71. AJA 1924, pp. 1-23, 142-169, 402-
69. BerlPhilolWoch 1911, pp. 299-307. 434.
70. See Frazer, Paus. II pp. 355-356,
34
Fig. 2. The arrangement of the area of the NE ascent according to Holland (AJA
1924, II, fig. 2).
off (Fig. 2), he recognised quite rightly that the gateway had already been
blocked during Mycenaean times. From traces on the poros slabs of the foun-
dation of the N wall of the Erechtheion and two small sections of limestone
wall, 72 he determined the existence of yet another Mycenaean terrace wall and
suggested that the area N of the Erechtheion had been used as a sort of arena
for sacred ceremonies, comparable to the steps of the palace at ·Phaistos.
Based on the interaxes of the two preserved poros column bases, which he
accepted as Mycenaean, he located the palace, specifically the palace me-
garon, in the area S of the Erechtheion. He recognised, furthermore, that
some of the walls found W of the Erechtheion and interpreted as prehistoric
foundations of courtiers' houses, were in fact much later, probably even Medi-
aeval. He dated the construction of the foundation north of the Erechtheion,
which he excavated down to the first phase of the supporting walls, and the
bastion by the entrance, to the 15th century B.C., on the basis of pottery from
both these excavations. He showed that the fortification wall (which was not
precisely dated) was significantly later.
72. See also Erechtheum pp. 137-142.
35
Holland's study was extremely important. This was the first time that such
acute and bold observations had been made, based on a method that was obvi-
ously correct and original. Recognition that most of the walls were terrace
walls contributed significantly to a correct interpretation of the remains NE
of the Erechtheion. Yet the fact that the entire work was carried out on the
basis of plans rather than on the few remaining and accessible walls them-
selves, led to a number of misconceptions. Moreover, his bold reconstruction
of the palace on the basis of very few facts, cast doubt on his other conclu-
sions; doubts which, it must be noted, never took the form of concrete objec-
tions. Holland based his work on two assumptions: that the walls NE of the
Erechtheion were all Mycenaean, and that the poros bases, which he attrib-
uted to the megaron of the palace, were both Mycenaean and in situ. As we
shall see below, these assumptions were mistaken and, as a result, his con-
clusions were not always right. Yet his publications constitute the most seri-
ous study of the Mycenaean remains on the Acropolis up to his time, if only
because he determined that the walls were terrace walls, to be distinguished
from the later fortification wall, and because he was the first to express justi-
fiable doubts that the walls of the so-called "settlement of the Eupatridai"
were really Mycenaean.
After Holland, nothing significant was added to the subject. Picard
accepted that the rock was surrounded by a strong fortification wall that in its
western part followed virtually a straight line, with a main entrance on the NE
and a second one on the NW leading to the Klepsydra. He accepted also that
to this wall, which surrounded the palace and dwellings of the aristocracy, the
Pelargikon and an entrance to the W were added during the llth and lOth
centuries B.C. 73 His argument was simply a combination of Koster's theories
with Holland's observations. Judeich 74 cautiously confined himself to some
general observations and Dorpfeld75 contributed nothing new to the discus-
sion.
73. L 'Acropole I pp. 11, 19-20. 75. Alt Ath. p. 3.

74. Top. pp. 54-55, 114, 115.
36
The discoveries of the period 1932-1939 added much important material.

Three serious problems were resolved: the formation of the W entrance, the
water supply of the fortress and the date of the Cyclopean wall. The discov-
eries of 1932-1939, moreover, contributed greatly to the general picture of the
Mycenaean Acropolis. As might be expected, the publication of these discov-
eries was followed by works of a general nature. None of these explored the
subject of the Mycenaean Acropolis as a whole. Instead, they all were limited
to inserting the new finds into the old frameworks or to examining various
special subjects, particularly those arising from the addition of the new finds.
Thus Welter, after the discovery of the Mycenaean bastion under the temple
of Athena Nike, argued in favour of connecting it with the already known for-
tification system of the W side. At the same time he proposed an arrange-
ment of the nine gates which went against the facts.7 6 Stevens, on finding the
traces of a Mycenaean terrace wall behind the base of the statue of the Pro-
machos, restored the entire western part of the terrace wall 77 without, how-
ever, connecting it with the rest of the terracing, the existence of which had
already been demonstrated by Holland. He examined and also drew the W
entrance of the fortification in connection with the bastion that had been
found and the trace of the W fortification wall, which he discovered in situ. 78
His solution was far more logical and plausible than Welter's. Yet it had a
serious drawback in that it left the bastion isolated and unconnected with the
wall.
To these studies we should perhaps add a restored drawing of the Myce-
naean wall shown by Dinsmoor in a text referring to the Archaic Acropolis. 79
The wall is shown in general along the line accepted by Koster, with one
change in the layout of the NW part, which runs in a direction suggested by
the Archaic propylon.
Broneer's observations,80 based primarily on the discovery of the North
Fountain and the NE approach, are completely sound. He approached the prob-
76. AA 1939, pp. 7-9, fig. 4. 79. AJA 1947, fig. 3, p. 122.
77. Hesperia V, 1936, pp. 499-503, 519. 80.AJA 1948, pp. 111-112, and Antiquity
78. Hesperia XV, 1946, pp. 73-79, fig. 2. 1956, pp. 9-13.
37
lem of the historical sequence of events with arguments based on a thorough

knowledge of the material, concluding that the wall should be dated in the
13th century. His publications, however, are devoted chiefly to the history of
the period rather than to the morphological development of the Acropolis.
Berard, interested mainly in a critical study of the ancient sources and the his-
torical problems connected with the descent of the Dorians, was not con-
cerned with a synthesis of the material except along very general lines. 81 In
her work on ancient Athens, Hill noted the various remains that had been dis-
covered without giving reasons for the way she relates them. In the accom-
panying plan the W bastion remains unconnected to the fortification wall, and
the NE entrance follows the turn toward the south that had been proposed
1.TDAY/\01:
19::>7
Fig. 3. The Acropolis of late Mycenaean times according to Travlos

(IloAEOD. fig. 7).
by Koster. Finally, Travlos82 believed that the wall had two periods of con-
struction. In the first, the wall takes in the top of the rock and goes back to
the 15th century B.C., with one entrance at the Wand another one where the
81. Stud. Rob. pp. 135-159. 82. IloAEOO. pp. 22, 24-26, fig. 7.
38
THE MYCENAEAN ACROPOLIS IN THE ·BIBLIOGRAPHY
NE approach ends. In the second period, in the 13th century B.C., the wall
circumscribes the entire rock. During this time the existing NE entrance is
closed, the NW postem gate toward the caves is opened and the W bastion
is built; the bastion he connects to the straight section of wall to the E of it,
i.e. to the piece preserved today S of the Propylaia. As for the rest, he defines
the course of the wall more or less along the accepted line (Fig. 3).
In general, the studies published after the researches of 1932-1939, which
brought about a real change in established perception, are those concerned
with the plan of the W entrance and especially with the incorporation of the
newly discovered bastion into the fortification system. For the rest, the dis-
covery of the North Fountain solved the problem of how the Acropolis was
supplied with water and the findings of Broneer and of Kolbe contributed to
the correct dating of the wall. Its line, however, was not modified, nor was
any attempt made to study that problem in detail.
39
CONCLUSIONS
It should be clear from the above that the ancient sources alone cannot
give a complete picture of the Mycenaean Acropolis. As is evident from pub-
lications before the big excavation of the Acropolis, the literary sources con-
tain only general information that may be interpreted in various ways, often
contradictory.
Yet the progress of research has led many scholars, especially during the
past few years, to examine a number of special topics. Most of this work has
been carried out with meticulous attention to detail. The results have been
exceptionally interesting but, as a rule, they pertain to special questions being
studied by the particular scholar at the time. The material collected in that
way was valuable and it contributed much topographical and chronological
information. Yet serious gaps remained, which no one undertook to examine.
To some extent this was due to the acceptance of various ideas, which were
given the authority of self-evident truth by time and constant repetition.
To date, two studies only rank as basic works: Koster's book about the
Mycenaean wall and Holland's publications on the area of the palace and on
one section of the wall. Koster was corrected on many basic points immedi-
ately after the publication of his work and subsequent discoveries showed him
to be mistaken. Holland had so little factual evidence as a basis for a number
of problems, that his views were met with considerable reserve.
Whether or not their conclusions were correct, will subsequently be shown.
Yet it is a fact that since then there has been no other systematic examina-
tion or analysis of the evidence as a whole.
That the Acropolis was walled during the Mycenaean period by a continu-
ous fortification, of which quite a few stretches have been preserved, is gen-
erally accepted. The main entrance to the fortification was at the W where it
was protected by a strong bastion. The construction of the North Fountain
made it self-sufficient in drinking water. On the N side of the plateau at the
top of the rock, where the Erechtheion later stood, a system of terraces sup-
40
CONCLUSIONS
posedly supported the palace. W of that complex are walls attributed to build-
ings of the same time. The tombs and house foundations that were found at
various other places on the rock suggest that the area was systematically inhab-
ited. In addition, ancient tradition holds that some of the most ancient sanc-
tuaries of Athens were in this same area. There is also the question of the
exact location, extent and boundaries of the Pelargikon, a problem that has
led to much discussion and to which most of the bibliography is devoted.
In the following pages these problems will be examined on the basis of
actual remains throughout the area in orde~ to compose a picture of the Myce-
naean Acropolis based as much as possible on tangible and verifiable facts.
41
PART TWO

ON THE BASIS OF THE FINDS
THE ROCK
The Acropolis rock consists of a large ellipsoidal mass of cretaceous -

upper jurasic limestone with neritic traces over a layer of Athenian schist
(kimilia). To the Wand E there are deposits of breccia adhering to limestone,
which elsewhere is deposited on the argillaceous schist mass in surface slides. 83
The soft argillaceous schist layers on which the limestone is deposited have
been eroded by moisture. As a result, pieces of the rock have pulled apart
from time to time and broken off from the main mass, thus creating its pre-
cipitous slopes. The N side of the rock is the most exposed to the effects of
weather, and here the erosion is more evident and has progressed further.
This destruction is continuous so that technical buttressing has been necess-
ary in many places. It is erosion of this sort that has formed the various caves,
large and small, in the rock.
The precipitous slopes, which make the Acropolis the best natural strong-
hold of the area, have not the same formation everywhere. At approximately
the middle of the W side, the gradient is less and the schist lower down slopes
also gently uphill. Thus the incline here is gradual and this side provides the
only relatively accessible approach to the rock. At the SW corner, however,
there is a projection toward the W which is quite steep (Plan 1, 1), and at the
NW corner the projecting rock drops off almost vertically (Plan 1, 2).
The N side is the steepest and the most eroded. Toward the W end, below
the brow of the rock itself, there is a second lower and narrower level space,
opening along the face of which are a series of caves (Plan 1, 3). Three of
these were later dedicated to the cults of Apollo, Zeus Olympics and Pan.
From the surface of this level area a rather uneven ascent leads eastward to
the top of the rock, where later one of the entrances in the wall was built.
To the E, this small level stretch is closed by a high and narrow piece of
the rock. This eroded down to the bottom, broke off and came to rest at a
83. See especially Philippson, Die griechischen Landschaften I, 3, pp. 894-895.
45
THE ROCK
Fig. 4. The large piece of the rock broken off at the NW part of the Acropolis.
slight angle (Plan 1, 4), so that the top of it leans almost entirely on the main
mass, while the bottom rests further out (Fig. 4). The space opened between
this and the rock itself is wide enough to form a passage, which is narrower
toward the W. Toward the E, it widens out to form a cave-like opening in the
spot where the sanctuary of Aglauros has been located. Within this crevice
the North Fountain was opened.
At about the middle of the N side, at the base of the rock where many
large pieces have broken off and slid down from time to time, a deep crevice
opens toward the E (Plan 1, 5) giving access to the top of the rock. From this
point on around to the NE corner the incline is practically vertical. Along the
E side it is gentler, but not enough to be easily passable. At approximately
the middle of this side is the mouth of a large cave (Plan 1, 6), the largest in
the entire rock. The lower part of the S side is practically vertical, excluding
46
THE ROCK
156,16
........ =•-a
1
oat a a AZi4114wa
EAST-WEST SECTION
N
105~ 110.../
115
+
NORTH-SOUTH SECTION
105 105
0 10 50 100 150 200 250 300 350
Plan 1. The Acropolis rock.
47
THE ROCK
any possibility of scaling (Plan 1, section N-S). The impression made by this
vertical face is even more striking in the area of the Asklepieion, where the
rock has been quarried, evidently at the same time as the construction of the
sanctuary. At about the middle of the rock face, the incline has a gentle but
definite slope up to an elevation of about 153 m. above sea-level. Here it forms
a second, less marked brow. There is no level ground anywhere on the entire
slope. Opening along it are a few caves, most of them small.
The top of the rock forms a plateau surrounded by precipitous slopes, but
it is far from being level. It rises gradually from W to E and less gradually
from S to N (Plan 1, sections E-W, N-S), so that the main plateau occupies
the middle of the rock, levelling out to the NE. Next to this area the rock
makes a small rise, roughly oval in shape, the top of which at 156,16 m. above
sea-level is its highest point. 84 Thus the top forms a smaller oblong plateau
bounded on the W, N and E by the edge of the rock, and at the S by the
beginning of the slope to the edge. This plateau is 270 m. long, 94 m. wide
and has an area of around 15.000 sq. m. If to this is added the slope down to
the brow toward the S, the width becomes 140-150 m. and the area increases
to around 23.000 sq. m.
The levels along the brow of the rock on the W side vary from 142,40 m.
at the SW corner to 138,40 m. at the NW. On the N side, they range between
84. Judeich, Top. p. 43, and Travlos, flo- public areas of Athens (Kavvadias-Kawerau
A.t:oo. p. 6, accept 156,20 m. as the height, pp. 55-57). That original measurement, how-
based on the measurements of Kiepert, ever, was not correct, and for this reason
made in 1875. These measurements are Kawerau's measurements in general show
meticulous and accurate. They were based, a consistent difference of an additional
however, on a datum point on the coast of 0,60 m. On realising these differences, I had
Piraeus determined as zero-level, which new measurements made with a level, using
more recent and more systematic observa- as a basis the trigonometric point n° 109
tions have altered somewhat. Kawerau on (elevation 157,580 m. above sea-level)
the other hand depended on the elevation placed on the Belvedere. This point is part
taken on the threshold of the Beule gate, of the new network used for the Athens
which was given him by the French Mission area and it is based on the newly deter-
of Public Works, and which was taken in mined zero-level. All altitudes given here-
1885-1890 during the planning for various after are based on these measurements.
48
THE ROCK
137,71 m. above the cave of Pan, 147,57 at the top of fissure 5 on Plan 1 and
152,40 m. at the NE comer. On the E side they fall from 152,35 m. by the NE
comer to 150,28 m. at the SE. Finally, the brow of the S side rises from 132,50
m. at the E to 144 m. at the W. Thus the NE part of the rock is the highest.
There are no traces of levelling on the top of the rock before the Classi-
cal period, and the levelling of Classical times was sp<;>radic and on a small
scale. Thus the ilTie8isov of Cleidemos85 must have a meaning other than the
cutting off of the surface protrusions of the rock.
If the top of the rock was to be inhabited, the most suitable place was the
high level area toward the NW. To this led the relatively regular approach
from the W. It was also accessible from the NE through a fissure that was dif-
ficult to climb (Plan 1, 5), From the NW part a path connected it with the ter-
race of the caves (Plan 1, 3). There is no other approach anywhere.
There is no water on top of the rock. Further down, veins formed between
the limestone and the schist emerge at various points around the base of the
rock in the form of small springs. A few of these flow throughout the year.
Thus, at about the middle of the S slope there is the spring of the Asklepieion,
and at the NW the underground spring of Klepsydra. This has a small amount
of brackish water, which does not reach the surface but collects in wells that
were dug in the area from time to time. Finally, there is also the spring that
emerges at the base of the space between the main rock and the piece that
has pulled away (Plan 1, 4). Precisely because it had no exit, this was .a more
plentiful water supply than the others, reaching a level some 4 m. above that
of Klepsydra and 5 m. above the level of the Asklepieion spring. 86 This source
is the only one directly accessible from the inhabited top of the rock.
Thus, of all the hills of Athens, this rock was the most suitable for use as
a citadel because of its form and location. Lycabetus is very high and peaked
at the top, the Areopagus small and low and the complex of hills of the
~ymphs and the Muses too spread out and easy of access. The Acropolis rock,
naturally unassailable, of the right size and shape and encompassed by springs,
was indeed the appropriate and obvious choice as a place to live and as a
refuge in time of danger.
85. Seen. 5.
86. Broneer, Hesperia VIII, 1939, p. 429, n. 194.
49
THE ACROPOLIS DURING PRE-MYCENAEAN TIMES
As early as Neolithic times, the Acropolis and the area around it was inhab-
ited. Localised and clear traces of this settlement have been found at two
places below the base of the rock. On the S slope, to the W of the Asklepieion
(Plan 2, 1) the Italian excavations of 1922 uncovered a small Neolithic house,
and a fair amount of pottery of Sesklo type indicating the existence of a set-
tlement.87 A stone axe of the same period was recovered 88 from the area of
the Odeion of Perikles. Near the Klepsydra, the Americans found 21 wells
scattered over a fairly wide area, 89 which had been used as dumps when they
ran dry. These were full of sherds of the Late Neolithic period, with some
from the beginning of Early Helladic times (Plan 2, 2). With the sherds were
obsidian blades, stone utensils and also animal and fish bones. Groups of sim-
ilar sherds were collected as well from various natural hollows of the rock near
the wells; they constitute sound evidence that the area was inhabited in an
organised fashion. A number of Neolithic sherds were found mixed with later
material in disturbed layers of the N slope, 90 and a few pieces came from the
top of the rock. 91 These are sparse remains. They are, however, sufficient to
show that during Neolithic times, both around the Acropolis and on it, a rural
population was established in permanent settlements, had household animals
and used implements of stone and bone. 92
87. Levi, ASAtene 1930/31, pp. 411-498 221 and IX, 1940, pp. 297-298 and figs 38-
and particularly 482-484. 39), Parsons (Hesperia XII, 1943, p. 206)
88. Kastriotes, Praktika 1914, p. 95. does not give the number. Travlos (lloA.EoO.
89. There were 21 of these wells accord- p. 20) mentions 14.
ing to the day books of the American exca- 90. Hansen, Hesperia VI, 1937, pp. 540-
vators, which I had the opportunity of con- 542.
sulting. Of these, Shear reports 17 (Hespe- 91. See Appendix II, Group 3, a.
ria VIII, 1938, pp. 335-338; VIII, 1939, p. 92. See also Blegen, HSCP p. 6.
50
From the Early Helladic period, the finds are more copious, showing a sys-
tematic and uninterrupted use of the Acropolis area. Their findspots co-
incide, as a rule, with the Neolithic. On the S slope directly above the Neolithic
house and in small caves in the rock (Plan 2, 3), the Italian excavations
revealed evidence of habitation going back to Early Helladic times. For the
most part this was pottery belonging to the category of Urfirnis painted ware.93
Contemporary sherds and tools were collected from disturbed levels in the
area of the precinct of Dionysos. 94 Near the Klepsydra, sherds of this same
time formed the upper deposit 95 in some of the Neolithic wells. On the N
slope, in the area of the sanctuary of Aphrodite and Eros and in various other
spots, Early Helladic sherds were found in plenty, mixed with pottery of other
periods. 96 From the W slope comes the askoid vase with incised decoration
published by E. Pelekides, probably also another five vases of the same period,
for which the findspot is not given. 97
The excavations of 1885-1890 on the Acropolis yielded obsidian blades and
a number of sherds, most of them with incised decoration. 98 To these should
be added the sherds found by Kolbe inside the Mycenaean wall E of the
Museum, 99 those found by the Americans in the Pandroseion area in the low-
est level of the fill, practically on the rock, 100 as well as a few collected sub-
sequently at various places on top of the Acropolis. 101
Thus the Acropolis, which had been inhabited to some extent in the pre-
vious period, continued to be used more intensively, probably by the same
people. The name /\61lva1 or probably /\6'fivl'l which is indeed of prehellenic
origin, may well have been employed for the first time by these inhabitants. 102
93. Levi, ASAtene 1930/31, p. 490. 97. Deltion I, 1915, suppl. pp. 34-35,
94. These finds come from recent ex- fig. 1.
ploratory excavations of the Archaeological 98. Furtwangler-LOschke, M. V. p. 35,
Service, which are still in progress. A re- and Graef-Langlotz I n°5 1-9.
port has been published in the newspaper 99. AA 1939, p. 235.
"Kathemerini" for 27 February 1962. 100. Erechtheum p. 581.
95. Shear, Hesperia VII, 1938, pp. 335- 101. See Appendix II, group 1, a-b, group
338. 3, b, group 4, a (figs 50-51).
96. Broneer, Hesperia I, 1932, p. 35 and 102. Blegen, HSCP p. 2.
Hansen, Hesperia VI, 1937, pp. 542-546.
51
110__.../
105----------
• • •
•
r . .. ~~- :ttTI
!' ;:t>
:
/_.J n
::0
"~ ~,
0
"• ''"•.,
'"C
0
\, .. ,,,, '''\,,,~~ t:
..... ... en
0
c::
Vl
,.,.•''' '·~
..,.
-z
::0
0
N
' ·- '
'"C
::0
tTI
~
n
tTI
/ z
~
"~~
115
110
~
:j
0 10 50 100 150 200 250

/____:__
300 350
~
tTI
en
Plan 2. The Acropolis in Neolithic, Early Helladic and Middle Helladic times.
• NL and EH wells • MH wells.
After the first centuries of the Early Helladic period, there appears to have
been a gap in the habitation of the area. This, at least, is suggested by the
finds, which leap from relatively early Early Helladic to advanced Middle Hel-
ladic times. Yet this gap, which may be simply coincidental and which may
well be filled in the future by new finds, is the only such break the long his-
tory of the Acropolis has to show. For from the advanced Middle Helladic
period on, continuity of habitation is unbroken. The belief of the Athenians
that they were autochthonous evidently rested on this very continuity. 103
The Middle Helladic inhabitants left traces of their presence over practi-
cally the entire area of the Acropolis and its immediate environs. On the S
slope, the Minyan pottery found by the Italians, 104 the grave mound (Plan 2,
1) with 6 burials and funeral gifts consisting of obsidian arrow heads and a
handmade jug, from the fill of which Skias collected Minyan sherds, as well
as the matt-painted pottery found at the SW foot of the rock, 105 all bear wit-
ness to the use of the Acropolis area after a long interruption. 106 Discovered
further south, at the corner of present-day Kallisperi and Parthenon streets,
were two small graves and house walls, somewhat later than the graves,
belonging to the final years of Middle Helladic times. 107 In the area of the
Odeion of Perikles, J. Travlos' excavation brought to light a number of Minyan
sherds 108 and the precinct of Dionysos yielded many examples of all types of
MH pottery, with some pieces going down into the first LH years. These come
from undisturbed levels within the cella of the later temple and from a point
slightly SE of this. 109 Five new wells were dug next to the Klepsydra, deeper
103. Herod. VII 161, Thuc. I 2, II 36, Archaic" ()..Lav UQXa°Lld'j~ xataaxEui]~) ,
Eur. Erechth. fr. 360, 7-8 (Nauck), Plato which held a few "exceedingly crumbly"
Menex. VI, Isocr. Paneg. 24. bones (Praktika 1876-1877, pp. 31-32).
104. Levi, ASAtene 1930/31, p. 488. 107. I owe this information to G. Don-
105. Skias, Ephemeris 1902, pp. 123-130 tas, the Ephor who excavated the building
and Furtwangler-LOschke, M. V p. 34. lot.
106. To this period should perhaps also 108. Praktika 1951, p. 44 and fig. 3.
be ascribed the burial found in 1876 W of 109. Newspaper Kathemerini 27 Feb.
the Asklepieion. Discovered in a hollow 1962.
dug into the schist was a clay vessel, "very
53
and far better constructed than the earlier ones. These were found filled with
matt-painted and Minyan pottery. 110 The N slope has likewise yielded much
pottery, all from disturbed levels. 111
The finds from the top of the rock are mainly ceramic. They come from
the excavations for the foundations of the Museum, from the excavations of
1885-1890112 and from other, more recent explorations. Some were also col-
lected from various spots, especially the area of the Erechtheion (Plan 2, 4) 113
and from inside the Mycenaean fortification wall E of the Museum. 114 These
sherds, matt-painted and grey and red Minyan, belong to practically all the
known categories of developed Middle Helladic pottery. Their precise finding
places are not always known. Other evidence, of a concrete nature, comes
from the rock itself, showing that it was used from that time on as a place of
permanent habitation.
During the big excavation of the Acropolis, five small cist-graves of chil-
dren were found at three different places, as shown on Plan 2, numbers 5, 6,
and 7. 115 Each was constructed with five stone slabs, one vertical on each side
and a horizontal cover slab. The rock formed the floor. None had funerary
offerings and no bones were preserved. Form and construction of the graves
are characteristically Middle Helladic, as is evident from the excavation of
Eleusis in particular. 116 That there were no grave offerings whatsoever is more
likely an indication that they precede late Middle Helladic times than because
they were the graves of children.
110. Shear, Hesperia VII, 1938, pp. 335- Graef-Langlotz I n°8 10-31. See also Stub-
338, Parsons, Hesperia XII, 1943, p. 206. bings, BSA 1947, p. 4.
Shear reports 8, but in the day books of the 113. Judeich, Top. p. 52, Holland, AJA
Agora excavations 5 are mentioned, with 1924, p. 155.
the references OAE, OAM, OAN, OAQ, 114. Kolbe, AA 1939, p. 235.
OAT. 115. Kawadias-Kawerau pp. 31, 77-79,
111. Broneer, Hesperia I, 1932, p. 35; II, 143.
1933, pp. 359-363, figs 28-34, Hansen, Hes- 116. See Mylonas - Travlos, Praktika
peria VI, 1937, pp. 546-557. 1952, p. 59 and Mylonas, Eleusis p. 31.
112. Furtwangler-LOschke, M . V p. 34,
54
Located on the Acropolis by the Athenian myths were the grave of

Kekrops and, on the S slope, the graves of Talos or Kalos 117 and Hippoly-
tos.118 It has been observed repeatedly that myths of this sort are usually built
around an existing and substantial fact, the true historical basis and origin of
which are hidden in time. The myths about Hippolytos and Talos, who was
said to have been the nephew of Daidalos, are clearly very ancient. They took
form perhaps late in the Mycenaean period, but certainly in Mycenaean times.
It is surely not by chance that these myths located the graves of heroes in
places where, as excavations have shown, there were graves so old that even
for the Mycenaeans themselves they would have been a faint memory at best.
They may have been forgotten altogether until some chance find drew them
to attention. Be that as it may, it is more than likely that the graves ascribed
to these mythical figures were actual graves, Middle Helladic, and similar to
those found in the same areas. People may have believed that they held the
remains of those old heroes themselves. 119
117. He is referred to as T aAws in practice before historical times (see also

Apoll. Bibi. III 214, Diod. IV 76, 4 and Luc. Mylonas, Stud. Rob. pp. 64-105, Tiµrrnxo r;
Pesc. 42. Paus. (I 21, 4)) employs the form r6µor; 'A. 'A..1.t,Bt~arov (Athens 1958) pp. 3-
Ka:\ws. 9, Eleusis p. 62, Marinatos, Ephemeris 1933,
118. See, among others, Paus. I 22, 1. pp. 97 ff. , Praktika 1953, pp. 239-240, 244-
119. This does not mean that there was 245, Altertum I pp. 147-148). It means sim-
worship of the entombed dead from then ply that a number of existing graves have
on, for as far as is known this was not a been ascribed to mythical people.
55
THE MYCENAEAN PERIOD
I. THE LATE HELLADIC REMAINS ON THE ACROPOLIS
It was during Late Helladic times that the Acropolis finally took form not
simply as a place of habitation, but as an organised fortified entity. It now
became the seat of the ruler of the area and possibly also a refuge for its
inhabitants in case of danger. The traces left by this use of the rock, as we
shall see, are enough to provide a clear picture of the entire complex. Before
we can follow these traces step by step and in detail, a preliminary general
examination of them is in order so that we can distinguish them from other,
later remains that have in many cases been ascribed to Mycenaean times.
The visitor entering the Acropolis encounters first of all the Mycenaean
bastion. This is hidden by the Classical bastion of Athena Nike and it was dis-
covered and published by Balanos and Welter. 120 To the W of the bastion and
lower down, traces of the Mycenaean ascent are visible on the surface of the
rock. 121 The polygonal wall W of and on the same axis as the Propylaia, and
mentioned by all the authors, is much later. 122 It remains from an Archaic sup-
porting wall the continuation of which is visible much further west, outside
the Beule gate. 123 Likewise Archaic, as Keramopoullos discerned, are the var-
ious other walls on the SW side of the rock between the Asklepieion and the
120. Balanos, Ephemeris 1937 r , pp. 131 ff., Keramopoullos, Ephemeris 1934/35,
776-807, and Welter, AA 1939, pp. 1-22. p. 87, Stevens, Hesperia XV, 1946, p. 77,
121. Beule, L 'Acropole p. 85. Beule, L 'Acropole p. 84, Bohn, Prop. p. 15,
122. Miller, AJA 1893, p. 486, Welter, Botticher,Akropolis p. 59, Weller, AJA 1904,
AM 1923, p. 195, Keramopoullos, Deltion p. 60.
1929, p. 74, n. 3, Picard, L'Acropole Ip. 17, 123. Keramopoullos, Ephemeris 1934/35,
Judeich, Top. p. 213, Dorpfeld,AM 1885, pp. p. 87.
56
THE LATE HELLADIC REMAINS ON THE ACROPOLIS
Odeion of Herodes Atticus. They had been attributed to the Mycenaean

Pelargikon. 124
To the north of the Nike bastion, in front of the Propylaia, part of the
curved section of the W Cyclopean wall is preserved. 125 Further north, inside
the Pinakotheke, stones from this wall were found as well as intact Mycenaean
levels and the walls of a house. 126
Outside the Classical fortification of the Acropolis, near the base of the
big Mediaeval buttress of the north wall and N of the Archaic cistern, is a
series of stones facing north. Kavvadias included these in his plan of the area
of the caves without referring to them in his text. 127 Koster 128 interpreted the
stones as the beginning of the N stretch of the Pelargikon, while Judeich 129
considered them to belong to a house wall. Further west, near the NW descent
to the caves, part of the Mycenaean wall is preserved,130 as well as the pre-
historic descent itself. 131 The beginning of the descent to the North Fountain
can be seen 132 still further W, beside the house of the Arrephoroi.
Preserved in the area to the N, NE and E of the Erechtheion is a complex
of walls, a section of the N Cyclopean wall and the end of the NE ascent.
These remains were ascribed to the palace. 133 The explorations of the Amer-
124. D'Oodge, Acropolis p. 25, Botti- 129. Judeich, Top. p. 116, n. 2.

cher, Akropolis p. 57, fig. 7, Harrison, M. 130. Kavvadias - Kawerau pl. r , n° 12.
and M. p. 330 and fig. 35, Miller, AJA 1893, 131. Kavvadias, Ephemeris 1897, p. 29,
pp. 485-486, fig. 1, Harrison, P.A. p. 35, pl. 1, 3.
Judeich, Top. p. 116, n. 2, where it is stated 132. Ibid. pp. 28-29, and Broneer, Hes-
with much reservation, Keramopoullos, peria VIII, 1939, pp. 317-433.
PraktAkAth 1932, pp. 114-115, 122, Kera- 133. Kavvadias - Kawerau pp. 33, 87, 89,
mopoullos, Ephemeris 1934/35, pp. 95-96, 91, pls r and Li, Wachsmuth, Ber. sachs.
98, Dorpfeld, Alt Ath. I p. 3, pl. II, Travlos, Ges. Wissensch. 1887, pp. 401-402, 403, Mil-
Ephemeris 1939/41 , pp. 59-62, Travlos, II.o- ler, AJA 1893, pp. 476-477, Judeich, Top.
A.eoc5. p. 25. pp. 259-260, Frazer, Paus. II pp. 355-356,
125. Stevens, Hesperia XV, 1946, pp. 73- AM (Funde) 1887, p. 141, Curtius, Stadt-
75. gesch. p. 45, Tsountas, :4.x{>0.7WAt~ p. 10,
126. Kavvadias - Kawerau pp. 41and59. Koster, Pelargikon pp. 13-14 and pl. IVa,
127. Kavvadias, Ephemeris 1897, pl. 1. Kavvadias, II.eoimoeix~ :4.exawA.oyia p.
128. Pelargikon p. 18. 300, Schede, Die Burg von Athen p. 10.
57
icans added traces of walls inside the Erechtheion and in the area of the Pan-
droseion.134 All these walls were examined and classified by Holland,135 and,
in addition, Broneer discovered the continuation of the NE ascent outside the
wall. 136
The next manifest section of the wall is preserved at the SE corner of the
rock, beside the Museum. Here were found house walls, graves and a cache
containing bronze weapons and objects together with a few sherds. 137 Con-
tinuing along the length of the S slope, another section of the wall emerges
from beneath the SW corner of the crepidoma of the Parthenon, part of it
visible today. 138 After this, the last piece of the wall that has survived is the
section preserved on the SW end of the rock,139 S of the Propylaia.
134. Erechtheum pp. 13-14, 138-142. 23, Koster, Pelargikon pp. 6-10, pl. IVb,
135. Holland, AJA 1924, pp. 1-23, 142- Kawadias, lleoiar:oetxi/ .'.AexawA.oyia p.
169, 402-434, Picard, L 'Acropole II pp. 19- 300, fig. 360, Frazer, Paus. II p. 355, Ju-
20, Judeich, Top. p. 260, Berard, Stud. Rob. deich, Top. p. 115, Kolbe, AA 1939, p. 235,
p. 138, Hill, A thens pp. 13-14, Broneer, FuF 1939, p. 394, Bericht VI Intern. Kongr.
Antiquity 1956, pp. 9-10. Archiiologie 1940, pp. 344-345, Research and
136. Hesperia II, 1933, pp. 351-355; IV, Progress 1940, pp. 254-258, Berard, Stud.
1935, pp. 109-113. Rob. p. 139.
137. Kawadias - Kawerau pp. 33, 35, 37, 139. Beule, L 'Acropole p. 83, Bohn,
39, 95, 99, 101, 103, pl. E ., Montelius, Prop. p. 16, Dorpfeld, AM 1885, pp. 131-139,
VHAM 1889, pp. 49-60, and La Grece p. Botticher, Akropolis pp. 59-61, Lolling, Topo-
155, AM 1888 (Funde), pp. 107-108, (Mis- graphie p. 337, Harrison, M. and M. p. 356
cellen), p. 228, BCH 1888, pp. 244-245, and fig. 35, Curtius, Stadtgesch. p. 45, fig.
Harrison, M. and M. p. 536, and PA. p. 13, 13, Miller, AJA 1893, fig. 1, Weller, AJA
Koster, Pelargikon pp. 6-10, pl. IVb, Kav- 1904, p. 60, pl. I, D'Oodge, Acropolis pp.
vadias, lleoiar:oe txi/ .'.AexawA.oyia p. 300, 23, 364, White, Ephemeris 1894, pp. 25-62,
fig. 360, Heberdey, Olh 1910, fig. I, Frazer, Koster, Pelargikon pp. 6-10, pl. IVb, Kav-
Paus. II p. 355, Graef - Langlotz (Wolters) vadias, lleoiar:oetxi/ .'.AexawA.oyia p. 300,
I pp. XXV, XXXIII-XXXIV, figs 5 and 6, fig. 360, Heberdey, 6Jh 1910, pp. 1-4, fig. 1,
Judeich, Top. p. 115, Kolbe, AA 1939, p. Middleton, JHS suppl. 3, pl. 7, X, Frazer,
235, Berard, Stud. Rob. p. 139. Paus. II p. 355, Walter, Athen, Akropolis p.
138. Kawadias - Kawerau pp. 39, 113, 15, Picard, L'Acropole I p. 11, Judeich, Top.
117, 119, pl. Z, Harrison, M. and M. p. 536, p. 115, Wrede, Attische Mauem p. 5, Welter,
and PA . p. 13, D'Oodge, A cropolis pp. 21- AA 1939, fig. 4, Dirlmeier, Die Pelasger-
58
Other remains of buildings that have been ascribed to Mycenaean times

also lie within the area enclosed by the walls and at the top of the Acropolis.
Visible just behind the base of the Athena Promachos, is a long shallow trench
cut in the surface of the rock and oriented S to N. Stevens considered this to
be a bed for the foundations of a Mycenaean terrace. 140 In the area between
the Propylaia and the Erechtheion, and also S of the Erechtheion amongst
the foundations of the big Archaic temple 141 are a great many walls of all sizes
that have been considered prehistoric. They are shown on Kawerau's plan 142
as "remains of Pelasgian walls", an identification that was generally accepted.
Following the excavation of the Acropolis, the walls were attributed in the
bibliography either to the palace, 143 to houses of the ruler's retainers, 144 or
just to ordinary houses. 145 Holland alone observed that some of these walls
appeared on the plan to have been built on Classical foundations. He there-
fore concluded that these at least could hardly be Mycenaean and he decided
that they were considerably later, perhaps even Mediaeval. 146
Holland's sound observation appears not to have drawn the attention it
deserved. The walls were never investigated in detail, and they continued to
be referred to as prehistoric by scholars whose publications came after Hol-
land's. It was therefore absolutely necessary to study all the walls meticulously
and on the spot, so as to have sound results based on the material itself. This
was not always feasible since some had been covered over by thick fill after the
mauer p. 37, Stevens, Hesperia XV, 1946, pp. pie" to the above mentioned foundations.
75, 78, Hill, Athens p. 8, Marinatos - Hirmer, 142. Kavvadias - Kawerau pl. f.
KefJrrJ xai MvxrJvaixi] <E.A..A.ac; p. 57. 143. Middleton, JHS Suppl. pl. 1, n° 37
140. Stevens, Hesperia V, 1936, pp. 499- and 74, Schede, Die Burg van Athen p. 10.
503. 144. Harrison, M. and M . p. 536, Miller,
141. Known in general as the Heka- AJA 1893, p. 477, Picard, L 'Acropole I p. 11,
tompedon. Since then the question of the II p. 19.
exact location of the so-called Hekatompe- 145. Kavvadias - Kawerau pp. 31, 63, 73,
don has arisen and there is some doubt as Curtius, Stadtgesch. p. 45, Kavvadias, fleo-
to whether this Archaic temple has been wroetxi/ ~exmo.A.oyia p. 301, Frazer, Paus.
correctly identified. The discussion contin- II p. 355, Judeich, Top. p. 282, Stevens,
ues at present. It is therefore preferable to Hesperia V, 1936, p. 502.
apply the term "ancient" or "Archaic tern- 146. AJA 1924, pp. 144, 162 and pl. VII.
59
L~R 16
_, 23
~
~ ~ ' I
!J
r-'
5
r- 111 til12 ~ "-.. ~ "°
"'I
J-. '
\ ' I
-~--~~-
I ' ,
L- - -~
- -8!1:618
4
..,, f 24
2
~
~ \:3 ~
~
~
20
19
20a
21 I
!F)
D
- - -\@-
ERECHTHEION
~ -.
~
~ [ =:= ~ 1
~ ~38
f;il[ -~]
~
Sa
35
A.- l
O='
Q
L:J t:: J
0 5 10 20 30 40
~:S:::et====::l=:=:=:=:===:!==========::l=;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;~M
Plan 3. Walls between the Propylaia and the Erechtheion area.

Acropolis excavation and were inaccessible. Wherever conditions permitted,

however, the walls were uncovered anew, cleaned and studied. In the course
of this work, a number of corrections were made to the plan of Kawerau who,
as Holland noted, had not surveyed these walls with his usual care. 147 Special
attention was given to the construction of the walls and their building mate-
rial, their relation to other buildings in the same area (sometimes informative
about the succession of buildings) and, wherever undisturbed, the fill in which
they lay. The results of this work are as follows (see Plan 3):
Wall complex 1: This forms a unit and the walls comprising it are clearly
contemporary with each other. Built into one of the corners is a rectangular
poros stone block, dressed in a similar fashion to those in Classical buildings.
It is not clear from the plan whether the block was used as building material
for the wall or if it was already there and the wall was added to it. In either
case the wall is later than Classical times.
Wall 2: Built of stones of various sizes and a few pieces of kiln-fired brick,
with traces of lime-mortar in the joins. It belongs definitely to the period of
Turkish domination.
Wall 3: Kiln-fired bricks, pieces of marble and plenty of lime-mortar were
used for its construction. This too is Turkish.
Walls 4 and 5: Available only on the plan. 4 is curving, 5 angular, the dis-
tance between them very small. If they are not contemporary with each other,
one must have cut through the other, but it is uncertain which one is the ear-
lier of the two. If, as their similarity of construction suggests, they are syn-
chronous, they must belong together as there is too little space between them
for them to belong to separate buildings. In fact, taken together with the end
of complex 1 to the east they form a curving unit. This is explainable if we
take into consideration the gun-emplacement that stood here during the Turk-
ish occupation. 148 Thus, as complex 1, the walls must be Turkish.
Walls 6: These walls, which form a single construction, contain pieces of
poros stone together with small fragments of kiln-fired brick, and are there-
fore later.
147. Holland, AJA 1924, p. 144. 148. Travlos, lloA.t:oO. fig. 138.
61
Walls 7: They have been built in part on top of the neighbouring rectan-
gular construction 8. Included in the building material are pieces of poros,
showing that the walls are later.
Rectangular construction 8: It is earlier than complex 7, which lies on top
of its W side. Built into the E fa~ade and the SE corner are pieces of white
marble destroyed by fire, pieces of Karra stone and fragments of Eleusinian
marble. Therefore it must have been built considerably later than Classical
times. It is probably Mediaeval.
Wall 9: The eastern end is built on the foundation of the house of the
Arrephoroi and it is therefore later than this building.
Walls 10, 11 and 12: As wall 9, they were built later than the house of the
Arrephoroi, their northern ends abutting the foundations of that building.
They contain pieces of green marble. Walls lOa and lla, which were built
above the preceding walls, are of course later still.
Walls 13 and 14: As was evident from the cleaning of these walls, they too
have been built against the E part of the foundation of the house of the
Arrephoroi, the construction of which therefore precedes them.
Walls 15 and 16: They form a corner, the end of which is built on the Clas-
sical poros wall to the N and on the SW corner of the neighbouring Medi-
aeval cistern. Lime-mortar has been used abundantly as bonding material.
They are definitely even later than the cistern.
Wall 17: Insignificant little wall of rough and careless construction, form-
ing a curved line. It cannot be Mycenaean.
Walls 18: The northernmost is unquestionably the continuation of the small
section running at right angles to 12. The southernmost contains pieces of
green marble. Both walls must be contemporary with 12.
Wall 19: Built of a variety of materials. At its eastern end it turns slightly
to the N, forming a sort of anta.
Wall 20: Little wall with a piece of poros built into it.
Walls 20a and 21: 21 cannot have been built while 20a was in existence; 20a
is therefore the later of the two. The excavation, however, showed that 21 is
a continuation of 12. The two walls are therefore later.
Wall 22: Pieces of carved poros, marble and kiln-fired brick have been in-
corporated in the east end. It belongs to Mediaeval or Turkish times (Fig. 5).
62
Fig. 5. Wall 22 of Plan 3, from E.
Wall 23: The Send of the wall rests on the foundation of the Erechtheion,
which therefore preceded it.
Wall 24: It contains fragments of marble and poros with traces of Classi-
cal tooling.
Wall 25: It is constructed of various types and sizes of stones, joined in dis-
orderly fashion with lime-mortar.
Flagstone paving 26: This is a somewhat irregular paving of large stones,
set with the smoothest surface up. It is surrounded by walls of the Turkish
63
period. The paving stones, which include a piece of poros with marks of a
coarse hammer, were found lying on a compact fill composed of hard earth
with traces of yellowish clay, undisturbed by the excavation of 1885-1890 and
containing sherds chiefly prehistoric but also a fragment of a Corinthian ary-
ballos. Thus it cannot be earlier than the Archaic period.
Corner wall 27: The two ends toward the S are built, the W one against
the foundation of the north porch of the Erechtheion, the E one against the
Turkish cistern. Thus the wall has to be later than the cistern.
Wall 28: Archaic, carefully constructed in polygonal style with well ham-
mered stones.
Wall 29: It is built in rough and careless fashion of small stones, kiln-fired
brick and lime-mortar; consequently this too must belong to the years of Turk-
ish domination.
Walls 30 and 31: They have been built on the foundations of the cella and
peristyle of the Archaic temple. The W end of one of the walls in one place
penetrates the wall of the foundation. Thus they were built later than the
temple.
Wall 32: It is made of relatively large stones and it is fairly well built. A
late Roman sherd was found beneath the third layer of stones.
Wall 33: A piece of poros is built into the top of the wall. Beneath this,
the stones are very well joined. Sherds of the Sth century B.C. were recovered
from beneath the wall.
Wall 34: Holland considered it to be a foundation for the support of the
cella roof of the Archaic temple. 149 It is more likely, however, that it was built
prior to that, and was cut off during the setting of the temple foundations. Its
position and construction connect it with 35.
Wall 35: Higher than the neighbouring walls, it differs from these in con-
struction as well. It is built of large dressed stones that are flat and horizon-
tal on top. Archaic in all probability, like 34, it must be either a foundation
for a column base of the temple, as suggested by Holland, or somewhat earlier
than the temple.
149. Holland, AJA 1924, p. 162.
64
Walls 35a, 35p, 36, 37: They are later than the Archaic temple and wall 34
since they are built against them.
Thus, of all these walls that were considered to be prehistoric and attrib-
uted to the palace complex or to other simpler houses, none are Mycenaean.
A few, to be sure, 28, 35, and probably 34 and 26, are ancient, but they date
to historical times. Some, 8, 32 and 33, are later still, although it is not poss-
ible to determine exactly when they were built. All the rest are remains of lit-
tle buildings that had been put up within the Acropolis fortress during the
Turkish domination.
There remain the two poros column bases lying within the Archaic tem-
ple, to the S of the Porch of the Maidens (Plan 3, 38). They were found dur-
ing the excavation of the Acropolis at a level lower than the top of the tem-
ple walls that surround them, as is explicitly stated by the excavators. 150 This
is why they were thought to precede the temple chronologically. Since in form
they resemble Mycenaean bases, they were generally accepted as belonging to
the Mycenaean palace, as also that they were still in their original position. 151
Presupposing this, Holland based his reconstruction of the palace megaron on
them. 152
With the passage of time, doubts arose as to whether these bases really
were in situ. It was observed that their tops were not exactly level, and it was
suggested that at least one, the northernmost, had been moved. 153
Let us examine them. They are made of soft, yellowish poros. Each con-
sists of a cube, from the top of which projects a low cylinder cut out of the
same piece and forming the main base of the column. Along the sides of the
cubes are point marks (Fig. 6). The S base is 0,94 m. long, 0,76 m. wide and
0,27 m. in height; the top of the cylindrical part has a diameter of 0,55, and
150. Kavvadias - Kawerau p. 83. determining that the top of the cubical
151. Ibid. p. 83, Middleton, JHS Suppl. lower part of the northernmost base was
pl. 1, n° 67, Jahn - Michaelis, Atx Ath. pp. 0,076 lower than the equivalent point on
VI-VII, Dorpfeld, Jdl 1919, p. 4, and Pauly - the southernmost. This is repeated by Bro-
Wissowa, RE s.v. Athena, p. 1952, Holland, neer, AntUjuity 1956, p. 9, adding that Dorp-
AJA 1924, p. 162, AJA 1939, p. 289. feld too expressed similar doubts during his
152. AJA 1924, pp. 162-168. lectures on the Acropolis.
153. Paton, Erechtheum pp. 427-428,
65
Fig. 6. The W side of the S column base, showing traces of working with a point.
rises 0,08 m. above the cube. The N one is 0,94 m. long, has a maximum pre-
served width of 0,70, height 0,31 and height of cylinder 0,14 m. The original
diameter of the cylinder cannot now be ascertained because it is damaged. It
will, however, have agreed with the other base. The cylindrical projections are
flat on top. The lack of tenons shows that the columns they held were wooden.
Both rest on a substructure consisting of small stones; among the stones of
the N base, however, are fragments of brick.
The S base is in fairly good condition and it is complete, whereas the N
one is not well preserved and has broken into pieces, particularly at the top,
which are held together at present with mortar. In addition, its N side, that
facing the outer wall of the cella of the Archaic temple, to which it is fairly
close, is uneven and incomplete. The missing part clearly was removed with a
hammer. It is very likely that these same hammer blows were responsible for
the cracks which, widening in the course of time, caused pieces to pull off at
the top. The builders of the Archaic temple evidently found that the base was
close enough to the line of the foundations to hinder them in their work,
hence the mutilation.
66
The bases, moreover, are not on the same level. The top of the cylinder of
the N base is 0,02 higher than that of the S, and the tops of their cubical mem-
bers show a still greater discrepancy (see supra, n. 153). These differences are
insignificant and in themselves would not be enough to show whether the
bases are in their original position or not. Yet the fact that the N base was
cut back in order to lay the foundation for the cella of the temple, and in par-
ticular the existence of a kiln-fired brick in its foundation shows that this one
at least had been moved, and more than once.
Thus at least one of the bases is not in situ. Are they really Mycenaean?
That they were found beneath the level of the temple of the sixth century
B.C. means that they belong to an earlier, but not necessarily Mycenaean,
building. The material of which they were made is the soft yellowish poros
that was used especially for the buildings and sculpture of Archaic times.
Poros in general as a rock was not unknown in the Mycenaean period. The
term "poros" is generic and includes numerous varieties. 154 This particular
stone, which comes from one of a few quarries, is characteristic of the Archaic
period. Moreover, the careful working of the cylindrical tops of the bases and
their plinths, which have been cut precisely to a regular four sided shape, are
features unknown in Mycenaean architecture, in which the column bases,
except for the flat top, would have been covered by stucco or by the flooring
in general. These two bases indeed belong to a period that had other con-
ceptions about the form of a base. A comparison with the limestone base,
indisputably Mycenaean, that was found E of the Erechtheion 155 makes this
particularly clear. The base is totally different in material, measurements and
method of working. Finally, and even more significant, the use of the point
for working the sides of the cubes shows that the bases were made in post-
Mycenaean times.
If this form had disappeared after the end of the Mycenaean period, we
would have to accept them as Mycenaean. Yet this is not the case. Wooden
columns had bases of similar form down to the time when the columns of
buildings were set on a continuous and unified stylobate. This was dictated by
154. See also Orlandos, 'YA.txa ooµijr; II 155. Kavvadias - Kawerau p. 68. See
p. 68. infra, p. 192, figs 30, 31.
67
technical necessity. The lower part of the column had to rest on a stable and
flat surface. It had also to stand at a height greater than the rest of the floor
so as to avoid exposure to stagnant water and ground dampness. This method
of supporting columns is therefore not unusual during Geometric and even
Archaic times. 156 The form and working of bases is sometimes even more
primitive than those of the Acropolis.
Thus there is no evidence at all that the bases are Mycenaean. Instead,
there are serious reasons for believing them to be considerably later. It is not
impossible that they belong to the same building as walls 34 and 35. If so, they
should probably be attributed to a temple built on that site after the Myce-
naean megaron and before the Archaic temple, to a temple the existence of
which Dinsmoor considered to be "hypothetical but necessary." 157
156. Bases of this sort were found in the ASAtene 1914, pp. 33-34, figs 6-7, 9-12), and
Archaic temple of Thermon (Soteriades, similar bases may be seen on the Fran~ois
Ephemeris 1900, p. 173), where they con- vase. In the Geometric temple of the end
tinued in use for a long time, in the 7th cen- of the 8th century B.C. at Dreros, instead
tury B.C., the Archaic building at Dreros of a base there is a rudimentarily worked
(Xanthoudides, Deltion 1918, appendix p. stone without any cylindrical rise (Mari-
26), the Archaic temple of Prinias of the natos, BCH 1936, p. 227, fig. 12).
beginning of the 7th century B.C. (Pemier, 157. Dinsmoor, AJA 1947, pp. 109-110.
68
II. THE BUILDING PHASES
Preserved to the E and NE of the Erechtheion are complexes of walls.

There is also a section of the Mycenaean fortification wall. The relation of
these walls to each other and all together to the Mycenaean wall shows that
they are not contemporary, but belong to different building phases. 158 The
phases can be distinguished and defined through a study of the walls.
The stairway rising from E to W (A on Plan 4) is blocked at the end by a
wall, n° 5. The preserved height of that wall, still visible today, is such as to
prevent the stairway from continuing westward. Thus it was built later, specifi-
cally to put this approach out of use (Fig. 7). Yet this wall is manifestly later
than walls 1 and 2, since its stones are built against these walls without bond-
ing. Furthermore the E face of wall 5 is stepped back some 0,50 m. more than
the line of the E faces of 1 and 2. Wall 5 is therefore a later addition and, for
the same reasons, the same is true of the two parallel walls 6 and 7.
If these later additions are excluded, we see that walls 1 and 2, built at right
angles to the S and N of the ascent, leave an open space between them,
approximately 4,40 m. wide, that allows the ascending pathway to continue to
the W. Thus two building phases are evident: the first, during which the
ascending pathway leads to the top of the rock and continues westward
between walls 1 and 2; and the second, during which the end of the ascent is
blocked by wall 5 and by walls 6 and 7 which are similar to it. To the E of
wall 1 (Plan 4) is preserved the corner wall 3. The space between walls 3 and
1 has also been closed, by wall 4, likewise later as is clear from its construc-
tion. It is reasonable to conclude, therefore, that just as walls 1 and 2 are ear-
lier than 5, walls 1 and 3 are earlier than 4.
Wall 3, however, also precedes the N Cyclopean wall (Plan 4, 8), which at
this point has been built against the N face of 3.
158. See Holland, AJA 1924, pp. 142-157.
69
THE BUILDING PHASES
Fig. 7. Wall 5 of Plan 4, from E. To the left is 1, to the right, 2 (phot. DAI n° 48).
We can now define two successive building phases: during the first, walls
1, 2 and 3 exist, and the ascent along the pathway functions; during the sec-
ond, the ascent is blocked by the addition of walls 4, S, 6 and 7, and the for-
tification wall is built.
This does not necessarily mean that the constructions belonging to this
period are synchronous with each other. That remains to be seen. It does,
however, mean that beyond any shadow of a doubt we have at least two suc-
cessive phases.
Just N of the E cella of the Erechtheion, below the Classical poros slab
pavement, a wall was built in the form of a 11 opening to the N. It is desig-
nated wall 9 on Plan 4. 159 Another, narrower wall (10), runs along the length
159. Kawadias - Kawerau pl. r , and Holland, AJA 1924, pp. 151-156, fig. 12.
70
THE BUILDING PHASES
Plan 4. Walls in the area N and NE of the Erechtheion.
of its S leg and on top of it. It is of a different sort of construction and it fol-
lows a line at a slight angle to the wall beneath it. Wall 10 is clearly later than
9: it has been built on top of it and as it does not follow the N legs of 9 it
cannot be interpreted as the upward continuation of the same wall, with 9 as
the foundation. Instead it continues eastward to a point where it meets yet
another wall (11), running from N to S. Where the two walls meet, 10 is con-
structed of small stones that abut the regular W face of 11. It is built against
11, and is therefore later. As we saw, however, it is later than 9 as well. Walls
9 and 11, however, cannot be synchronous, since 10 lies entirely on 9 without
following its turn to the N, whereas it simply abuts on 11. This means that
when 10 was built, 9 was no longer used as part of a building, while 11 existed
and was still in use. Thus 9 is not only earlier than 10, but it precedes 11 as
71
THE BUILDING PHASES
well. Yet 11, which was found in part covered over by the Classical slab paving
of the N courtyard of the Erechtheion, is unquestionably ancient. In con-
struction, size of stones and orientation, it resembles the walls around the end
of the ascent. Its position, thickness and construction, however, exclude the
possibility that it is part of the fortification wall. It must therefore belong to
the same phase as walls 1, 2 and 3 on Plan 4. Since wall 11 is later than 9 but
earlier than the fortification wall, we have at least three building phases. The
first phase is represented by wall 9. To the second phase belong walls 1, 2, 3
and 11. The third and final phase includes walls 4, 5, 6, 7 and wall 8. On the
basis of these conclusions we shall now examine the remains of the Myce-
naean Acropolis.
72
1. THE FIRST PHASE
The wall lying beneath the Classical slab paving of the courtyard N of the
Erechtheion (9 on Plan 4; Plan 5) is the only construction belonging to this
phase. Kavvadias was the first to excavate it and, although it is not mentioned
in the text, it is recorded on the relevant plan of the Acropolis excavation. 160
With the American excavations in the area of the Erechtheion, it was re-exca-
vated in 1923 by Holland, who studied and published it in detail. 161
~ .. I 1I +
~
1 1 1
0
0 1 5 10
Plan 5. Remains of the LH I house N of the Erechtheion.
It represents the remains of a four-sided area, the N part of which is not

preserved. It is oriented E to W. At each end the walls make an approximately
right angled turn to the N. Built of rough limestone blocks, with a maximum
160. Kavvadias - Kawerau pl. r, under 161. AJA 1924, pp. 151-156, fig. 12.
the number 36.
73
110_/
105---------
..
.---
,~ ;.
/
!:.... '-
~,, , ..,,_ ~
,
.,,,
'····
<
i
,.,,,,
:itTl
,
-....J
~
·-.,
'''••,,,,,,, ....
'
...~~
-
"Tl
~
r.n
>-3
.... •···· ~
......··' ""ti
~
r.n
tTl
/
115
~ 1 10
~ 110~
~ / ,,.-- 105
0 10 50 100 150 200 250 300 350

M
Plan 6. The Acropolis at the begi,nning of the LH period.

THE FIRST PHASE
measurement of 0,30-0,50, and standing to a height of 0,40-0,80 from the rock,

they are bedded partly on the rock itself, and partly on a thin layer of fill. The
area is at least 6,60 m. long, and ea. 0,75 m. wide; of the legs running north,
the easternmost is preserved to a length of 2,50 and that on the west to about
3 m. It will thus have included a fairly large space.
There is no trace whatsoever of any flooring on the outer side of its S face
and this side was evidently the exterior. The pathway to which the NE ascent
led probably ran in front of this. The floor of the interior was 0,025 thick,
carefully constructed of tamped white clay, some 0,10-0,15 above the rock, and
it covered all the surface enclosed by the three walls. Next to the SW corner,
n° 1 on Plan 5, a small four-sided hollow was found, a sort of small bothros,
the purpose of which is not clear. Beneath the floor and covered by it lay a
child's skull.
The type of flooring indicates that the construction (Plan 4, 9; Plan 5) is
the remainder of a roofed area, which, as seen in Plan 6, had been built at
the top of the rock very near the brow.
The date of the building is provided by sherds collected from the fill within
the room, above and below the floor. Since the sherds from both groups
showed little difference between each other, it appears that the room was not
long in use and was soon abandoned.
Most of the sherds are Middle Helladic of various categories, including also
a considerable number of reddish monochrome, LH I in date.162 If the skull
found beneath the floor represents a burial, at a stretch it may be of this same
date. Yet its presence here is more likely to be a coincidence given the short
distance between the room and the Middle Helladic children's graves to the
NW (Plan 2, 7).
Be that as it may, the building represents the earliest Mycenaean evidence
of habitation on the Acropolis, and it belongs to the beginning of the period.
162. As Holland reports, the sherds dated them (AJA 1924, p. 151, n. 1, 155-
were examined by Wace and Blegen, who 156).
75
2. THE SECOND PHASE
THE AREA W OF THE TERMINUS OF THE NE ASCENT
Preserved in the area N and NE of the Erechtheion (see also Plan 4) is

the inner face (see Plan 8, T) of the Cyclopean wall, as well as a complex of
walls (la, tp, 2a, 2p, 2y, 3, 4, 5, 6, Tl, T2, T3, T4, T5 and T6 on Plan 7).
The face of the Cyclopean wall along this stretch has undergone many
changes through time. In the Classical era, after the Persians destroyed it, it
was levelled to 148,82 m. and used as a foundation for the column drums,
which were set there as part of the Themistoclean fortification wall. Later on
it was deformed by walls belonging to the period of Turkish domination, which
rested on top of it. Despite all this it is well enough preserved to be recog-
nisable. Parts of all these walls are visible and accessible today. Discernible
specifically are the point where T2 and T3 meet, the N end of T4 and practi-
cally all of T5. Likewise visible are the column drums a, p, y, b, £, ~ and, in
part, 11· Orum Q, which Kawerau notes as upright and in position, now lies
fallen on its side over the N end of wall T3.
Cleaning of the visible stretches of these walls revealed the following: T2
and T3 are built of stones of various sorts, for the most part Acropolis lime-
stone but some from elsewhere. Since they are constructed loosely and care-
lessly, there are many open spaces between the stones that have been filled
in with smaller stones, likewise carelessly placed as chance dictated. Among
these are pieces of fired brick and quite a few pieces of marble. All of it is
plastered together with lime mortar containing a high percentage of lime. The
two walls are not of the same height, T3 being preserved to a higher level than
T2. Furthermore, about half of the area of the top of the poros column drum
Q is covered by remains of lime plaster and small stones built up to two lay-
ers. If drum Q, which, as we noted, is now lying on a slant on T3, is placed
upright in its original position, these constructions form a continuation of T3
to the N. The wall in any case continues one way or another on the preserved
top of the Cyclopean wall almost to drum ~. Thus, as T3 lies on top of both
76
THE SECOND PHASE
2y
0 5 10
~H~f"'-l~~H-i~~I~~~~~I M
Plan 7. Mycenaean and Mediaeval walls NE of the Erechtheion.
the Cyclopean wall and the Classical wall and since it is constructed with lime
mortar, both T3 and T2 must be notably later (Fig. 8).
As for T4 and TS, the join between them shown by Kawerau on his plan is
in fact non-existent. It is clear, however, that they are connected and form a
corner. In any case, both the N end of T4 and the N side of TS are built to a
width of about 0,10-0,15 m. on top of the Cyclopean wall and they are pre-
served to a notably higher level. Type, variety and arrangement of material
are the same as in T2 and T3. TS indeed contains pieces of marble, not only
at the top but also in a few courses lower down. Lime mortar has been used
in these too. Like the previous walls, they are therefore later (Fig. 9).
What holds for T2 and T3 holds as well for Tl, which is the continuation
of T2 and together with it forms a regular corner. In addition to the argu-
77
THE SECOND PHASE
Fig. 8. The N end of wall T3, which continues on the preserved top of the Myce-
naean fortification wall (right, below the column drums). Above, right, column
drum e with traces of the wall construction on top of it.
ments based on construction, which are already ample, there is one further
piece of evidence related to the Cyclopean wall: the north side of Tl, specifi-
cally its northwest corner, is so close to the brow of the rock, without even
being parallel to it, that there is less than 2,50 m. to the edge. There would
indeed have been no space for the wall had Tl existed when the wall was built.
T6, on the other hand, not only abuts on T4, with which it forms a corner,
but it is higher than 2y, the S end of which it has covered. It has, as seen
clearly on Kawerau's plan, the same kind of projecting sub-foundation of
smaller stones as have the others. It is the same as the others, associated in
any case with T4.
It follows that the walls Tl, T2, T3, T4, TS and T6 are all later and they are
roughly synchronous with each other. The lime mortar and bricks used in their
construction place them in Christian or Turkish times. Kawerau informs us
that in this place there was a little house and a Turkish "domed construction"
78
THE SECOND PHASE
Fig. 9. Walls T4 and TS of Plan 7, from the S.
which had been used by Pittakis to store his finds. 163 This would suggest that
the walls, which seem to have incorporated whatever building material was at
hand, were associated with the foundations of these buildings.
The remaining walls of the area, shown in Plan 8, may now be examined.
Of these, wall 5 blocked the ascending path, which was in use during Myce-
naean times. As we shall see below, the blocking of the path had already
occurred by then. Wall 5 is therefore Mycenaean, as are also walls la and 2a,
which precede it. Wall la, moreover, continues to the Was lP and, as observed
by Holland,164 these are terrace walls. This is evident from their thickness,
up to 1,50 m., which is too narrow for a fortification wall but very wide for a
simple building wall. 165 Still more significant, they have only one face, toward
163. Kawadias - Kawerau pp. 15, 33 and 165. The walls of the megaron at Myce-
fig. 2, where Turkish houses are shown in nae vary in thickness from place to place,
this place. but do not as a rule exceed 1 m. At Tiryns
164. AJA 1924, p. 145. the thickest are 1,30 m.
79
THE SECOND PHASE
~~\. ~
,I
--
~="015- - - - - ~- - - - - -~mP- ~
- - - - - - - - - - - 'ii)
---@---~'%
7[3
--- 7y
\ 151 ,15
o~~- --
I
I
0
8
0 5 10 20 40
~~~:l========~===========l:~~~~~~~~~~~~M
Plan 8. Remains of terrace walls NE of the Erechtheion.
the N; on the inner side the stones are not set in a line, but are adapted to
the uneven lie of the rock. This is therefore a terrace wall facing N along line
la-lp, which obviously supported a level space to the south of it. So, its height,
that is the distance from the rock to the top of the terrace wall, will have
varied according to the level of the foundations so as to keep the top of the
terrace level.
Some 5,50 m. S of wall la-lp, there are traces of another terrace wall, par-
allel to the first (7a-7P-7'Y on Plan 8). The E end of this terrace wall abuts on
the S extension of wall la, without being bonded into it. It is thus clear that
80
THE SECOND PHASE
wall 7a was built after wall la; not much later, however, since the two suc-
cessive terraces appear to have been planned and constructed as part of one
and the same programme. Wall 7a, the eastern end of the second terrace wall,
is founded on the rock at an approximate elevation of 150,40 m., and the same
holds for 7y. 166 Since the two terraces, placed one behind the other, can only
have been stepped, the level of foundation of the second will give us the level
of the top of the first. Thus the top of the terrace bounded by wall la-IP was
at an elevation of around 150,40-150,50 m. Because of the unevenness of the
rock, which rises markedly toward the W just where la and IP were built, la
is founded at a height of 148,40, while IP is at 149,56 m. Since, as we saw, the
top of the terrace is at ea. 150,50 m., wall la will have been 2-2,10 m. high,
and IP around 1 m.
The top of the terrace wall 7a-7y can only have agreed with the level of
the space S of it, where later on the Archaic temple was built and the S side
of the Erechtheion was founded. That the Archaic temple was built on a ter-
race that was as high as the euthynteria, is well known. It is evident from the
fact that the foundation stones of its colonnade are coarsely worked on the
outer face and were clearly not meant to be seen. Because of the steep decliv-
ity of the rock from S to N, only the SE corner of the temple was founded
directly on the rock. The rest was constructed on an artificial terrace that hid
the foundations. Such a building, however, with compact and massive con-
struction, could not possibly disappear without leaving the slightest trace in a
place that had been filled in and covered over from Classical times on. Since
no such traces were preserved, the temple was evidently built on the already
existing Mycenaean terrace, which was covered over with fill after the Themis-
toclean wall was built. The construction of a new terrace would in any case
have been superfluous. The euthynteria of the Archaic temple is at the level
of 152,54 m. This we may take as the elevation of the top of the terrace wall
7a-7y, which, since it was founded at 150,40 m., will have been slightly over 2
m. high.
N of the end of the NE ascent, opposite wall la, is a corresponding cor-
ner wall, 2a on Plan 8. At right angles to this wall are the two walls that are
166. Kavvadias - Kawerau pl. A.
81
THE SECOND PHASE
Fig. 10. Walls 2P and 3 of Plan 8, from the N, at their point of junction. In front
the inner side of 2p. Wall 3 is next to the wooden stairway (phot. DAI n° 46).
parallel to wall 5, numbered 3 and 4. Since they too block the pathway, these
walls must be contemporary with 5 and later than 2a. Wall 2p likewise pre-
cedes them. In fact, both the plan and the German Institute photograph n°
46 (Fig. 10), which shows the point where 3 touches 2P viewed from the N,
indicate that the northernmost stone of wall 3 is not bonded into wall 2p but
lies parallel to the S face of the wall and on a different level. Wall 3 was clearly
built at a later time against 2p. Walls 2P and 2a are accordingly contemporary
with each other. They are built along the same line and 2p is actually a con-
tinuation of 2a to the W. Thus, there is also a wall along the north side of the
ascent, approximately parallel to la-Ip, and the pathway ran between them.
The width, construction and length of this wall (greater than that preserved
since the wall did not end at the point to which it was preserved at the W)
exclude its identification as a house wall. Moreover with a width of only 1,40
m. it cannot have been part of the fortification wall. Thus there will have been
82
THE SECOND PHASE
yet another terrace wall, represented by walls 2a-2p, similar to and facing S
toward la-lp.
Some 10 m. from the preserved end of 2p, to the W, there is a stretch of
wall running S to N, wall 6. This has been discussed above (see Plan 4, wall
11, and Plan 7, wall 6). The wall is covered in part by the paving of the court-
yard N of the Erechtheion and is therefore earlier than the courtyard. Since
its construction excludes the possibility of its being either Geometric or
Archaic, it may be considered Mycenaean. The W face is made up of stones
that are larger and more regular than those of the E side of the wall, so the
wall faced W. Up to 1,20 m. in width, it is comparable to the previous walls.
To the W, moreover, was the LH I room (see Plan 5), on the foundations of
which rested the narrower, later wall that was built against wall 6. This means
that the space west of 6 was open. Wall 6, therefore, forms the W end of the
terrace wall represented by 2a-2p. If we extend the line of 2a-2P to the W, and
the line of 6 to the S, these two extensions meet at a point about 0,60 m. south
of 6 and they give us the line of the S face of the terrace wall. This, to be
sure, is an approximation since Mycenaean walls, especially retaining walls,
were never perfectly straight.
The north boundary of the terrace is more difficult to determine. Yet there
are a number of clues, one being the inner face of the N Cyclopean wall at T
on Plan 8. The fortification wall, built after the terrace, along the brow of the
rock, which at this point is some 4 m. further north, will either have been built
against the outer face of the terrace wall or it will have stood on top of it. In
other words, the terrace wall extended to the brow of the rock and followed
it, or it lay within the fortification wall S of line T, or it will have run some-
where between these two lines. The most likely solution appears to be the lat-
ter, for two reasons. The first is the construction of the inner face of the for-
tification wall, along its foundation. It is built of stones that are smaller than
usual and do not form a regular face, recalling the construction of the inner
side of the terrace walls of the area. This means that the inner side of the for-
tification wall probably rested on the already existing retaining wall, which was
incorporated in this way into the width of the wall, serving as a foundation on
the inner side. Construction T, therefore, can be considered as part of the
retaining wall before it was incorporated into the fortification wall. The sec-
ond reason is the arrangement of the stones at the NE end of wall 6. The
westernmost of the two last preserved stones at this point is in normal posi-
83
THE SECOND PHASE
tion on the inner building line of the wall, but the easternmost lies at an angle
to the wall and appears to make a wide angled turn toward the E. If the line
indicated by this stone is followed it joins easily the line of T. In my opinion,
this is the line of the inner face of the north side of the retaining wall. The
outer face can be restored parallel to this, assuming an average thickness for
the wall of 1,20-1,40 m., as is evident from the walls preserved.
Walls 2a and 2'Y represent the E boundary of the terrace. Wall 2'Y is stepped
back about 1,50 m. to the W of the line of 2a. This was dictated by the con-
figuration of the rock, which at that point is divided by a deep cleft into two
tangent masses. To avoid the cleft, wall 2a had to be jogged slightly westward
before continuing toward the N. Accepting this, if wall 2a is extended to the
N and W, and 2'Y to the S, they meet to complete the line of the retaining
wall. It bordered and retained a terrace N of the end of the NE ascent, cor-
responding to that on the opposite side. The height of the terrace cannot be
estimated with any certainty, but it is likely to have agreed with that of retain-
ing walls la-lp.
We now return to the terraces S of the ascent in order to determine their
continuation to the W. It is clear that walls lP and 7'Y were cut off in the
process of laying the foundation for the E pronaos of the Erechtheion, 167 and
that they therefore originally continued further than the point where they
appear to end (Fig. 11).
A study of Kawerau's plan 168 reveals that the three last stones of wall lP
to the W do not follow exactly the same direction as the others, the line of
which connects them with la. Instead they turn slightly northward forming a
very wide, almost imperceptible angle. This is not exceptional since, as noted
above, Mycenaean retaining walls are never precisely straight. If this adjusted
line is continued toward the interior of the Erechtheion, it coincides approx-
imately with the line of the interior of the foundation for its N wall. The foun-
dation projects out some 0,25-0,30 m. south of the line of its overlying wall.
167. This is visible on Kavvadias - Kawe- they actually are. Proof of this lies in Fig.
rau pl. r, where the underpinnings of the 11 (phot. DAI n° 742). See also Holland,
foundation of the E pronaos of the Erech- AJA 1924, p. 419, fig. 5.
theion are shown as much narrower than 168. Kavvadias - Kawerau pl. f.
84
THE SECOND PHASE
Fig. 11. Wall IP of Plan 8 at left, and the foundation of the E porch of the
Erechtheion, from the N (phot. DAI n° 742).
Along its length the faces of the poros foundation blocks preserve · traces of
tooling showing that they had been placed against an already existing rough
wall and had been worked so as to conform with it. At two points (Plan 8, ly
and lb) two sections of this wall are preserved in situ with a width of some
1,50 m. 169 They are built of the same material and in the same way as the pre-
vious retaining walls. There can be no doubt that wall ly-lb is the westward
continuation of wall la-lp.
169. Erechtheum pp. 138-142, fig. 88, pl. 156, fig. 12). Thus he accepted the exist-
II, and Holland, AJA 1924, pp. 1-23, fig. 1. ence of two similar, parallel, heavy retaining
Holland correctly noted this wall but be- walls, A4 and F2, 0,60 m. apart, a situation
lieved it to be a different wall from lP (which that could never have existed.
he numbers A4 on his plan in AJA 1924, p.
85
THE SECOND PHASE
Preserved within the cella of the Erechtheion, in the corner formed

between the W cross-wall and the Christian foundation of the S aisle, is yet
another trace of a similar wall: 170 a long and fairly large stone (Plan 8, 7b).
The stone lies precisely on the line of extension of wall 7a-7y, if we continue
that wall toward the W as with la-lb. Thus the two retaining walls continue
towards the W and they can be followed as far as the W cross-wall of the
Erechtheion.
Clearly they did not stop here. There will have been a boundary further
W. Indeed the existence of a west retaining wall, hiding the foundation of the
Archaic temple, had been postulated by the principal researchers. Yet no one
had gone beyond that to determine its exact position. 171 Stevens 172 was the
first to observe that just behind the base of the Athena Promachos, there were
traces of a shallow trench that had been cut into the rock, running from S to
N. He considered this trench to belong to the foundation of a Mycenaean
retaining wall that supported the terrace on which the Archaic temple had
been built and that it was still extant in Classical times. He determined its S
boundary on the basis of the traces of the Processional Way of the Pana-
thenaia which made a detour around the SW corner of the terrace and then
continued its course toward the E. The N boundary of the terrace he surmised
both from the existence of the supposed Mycenaean house walls, beyond
which the retaining wall could not have gone, and particularly from the
oblique positioning of two Classical bases lying NE of the end of the wall. The
position of the bases makes sense only if there was a wall to the S of them.
The orientation of the bases showed Stevens in addition the line followed by
the N leg of the retaining wall toward the E. The direction of the S leg he
surmised from the existence of a Classical wall that was parallel to the foun-
dations of the Archaic temple, but not parallel to the Parthenon. Given that
this south leg begins at the S end of the trench and runs eastward between
170. Holland, AJA 1924, p. 2, fig. 1, g agrees that its height is 153-153,23 m. at the
and Erechtheum pl. II. top. Likewise Paton, Erechtheum p. 437, n.
171. Dorpfeld, Jdl 1919, pp. 12, 38 and 3, and Holland, AJA 1924, p. 145.
pls 1, 2 and 3, where he shows it with the 172. Hesperia V, 1936, pp. 499-503, figs
reference Y much further east than it actu- 42-43, 51-52. See also Holland, AJA 1924,
ally is. He does not locate its S leg, but he p. 77.
86
THE SECOND PHASE
two constructions that are parallel to each other, that is, the Classical wall and
the Archaic temple, it can only have been parallel to them. The Classical wall
formed the N boundary of the processional way, which ran E along the side
of the Parthenon. It could not have been planned and built at an angle to the
Parthenon unless it had to be adjusted to another already existing wall, in this
case the Mycenaean retaining wall.
He calculated the height of thi~ Mycenaean terrace on the basis of the elev-
ation of the rock at the NE corner of the Promachos base and the height of
the stylobate of the Archaic temple, which was 4,53 m. at the middle of its W
end. He dated it to the Mycenaean period on the basis of the chronology of
the buildings that had been built on top of it, and on the fact that the surface
of the rock within the terrace showed no trace of having ever been worked.
This showed that the area was covered by an earth accumulation from very
early times, thus ruling out any building activity in that space.
Stevens' discovery and conclusions were generally accepted as the correct
solution to a long-standing problem 173 and, as we shall see, exploration of this
area has fully supported his ideas. Cleaning in the area has shown that in gen-
eral the trench is just as Steven described it, long and narrow, and running
from S to N (Fig. 12). Its width varies from place to place, but in general,
where the rim has survived on both sides, it ranges between 1,20-1,40 m. At
only one place, exactly E of the base of the Promachos, it is 2,50 m. wide.
This, however, is due to the configuration of the rock. As Stevens rightly
observed, the S end is defined by a low rise in the surface of the rock, worn
smooth by the feet of those walking along the ceremonial road, and by the
cuttings in the rock which show the course of the road (Plan 9, 1). This end
is at a level of 149,58 m.
173. Reservations were expressed by ther a trench nor Mycenaean (Es gibt keine
Dinsmoor alone (AJA 1947, p. 122, n. 69 mykenische Mauerbettung dieser Art, Beyer)
and fig. 3) who changes somewhat the orien- but a natural hollow in the surface of the
tation of the N leg of the retaining wall and rock (Beyer, Wright) or a much later set-
has doubts about dating it to Mycenaean ting for a row of monuments. This unequivo-
times. Later, Immo Beyer (AA 1977, p. 50) cal statement is obviously based on their
followed by J. C. Wright (AM 95, 1980, pp. lack ·of familiarity with Mycenaean con-
64-65, n. 18) stated that the trench is nei- struction tools and practices.
87
THE SECOND PHASE
Fig. 12. Foundation trench for the te"ace wall behind the Pro machos base, from
the south.
The W rim of the trench is the straightest, well cut and deep, and it is eas-
ily discernible for practically its full length. At only a few places is the rock
surface so low that the cutting cannot be made out. Just S of the Promachos
base, at about the middle of its course, the bed of the trench is 0,35 m. lower
than the rim. This section is also the best preserved. Clearly the trench was
cut along a line where the rock had already a natural hollow, and this hollow
remained approximately as it was along the E side.
The E rim, which corresponds to the inner side and was invisible, is there-
fore neither as straight nor as carefully cut as the W rim. It is visible only
toward the S, even though the trench is very shallow there. Further N it dis-
appears for some 8 metres, to reappear again along an irregular line, deter-
mined more by the chance formation of the rock, and modified in places by
minor chipping rather than by systematic cutting.
88
THE SECOND PHASE
~ 148,19
10 0 5 10 15 20 25
~~;::;:t:::~~
~~f-'f
;;;;f:::======~l~~~~~l========~l~~~~I M
Plan 9. Foundation trench of the W terrace wall.
At about the middle of the E rim (2 on Plan 9) there is the trace of a short
straight cutting that begins just in front of the rim, on the interior of the trench
bed. It is at right angles to the trench and runs from E to W. At the begin-
ning, it is shallow but quite clear. Further to the E it deepens, and the edge
coincides with a natural fold in the rock about 3 m. long, which has been deep-
ened in part and modified to accommodate the base of a built pithos of Turk-
ish times (Fig. 13). There is no corresponding working of the rock N of it.
This would in any case have been superfluous as the rock itself here forms a
shallow trench-like hollow, about 0,70 m. wide, at right angles to the trench
of the wall. Some 2,50 m. N of the cutting (Plan 9, 2) there is a crack in the
surface of the rock, comparable in orientation and size (Plan 9, 3), and entirely
89
THE SECOND PHASE
Fig. 13. Working of the rock at 2 on Plan 9, from the W
natural. N of 3, the inner rim of the trench can barely be discerned. Indeed
in a number of places it is not even extant, especially near its N end, which
at an elevation of 147,56 m., is lower by 2,02 m. than the S end. The bed of
the trench is nowhere completely flat and in places it appears not even to have
been prepared. Rough spots in the surface of the rock have simply been
removed, evidently with the hammer, to create a surface more or less level
but hardly smooth. More care has been spent on the cutting of the W rim,
which is more uniform in appearance; yet there the lack of chisel or any other
similar tool marks is notable.
The creators of the trench, however, did not confine themselves simply to
removing protrusions in the rock. They filled in whatever hollows existed,
90
THE SECOND PHASE
Fig. 14. The stones of the terrace wall, in situ, next to 5 on Plan 9.
either in the bed of the trench or just in front of the edges of the trench, with
a mixture of ordinary yellowish mud plaster and gravel or small stones. The
purpose was to level it and to provide a more or less uniform surface for the
blocks of the retaining wall. Hollows filled in, in just that way, have been found
at various places in the trench, especially beneath the few stones of the retain-
ing wall that have remained in situ.
Exploration of the trench indeed revealed not only the bed itself, but stones
belonging to the lowest course of the retaining wall that were still in place.
The stone noted at point 4 on Plan 9 was found built with lime plaster into
an artificial hollow and it belongs to the period of Turkish domination. At
point 5, however, there were two fairly large stones, built on the rock with
simple mud plaster and small pebbles, without any sherds at all, and next to
them a few smaller stones (Fig. 14). Preserved at the beginning of the crack
3 at the W, are two more stones, and at the point where the E corner of the
Promachos base is close to the edge of the trench (Plan 9, 6) a block of large
91
THE SECOND PHASE
proportions remains, the S edge of which covers quite a deep hollow in the
rock. It was excavated earlier by Kavvadias. Lying next to this block is another,
smaller stone, and two more were found further north.
Close to the north end of the trench, at 7 on Plan 9, there is a hollow that
extends some way into the trench. Like the other hollows, this too had been
filled in as described above, and over this were set the stones of the NW cor-
ner of the retaining wall. These stones no longer existed, but in situ were a
number of smaller stones that had been placed as an underpinning for the
lowest course of the wall. To some extent these stones also sealed off the fill-
ing of the hollow which had therefore remained undisturbed, hard and corn- ,
pact. Within it were found a considerable number of sherds which were very
useful for dating the retaining wall; the latest are Mycenaean. 174 To the E of
7, at 8 on Plan 9, a stone was preserved at a spot coinciding with the NE end
of the trench.
There are in addition a few traces of the N leg of the retaining wall as well.
At location 9 on Plan 9, around 4 m. E of 8, a large block was found on the
rock, with two smaller stones beside it, together covering a fill containing
sherds agreeing with the evidence from the previous group. 175 Some 16 m. fur-
ther east along the line of 8-9 (Plan 9, 10), lies another group of stones, built
directly on the rock and unquestionably part of this leg of the retaining wall.
The line 8-9-10 runs in the direction that Stevens had proposed and it follows
a course about 2 m. S of and parallel to the Classical base (see Plan 10, 4).
Thus the retaining wall existed and it was indeed Mycenaean, just as was
the terrace it supported, on which the Archaic temple was later built. 176 Since
the rock slopes from S to N, and the tops of the walls must have been hori-
zontal, the heights of the terrace walls will have varied from place to place.
In order to determine these the height of the terrace must be found. We have
one very valuable piece of evidence for this: the euthynteria of the colonnade
of the Archaic temple which, as we saw, is at an elevation of 152,54 m. Assum-
174. See Appendix II, group 1. Dorpfeld, Jdl 1919, p. 4, according to which
175. See Appendix II, group 2. Mycenaean sherds were found between the
176. See also the information given by preserved walls of the Archaic temple.
92
THE SECOND PHASE
- -~
3 ARCHAIC TEMPLE
:r
149,58 • ..:. - - - - - - - - - - - - - - - - - - -
-
1
_.152 ,40
0 5 10 20 30 40 50
Plan 10. Terrace walls W of the Erechtheion.
ing that the top of the terrace sloped slightly from N to NW to facilitate
drainage, 177 152,40 m. must have been the average elevation of the terrace.
According to this calculation, the terrace wall will have been 2,82 m. high
at the S edge, and 4,84 m. high at the N because of the slope of the rock.
On the basis of the arguments given above, Stevens determined that the
line of the S leg of the retaining wall began at the S end of the trench and
ran parallel to the Archaic temple toward the E. He was most certainly right.
We can determine the length of this leg with accuracy, basing the calculation
on the differences in elevation. 178 The rock rises not only from N to S, but
177. Similar sloping exists in the Myce- 178. Stevens too used this method, but
naean palaces. See Holland, AJA 1924, p. 167. he depended on the altitudes given by Ka-
93
THE SECOND PHASE
also from W to E. Accordingly the wall of the S leg, retaining a horizontal

elevation of 152,40 m., as we have seen, must end at the point where the rock
reaches or approaches that height. This level is reached at location 1 on Plan
10, around 54 m. from the S end of the trench. As for its N leg, which we fol-
lowed as far as the stones at 10 on Plan 9, it will have continued somewhat
beyond the Classical bases (Plan 10, 4) that were erected in front of it.
Were there approaches to the terrace, whose boundaries we have thus
determined to the S, W and in part to the N? Traces of one are preserved at
the W side and, no doubt, there may have been more than one. The two small
quasi-trenches (Plan 10, 2 and 3, Plan 9, 2 and 3), of which 2 is man made, at
least in part, can only be explained as foundations for the support of a stair-
way that will have led to the top of the terrace, at right angles to the retain-
ing wall. 179 Assuming a width of about 0,70 m. for the two supports, using the
width of 3 as guide and in agreement with that of the N stairway of the Palace
of Mycenae, 180 the opening between them will have been 2,40 m. This gives
us the width of the stairway. The elevation of the rock just in front of the mid-
dle of that opening is 148,59 m. The difference between this and the top of
the terrace, estimated at 152,40, is 3,81 m. If this difference is divided into 16
steps, with risers of 0,24 m. and treads 0,35 m. wide, we have a stairway pen-
etrating the terrace some 5 m. This provides an easy ascent to the top of the
terrace in accordance with Mycenaean building practice. 181
werau, and in addition he accepts a some- and the measurements of the steps vary
what lower level for the terrace. greatly. They follow no strict canon, and it
179. The likelihood that a stairway is therefore impossible to restore precisely
existed at this point was suspected also by any Mycenaean stairway or ascent that is
Stevens, to whom I am obliged for the sug- not preserved. In the calculations for this
gestion. stairway I took into account measurements
180. Ergon 1959, pp. 98-99. drawn only from comparable Mycenaean
181. The number and measurements of constructions. What is impossible to esti-
the steps are only indicative. Compared to mate, however, is the asymmetry of the
later examples and to modern ones, Myce- steps. This holds not only for this particu-
naean stairways are steep and difficult to lar stairway, but also for all the others to be
climb. They are never symmetrical even in restored in the following pages.
luxurious constructions. Their arrangement
94
THE SECOND PHASE
To complete the picture of the complex of retaining walls W of the NE

ascent, we must examine the empty space between the termination of walls lb
and 7b (Plan 8) at the W and the eastward continuation of the N leg of the
retaining wall beside the Promachos, east of point 4 on Plan 10. There are no
longer any traces in the area lying between, but a study of the W side of the
Erechtheion, provides some evidence.
It is common knowledge that concentrated in this place, even as early as
Prehistoric times, were all the tokens associated throughout the centuries with
chthonic cult. 182 In the N portico of the building (Plan 11, 1) there are the
traces of the trident (or thunderbolt), and the tomb of Erechtheus. 183 The sea
of Erechtheus was located in the prostomiaion, in the W compartment of the
building (Plan 11, 2). 184 Outside the Erechtheion to the W (Plan 11, 3) grew
the sacred olive tree. 185 The tomb of Kekrops was in the SW corner (Plan 11,
4), 186 and further west, but easily accessible from the Porch of the Maidens,
was the sanctuary of Pandrosos (Plan 11, 5). 187 The way in which all this came
to be concentrated in the building of the Erechtheion, or rather, the way in
which the Erechtheion ·adapted itself to all these things, is most instructive.
First the Pandroseion: remaining today are traces of the foundation of its
peribolos at the N and to some extent at the W (Plan 11, Sa-5~), showing that
the wall was relatively narrow and set at an angle to the W side of the
Erechtheion; the SW corner of the N portico of the Classical building was
adapted to the E end of this peribolos wall. Fragments of broken marble slabs
built into the W wall of the building (Plan 11, Sy) verify the existence and at
the same time give the level of a Classical pavement that was laid in the
precinct after the Persian wars but preceding the construction of the
Erechtheion. 188 Near Sy this paving was founded on a relatively thin layer of
limestone over a fill containing Helladic sherds; 189 the top of the fill had been
182. Kontoleon, To ,Eeix8ewv w~ olxo- 186. Ibid. p. 69.

ooµrJµa xOovia~ ).ar:eda~ (Athens 1949). 187. Erechtheum pp. 119 f., likewise
183. Ibid. p. 81. Paus. I 27, 3, JG 12 372 (Erechtheum p. 286
184. Ibid. p. 34. f.), JG 12 373 (Erechtheum p. 322 f.).
185. Ibid. p. 37. The olive is thought to 188. Erechtheum pp. 125-126.
be a "marker" for the tomb of Kekrops 189. Ibid. pp. 122, 126.
from which it grew.
95
THE SECOND PHASE
• • •
2
I
I I
I I
I
I
0 5 10 20
H=~H~H
~===:j1~~~~~~1 M
Plan 11. The area of the Pandroseion and the Kekropion.
destroyed with the addition of the limestone layer. This had been observed
earlier also along the length of the W wall of the Erechtheion. As noted also
by the Erechtheion excavators, all this shows clearly that in this place there
was a deep prehistoric fill which the makers of the Classical building avoided
disturbing in so far as possible. Specifically, the elevation at location Sy is
148,19 m. and the Helladic fill beneath the limestone bedding for the slab
paving reached a level as high as 149,62, whereas the pavement itself, which
unquestionably maintained the old level, was at an elevation of 150,45. Thus
we may conclude that down to the Persian wars, in the area of the Pandro-
seion there was a fill some 2,26 m. high, which had been formed as early as
Helladic times. A fill this high, however, could not possibly be formed or sur-
vive unless retained by a wall, and it could not remain intact unless the wall
96
THE SECOND PHASE
was continuously there. Yet the N line of the Pandroseion peribolos not only
preceded the Erechtheion but was among the first to be repaired after the
Persian wars. Since it is aligned differently from the Erechtheion, it clearly
was adapted to another, pre-existing construction that had been there for cen-
turies and, as shown by the sherds in the fill, was in fact prehistoric. There-
fore the peribolos of the Pandroseion succeeded a prehistoric retaining wall
that had been preserved as it was or with alterations until the Persian wars.
That wall retained a fill 2 m. high, at an elevation of about 150,45 m. In this
grew the olive tree.
The Kekropion (Plan 11, 4), the tomb of Kekrops, is included in the Pan-
droseion, and takes up the SW corner of the Erechtheion. Its exact form in
Classical times, we do ~ot know. It probably consisted of a mound or small
open-air space. In any case, it was not in the form of a building. 19 Certain it °
is that it was surrounded by a peribolos wall, the E side of which coincided
with the W wall of the Erechtheion and continued as far as the Porch of the
Maidens. 191 Stevens accurately determined the line of this peribolos wall 192 on
the basis of traces preserved on the W wall of the Erechtheion and on the
euthynteria of the Archaic temple. Its elevation may be ascertained from other
traces preserved on the W wall of the Erechtheion within the Kekropion
boundaries. A number of blocks of the W wall of the Erechtheion have coarsely
worked surfaces. Moreover, at the height of the marble beam joining the
Porch of the Maidens with the W wall, there is a marble block with lifting
bosses that have been roughly and carelessly removed (Fig. 15). It is evident
that this unfinished part of the W face of the Erechtheion was not meant to
be seen, for it will have been buried at least to the height of the top of the
bosses, which are at 153 m. 193 Thus the Kekropion was at a level some 2,50
m. higher than the Pandroseion, which implies a second terrace, S of the first
and parallel to it.
190. Dorpfeld, Jdl 1919, p. 7, accepts 193. Stevens, Hesperia XV, 1946, p. 95,
that it was covered by a terrace. fig. 13, accepts a lower level for the Kekro-
191. Erechtheum pp. 127-137, and JG 12 pion as he does not include the last block
372. preserving the bosses among the unworked
192. Hesperia XV, 1946, pp. 93-97, figs blocks.
12-14.
97
THE SECOND PHASE
Fig. 15. The west wall of the Erechtheion. At the level of the big marble beam
and to its left, the wall block with the lifting bosses incompletely removed.
The next question is whether these two terraces of the Pandroseion and of
the Kekropion are associated with the retaining walls la-lb and 7a-7b, and
with the N wall of the terrace E of the Promachos base.
If line 7a-7b is extended toward the W some 5 metres, at point 2 on Plan
12 (at the level of the W wall of the Erechtheion), it meets the NE comer of
the retaining wall supporting the Kekropion terrace. If we extend the line of
the terrace wall from that point westward for about 27 metres, at the level of
the NW corner of the Archaic temple (Plan 12, 3) it meets the N wall of the
98
THE SECOND PHASE
~-----'-'-!
:
...8
. 7'6
147,56 ·
152.40
ARCHAIC TEMPLE
152 54.
149,58.' - - - - - - - - - - - - - - - - - - - '· 152.40
0 5 10 20 30 40 50 100
a:::a::Et===E=;~~;i=:=====~~~=======~;;;;;;;;;;;;~~~;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;~~~~~;;;;;;;;;;;;~ M
Plan 12. The complex of terraces S of the Erechtheion and around the Archaic temple.
terrace behind the Promachos, the course of which we have already followed
to this point. This gives us the entire N side of the terrace. Its elevation (152,54
m. at the E end, about 153 m. in the middle where the Kekropion is located,
and about 152,40 m. at the W end) is uniform, indicating that the _terrace was
continuous and uninterrupted. As is apparent from the curving line of the N
terrace wall, the Archaic temple was built at the very spot where it could be
positioned symmetrically.
The extension of line la-lb around 7 metres westward to point 4 on Plan
12, takes it to the E end of the Pandroseion retaining wall. Continuing the
line known and preserved in Classical times westward brings the wall to point
5 where the Pandroseion wall makes a right angle turn S to meet the higher
retaining wall at point 6 on Plan 12. It is thus the westward extension of retain-
ing wall la-lb, as shown also by the elevation, which at the E end is at 150,40
m., and in the Pandroseion area at 150,45. I consider it most likely that the
line 5-6 is the W boundary of this terrace. First of all, the existence of the
99
THE SECOND PHASE
Classical bases slightly west of 3 means that the terrace cannot have con-
tinued much beyond 5. Secondly, and most significant, the line coincides with
that of the post-Persian rebuilding of the Pandroseion peribolos and since the
inherited arrangement was respected in all else, an innovation here would be
most unlikely.
Thus we have two parallel, graduated terraces, forming long, narrow level
areas with the greatest width in the middle. At precisely that point and to the
N were the venerable sacred places and the "tokens," revered as early as Pre-
historic times: the tomb of Kekrops, perhaps covered by a mound-like con-
struction (Plan 12, 9), near it the sacred olive tree (10), the temenos of Pan-
drosos (11) and the sea of Erechtheus (12). Just to the N (at 13), and prob-
ably enclosed within a peribolos, the marks of the thunderbolt that killed
Erechtheus, or of the trident of Poseidon. These marks, which in Classical
times were kept uncovered and open directly to the sky through the open cof-
fer above them in the N portico (Ta EVT)Avcr1a), could not possibly have been
covered over in Prehistoric times. This perhaps provides yet another bit of evi-
dence for the line of the retaining wall south of them, as calculated.
We cannot know the precise arrangement of this corner of the terracing
during Mycenaean times. Perhaps the peribolos of the thunderbolt marks
communicated in some way with the sea of Erechtheus as it did in Classical
times. Perhaps also at the location of the old entrance to the Pandroseion,
which the architect of the Erechtheion retained by leaving an open doorway
in the SW part of the N porch, there was a stairway leading from the level of
the rock (which is at 148,20 here) to the top of the terrace, N of the olive
tree. The two stones from Prehistoric times preserved in the foundations of
the Erechtheion (Plan 12, 8 and 14) 194 are isolated and so far apart from each
other that only conjecture is possible. One hypothesis is that 14 is a remain-
der of a wall running N to S that separated the sea of Erechtheus from the
olive tree and the Pandroseion, and that 8 was part of a construction believed
to be the tomb of Kekrops.
In my opinion the arrangement of the terraces as reconstructed here
satisfactorily explains the peculiarities of the "building known as the
Erechtheion" 195 and the solutions imposed on its architect.
194. Erechtheum p. 126, fig. 80, and pl. IV. 195. Paus. I 26, 5.
100
THE SECOND PHASE
THE AREA SE OF THE END OF THE NE ASCENT
Wall lla (Plan 13) lies E of wall la-7a, opposite to it and roughly in line
with it. The N part of the wall is built of large, heavy blocks, still visible today.
Since the rock itself rises toward the S, the stones become progressively
smaller in this direction while the width of the wall, as much as 1,10 m.,
remains the same. This is to be expected, since the steeply rising gradient of
the rock meant that the mass and weight of fill the wall had to retain was
much less. It stops at a point corresponding to 7a on Plan 12. Both walls end
against an abrupt rise of the rock; continuation to the S is not preserved.
At its N end, wall lla makes an approximately right angle turn to the E,
continuing in a straight line (Plan 13, llP-lly) to a point near the brow of the
rock. A number of constructions were discovered S of this, some of them built
on the inner side of the wall and plotted by Kawerau. 196 Clearly they belong
to Turkish times and they are not included on the plan.
Like the walls W of it, lla-lly is thus a retaining wall, supporting a ter-
race and facing N like the others. Its Send is founded at 150,79, its NW cor-
ner at 149,57, and at the middle of its N side it is at about 150,02. In order
to determine its height, we shall have to examine the area of the rock to the
S of it.
Some 14-14,50 m. to the south, parallel to it but extending further E, is a
second retaining wall (Plan 13, 12). Its easternmost preserved end stops at a
point roughly opposite the Wend of lly. The N side of this wall is shown by
Kawerau 197 as a straight line. Cleaning this side of the wall, however, showed
that the line is irregular, following the conformation of the rock. Its exterior
side is constructed of relatively large blocks measuring between 0,85 x 0,30 and
0,40 x 0,25 m., with smaller stones in the interstices (Fig. 16). The wall varies
in width, ranging from 1,40 to 1,80 m., considerably thicker than noted by
Kawerau. The interior face of the wall is made up of much smaller stones so
as to adjust it to the uneven surface of the steeply rising rock. In other words,
the lower courses of retaining wall 12 were carefully adapted to the steeply
rising surface of that part of the rock, the top of which to the S coincides with
196. Kavvadias-Kawerau pl. !l.. 197. Ibid. pl. !l., 43.
101
THE SECOND PHASE
--~ ........ ,
11y '
''
',~15
' .....
I
I" 151 ,29
.....
I
I
\10 - - - - - -
0 5 10 15 20 30
SH~H~H~====s:====~====~
l===========~I M
Plan 13. Te"ace walls E of the Erechtheion.
the highest point of the Acropolis. All the stones of the retaining wall, espe-
cially the smaller ones on the inner side, are set with a plentiful amount of
yellowish clay, totally unlike the usual sort of earth lying on the rock. The clay
layer had not been disturbed in the 1887 excavation and it contained sherds.
These were collected from between the lowest courses of the wall where they
clearly had not been damaged since the building of the wall, thus providing
material for its dating. 198
198. See infra, Appendix II, groups 3 and 4.
102
THE SECOND PHASE
Fig. 16. Te"ace wall 12 of Plan 13 from the NW (phot. DAI n° 50).
At its W end, the wall stops without coming to a finished end. No doubt
it will have continued further west, although we cannot determine precisely
how far. Its line toward the E was interrupted in Classical times by the set-
ting of a series of regular rectangular conglomerate blocks, so that here too
the original end of the wall is missing.
Wall 12 is founded directly on the rock, which slopes steeply toward the N
and less so toward the E. Thus the level of its foundation ranges from an elev-
ation of 152,53 m. at the Wend to 151,68 m. at the E. The elevations enable
us to determine the level of the terrace supported by walls lla-lly as being
at about 152,50 m., probably with a slight slope to the E for the draining of
water. This level agrees with the elevation of the top of the large terrace to
the W. In addition, it enables us to determine, on the basis of the elevations
of the rock, the approximate termination of lla at the S and 12a at the W, as
shown at 10 on Plan 13.
Beside the E end of retaining wall 12 there is a wall about 1,40 m. wide,
today covered over. It is built of medium sized stones with yellowish clay of
the same sort as that used for terrace wall 12 and it contained comparable
103
THE SECOND PHASE
sherds (Plan 13, 13). On Kawerau's plan, 199 the S end appears to stop short
of the N side of 12. The measurements of the stones given are smaller than
they actually are and the angle of the corner is in fact wider than that shown.
There is no reference to it in the text. When it was excavated it was evident
that it formed a comer slightly wider than a right angle, and that it had two
unequal legs. The longer of the two, about 5 m., is perpendicular to 12, on
which it abuts. The shorter leg turns W and continues in that direction for
some 3 m. The rest of the wall is not preserved, but it is evident that it was
destroyed and that wall 13 did not end where it does today. On top of it were
remains of a wall built during the Turkish domination, not shown on the plan.
The direction of 13 and the fact that it is Mycenaean and has been built
against the N side of 12, makes it certain that it is the SE end of lly. In fact,
if the line of 11 y is continued along the line dictated by the NE edge of the
rock, which 11 follows elsewhere, it meets the westward running leg of 13.
Judging by its position and the given elevations, 13 will have been built to a
height of less than 1 m. Its construction resembles exactly that of the other
similar low retaining walls, recalling the S end of lla.
Whether the stones at 14 on Plan 13 actually belong to a Mycenaean build-
ing, is uncertain. They comprise two very short stretches of wall slightly out
of line with each other and they were noted by Kawerau. It is most unlikely
that they belong to a retaining wall built between and parallel to 11 and 12,
since in that case wall 11 could not have been higher than 1,40 m., and such
large stones and heavy construction would have been unnecessary.
In the preserved section of retaining wall 12, the tops of the blocks at its
Wend are at elevation 154,13 m., in the middle at 154,37 m. and at its E end
at 154,09 m., giving an average level of about 154,10 m. As we have already
noted, to the S and E of the terrace lies the oval hump of the rock with its
top at elevation 156,16 m. (see Plan 16). The extent of rock surface that
remained uncovered will have depended on the height of the terrace wall and
consequently also on the level of its top S of the terrace it retains. If the top
199. Kavvadias-Kawerau pl. ~.
104
THE SECOND PHASE
of the retaining wall as preserved today was actually the uppermost course,
the exposed surface of the rock will have been the same then as it was in Clas-
sical times and the wall will have had an average height of 1,60 m. If another
course is restored, the height of which must be calculated at 0,70 m. on the
basis of the blocks preserved, the wall will have been about 2,30 m. high. In
this case the top of the terrace will have had an elevation of about 154,80 m.
and the exposed surface of the rock will have been considerably less. The
traces preserved yield no positive conclusion on this score.
THE NE ASCENT
The deep fissure that cuts into the mass of the Acropolis rock from E to
W, ending at the top of the Acropolis E of the Erechtheion, forms a natural
access which will, indeed, have been used from the earliest times. While pass-
able, it is very narrow and steep. It requires climbing rather than walking, and
it is totally inadequate for animals, especially pack-animals. This in itself shows
that it could never have been the main entrance to the Acropolis as has been
thought by some.
The track had been made into a regular ascent during Late Helladic times.
Precisely when this was done, we do not know. We know, however, that dur-
ing the time of the terrace walls we have been discussing, the pathway had
already been formed and was in service as a supplementary approach, wind-
ing, as it did, through the anomalies of the rock.
The Mediaeval fortification wall of the Acropolis, built on the line of the
Classical wall, divides the NE ascent into two unequal parts: the western part
which is inside the wall and the longer, eastern part which lies outside it. The
first was excavated by Kavvadias in 1887.200 It was found covered over by a
fill with sherds showing that it was buried when the Themistoclean wall was
built. 201 The second was cleared in 1931-1934 by Broneer whose explorations
verified Holland's conclusion that the entrance was blocked when the Myce-
naean fortification wall was erected. In fact, his excavation showed that the
200. Ibid. pp. 33, 89, pls r and {),., 201. AM 1887, p. 141.
105
THE SECOND PHASE
0 5 10 20 30 40 50
~i=p~;;;;;;;;;;;;;;;;;;;;;t==========l~;;;;;;;;;;;;;;;;;;;;=:;;;;;;~:==========l=-m;;;;;;;;;;;;;;;;;;;;;;;;;~~M
Plan 14. The northeast ascent.
eastern part of the ascent had been covered over by little buildings contem-
porary with the construction of the wall, and the last steps beside the Classi-
cal wall were buried beneath a pure Mycenaean fill of that same time. 202
The ascent (Plan 14, 3) begins some 10 m. east of the Peripatos inscription
(Plan 14, 4) at the point where it meets with the Peripatos itself (Plan 14, 5),
at a level of about 125 m. It ends at the top of the rock, between retaining
walls 1 and 2 at a level of 147,57 m. The slope it had to surmount is precipi-
tous indeed,203 and for this reason steps and ramps were constructed in order
202. Hesperia II, 1933, pp. 351-355; IV, average incline of 34%. The slope has been
1935, pp. 109-113. modified in places by surface reworking,
203. Over the total length of approxi- but it is none the less exceedingly abrupt,
mately 66,50 m. , the pathway shows a dif- especially for frequent use.
ference in elevation of 22,57 m., that is an
106
THE SECOND PHASE
to make the climb easier. The first section of it, beginning at the Peripatos
and ending at about the middle of its course, is a steeply ascending path. At
this point the first step is encountered, constructed of small flagstones. Then
comes a section destroyed by the Mycenaean houses that were later built on
top of it (see infra). A little further on are three similar steps, the tops of
which have been trodden smooth. 204 Some 7 metres further west are another
seven steps, followed by a ramp leading to the point where the fissure in the
main mass of the rock begins. From here on the ascent is by means of a con-
tinuous stairway, cut by the later fortification wall, with steps in some places
built, in other places cut into the rock. Their measurements vary to such an
extent that they are totally unequal and irregular. The rock S of this last sec-
tion of the ascent rises steeply and evenly to the level where terrace II (Plan
16) is located and where later another section of the north Cyclopean wall was
added (Ton Plan 14). This applies also to the narrow space between walls 11
and 1, which is not accessible from the stairway. N of the ascent, the rock is
lower, but not climbable, and it is broken into a series of irregular masses by
deep fissures. The last step to the W is the widest, leading to a landing that
brings the climber to the passage between terrace walls 1 and 2. This charac-
teristically Mycenaean ascent, making use of the formation of the rock itself,
served the Acropolis throughout its entire second phase until the building of
the fortification wall.
THE NORTHWEST DESCENT TO THE CAVES
A Mycenaean descent is partially preserved on the low NW plateau of the

rock where the caves of Pan, Apollo and Zeus Olympios are situated. It runs
through the rough and uneven rock from the top down to the narrow shelf of
the plateau, to the mouth of the cave of Pan. Kavvadias was the first to exca-
vate the area, and Keramopoullos later carried out supplementary exploration
and cleaning. 205
204. Broneer, Hesperia IV, 1935, p. 113. pls 1-4, and Keramopoullos, Deltion 1929,
205. Kawadias, Ephemeris 1897, pp. 1-32, pp. 86-88.
107
THE SECOND PHASE
That the descent was constructed and used during Mycenaean times is evi-
dent from its construction, the remains preserved in situ, its layout and the
fact that it preceded the Classical wall, the building of which altered it at the
top of the rock. The lack of relevant material, however, makes it difficult to
determine the exact Late Helladic phase to which it belongs. Yet it seems to
me most unlikely that this descent would not have been in use when the
Acropolis was as systematically inhabited as it was during the period we are
investigating. I believe it more logical to assume that its construction was con-
temporary with the terrace walls rather than with the later third phase.
The point where the descent is next to the base of the Classical wall, which
coincides approximately with the beginning of the descent, is at a level of
137,14 m. There are no steps for the first 10 m. of its downward course (Plan
15, 1). They would in any case have been unnecessary since the rock, while
far from being "virtually horizontal," 206 slopes gently, so that a ramp was suf-
ficient. Preserved at 2 on Plan 15 is a stone that appears to belong to the
retaining wall of that part of the descent. 207 The incomplete rectangular con-
struction at 3, not mentioned by the excavator, should be much later since it
could not possibly have stood where it is while the retaining wall was extant.
Slightly W of 2, a series of uneven steps of various sizes (Plan 15, 4) begins.
They are cut into a natural cleft of the rock, 208 the course of which they fol-
low. Between the 16th and the 22nd step from the E, the bed of this cleft has
a deep and abrupt fissure. Here there are cuttings in each side, wider to the
N, that were made to accommodate built steps to bridge the empty space
between.
After the 23rd step, the stairway is interrupted by a four-sided construc-
tion of later times, partly cut into the rock (Plan 15, 5). Its construction cut
off the four curving steps (Plan 15, 6), which make a detour to the N in order
to continue the descent, whose traces are encountered again just W of 5 as a
series of eight more steps. These stop next to the precipitous brow of the rock
206. Kavvadias, Ephemeris 1897, p. 27. down (Ephemeris 1897, pl. 1, ~)and having
207. Kavvadias does not refer to that no relation to it.
stone nor does he mark it on his plan. In- 208. In his text, Kavvadias (p. 26) men-
stead, he attributes to the retaining wall tions 17 steps; on his pl. 1, however he
another series of stones, found much lower records 31, which is the correct number.
108
THE SECOND PHASE
0 5 10 20 30 40
M
Plan 15. The area of the northwest descent.
(Plan 15, 7), just before the cave of Pan (A on the Plan) where the partly
destroyed chapel of Aghios Ioannes Alaniares (St. John the Tramp) (Plan 15,
8) stands today. From this point on, the rock toward the W is virtually hori-
zontal, as is evident from the elevations marked on the plan.
From 1 to 7 the course of the path is clear. The question is whether it con-
tinued on down, and what this continuation might have been.
If it continued, it will have started at the level of the open area at 7 and,
descending the side of the rock, it will have come down to about the level of
the Peripatos. The brow of the rock at this point is about 10 m. above the
base, and the wall of the rock here forms an almost vertical cliff. To overcome
this difference of level, it would have been necessary to have either a ramp
parallel to the side of the rock and at least 50 m. long in order to obtain a
20% incline (twice that of the great E ascent at Tiryns), or else a stairway at
least 12 m. long. 209 Such large scale and massive constructions could hardly
209. Hypothesising steps with tread- which would make the climb anything but
boards 0,30 m. wide and risers 0,25 m. high, easy.
109
'
0 10 50 100 150 200 250 300 350

M
Plan 16. The Acropolis at the end of the second LH building phase.
THE SECOND PHASE
have disappeared leaving no trace at all. In fact, for such a construction, we

might even have expected to find a relevant comment in the ancient sources.
We may therefore exclude the existence of a descent from the brow to the
base of the rock. The path clearly ended near 7 and its sole purpose was to
connect the top of the rock with the NW plateau and the caves.
THE SECOND LATE HELLADIC PHASE AS A WHOLE
During the span of time to which the second building period belongs, the
Acropolis was still unfortified. On top of the rock at the N is a complex of
terraces forming a series of stepped level areas. On these we must visualise
the palace buildings and other installations. The tombs, to be sure, are not
visible. So too the LH I house which, judging by the sherds in the fill that cov-
ered it,210 had also been abandoned and buried.
There are five terraces. At the N the smallest of all, I on Plan 16, is an
irregular four-sided one in plan, its top probably at 150,45 m. Opposite it to
the S, with its top at the same level, is the long, narrow terrace II, the W end
of which includes the "martyria," the "tokens" of the time honoured cult on
the rock. S of II and higher is the large west terrace, III, with its top at a level
of 152,50 m., and with a stairway in the middle of its west side, serving the
old and natural approach to the summit of the rock from the W. Here, where
the Archaic temple stood, and, later on, part of the Erechtheion, the main
complex of the Mycenaean palace must have been located.
Opposite II and III and separated from them by a narrow strip of the rock
itself, left free and perhaps intended as a water run-off and for draining the
area, are the two terraces to the E. To the N is IV, dominating the final sec-
tion of the NE ascent, and with its N side following the configuration of the
rock so that it has an irregular plan that is almost triangular; its top is at the
same level as III. Rising above this to the S is terrace V, contiguous and on
the highest part of the summit of the rock. These terraces must have com-
municated with each other and with the rock on which they stood, so we must
210. Holland, AJA 1924, p. 155.
111
THE SECOND PHASE
assume the existence of steps leading from one level to another. The only one
the position of which is certain is the W stairway of terrace III. Not a trace
remains of the others. The only existing clues are the preserved sections of
the supporting walls of the terraces. This is purely negative evidence since,
being walls, they show us exactly where such stairways could not have been.
Thus, if we examine the various terrace walls (see Plans 8, 12 and 13), we see
the following:
Terrace I, isolated as it is from the others, must communicate with the
rock. This can be ruled out on the N and E sides as the rock here falls off
sharply; likewise on the W, where later stood the wall that was founded on
the LH I building. Only the S side remains. The E part of the terrace wall is
preserved to about the middle of this side. Thus the stairway to the top of ter-
race I will have been either in the middle of the S side or near the W end. It
was most likely in the middle.
The stairway connecting the rock with terrace II should, for the same rea-
sons, have been at a spot about opposite to the stairway of I. m has a stair-
way at the W; at the S no stailway is needed since the natural slope of the
rock rises gradually to the top of the terrace. By the same path and in the
same way, terrace IV is reached near its SW corner, and terrace V by way of
its SW end. Thus terrace III communicated with IV and V without need of a
stairway.
Several locations are possible for the stairway connecting II with m. One
is near its E end, between walls 7p and 7y on Plan 8, roughly opposite the
ascent from the rock to terrace II. The other or, more likely, the others, are
to be found further west between 7y on Plan 8 and the area of the "tokens"
the latter seems more likely. There remains the ascent from IV to V, which
can only have been at the westernmost part of the two terraces, between 10
and 12a on Plan 13.
These stairways were needed to provide circulation in the area of the ter-
races. The possibility that there were others as well is not excluded, but to
determine how many and where they were would take us into the realm of
pure conjecture. 211
211. J. Travlos in his Pictorial Dictionary restores one more terrace to the south of
of Ancient Athens (1971), plan on p. 57, 11-m, roughly in the location occupied later
112
THE SECOND PHASE
These terraces, which covered the top of the rock, explain the fpr£81~ov
of Cleidemos as well. They "levelled" the area indeed, but they levelled by
construction and filling, not by quarrying.
The ascents to the Acropolis were two. The main and most gradual ascent
was that on the west, preserved and reformed again and again through the
years that followed. There was, in addition, the auxiliary NE approach (Plan
16, 2), the one ending between terraces I and II. The NW descent (Plan 16,
3), as we have seen, went down only as far as the plateau of the caves.
CHRONOLOGY OF THE SECOND PHASE
An initial, though somewhat indefinite, dating of the terrace walls is pro-

vided by observations of the building sequence. It is a fact that they were con-
•
structed after the house to the N of the Erechtheion, which goes back to
LH I times, and they are earlier than the Cyclopean wall which, as we shall see,
was built toward the end of the LH IIIB period. This fairly wide range of pos-
sibility is significantly limited by the sherds recovered from four different
places in the terrace walls. They provide a much more precise dating. 212
The first find was in the shallow depression next to the NW end of the
west wall of terrace III (see Plan 9, 7). This, as we have already determined,
was filled in so as to grade the rock for the foundation of the terrace wall.
From then on it remained undisturbed. Among the sherds in the fill were EH,
MH and LH, for the most part early. The latest of all (Appendix II, group 1,
i) is the foot of a kylix of the initial years of LH IIIB. Found E of that depres-
sion, beneath stones of the terrace wall lying in situ (Plan 9, 9), was another
group of sherds, the latest of which is no later than advanced LH III times
(Appendix II, group 2, d, e ).
Two groups of sherds were collected from the point where the E wall of
terrace IV meets the N wall of V. One came from among the stones of the
wall itself, the other from beneath the stones of its foundation. The latest
by the Parthenon. Although not improba- left any material evidence whatsoever.
ble, this hypothetical construction has not 212. See Appendix II, groups 1-4.
113
THE SECOND PHASE
sherds in the first group (Appendix II, group 3, h, i) are LH I-II and the lat-
est and most characteristic of the second group (Appendix II, group 4, d), with
a clumsy attempt at panel decoration, which would date it in LH IIIBl rather
than in IIIB2.
The conclusion to be drawn from the pottery is that the terrace walls must
have been built in LH IIIA2 or early LH IIIB times. 213 Most of the sherds
belong to earlier periods, a further indication that the LH IIIB style is still at
its beginning. The first, unfortified, phase of the Mycenaean Acropolis of
Athens thus begins with a considerable delay after the initial fortifying of
Mycenae and the first period of Tiryns.214
213. Cf. Mountjoy 22-24. 1938, pp. 555-559, and Ergon 1959, pp. 93,
214. For the dating of Mycenae and 96-97, but mainly G. E. Mylonas, Mycenae
Tiryns, see among others Mackeprang, AJA and the Mycenaean Age (Princeton 1966) 33.
114
3. THE THIRD PHASE
THE TOP OF THE ROCK
THE WEST SIDE
The approach to the rock from the W is fairly gradual. At the SW end,
however, is a low but prominent protrusion that forms a sort of natural bas-
tion. When the Acropolis was fortified, the tower of the W entrance was built
on that projecting height. Later on, in Classical times, the bastion was hidden
in an ashlar sheathing of poros blocks. On top of this was built the temple of
Athena Nike.
The Classical encasement left little of the Cyclopean bastion visible.
Indeed, up to the time of Balanos' restoration, it had not even been noticed.
On the N side of the Classical bastion, some 9-10 m. from its NW corner, the
covering blocks are so close to the fa~ade of the Mycenaean tower that instead
of the usual alternating headers and stretchers, the blocks were laid only as
stretchers with false joints cut into the outer face. At the height of the ninth
and tenth course from the top, where there was no room even for stretchers,
an opening was left in which the stones of the Cyclopean wall could be seen
projecting (Plan 17, 1). Preserved above this, on top of the Cyclopean blocks
is part of a polygonal wall (Plan 17, 2). This is the remainder of the Archaic
rebuilding of the bastion and it is constructed of Acropolis and Karra lime-
stone. Still another section, built of limestone (Plan 17, 3), is preserved at the
E end of the bastion. It is 3-3,50 m. W of and virtually parallel to the leg of
the Mycenaean wall S of the Propylaia, which it faces. Only one course
remains and today it is no longer visible.
Most of those who have previously studied this part of the Acropolis
accepted the idea that the bastion of historical times was built as part of the
fortification. It was thought to be earlier than the construction of the temple
of Athena Nike, but no date was suggested and there was no reference to the
115
THE THIRD PHASE
visible Mycenaean remains. 215 Koster even denied the possibility that the
Mycenaean fortification had a bastion. 216 Kawerau interpreted wall 2 as a con-
tinuation of 1, contemporary with the Archaic terrace wall W of the Propy-
laia. 217 He described wall 3 only briefly without relating it to the other walls. 218
The first to explore the interior of the Classical bastion systematically was
Welter, who wanted to find out if it had been built before or after the Propy-
laia. At the depth to which he excavated, he established that there was indeed
an earlier Archaic construction. Yet he did not reach the Mycenaean re-
mains.219 These were discovered later, when Balanos was carrying out the
restoration of the temple of Athena Nike. In the course of consolidating the
Classical bastion, which had settled, he removed the outer part of the wall in
sections. 220 The task was continued by Orlandos along the E part of the S side.
The bastion of the Mycenaean fortification that now came to light is rec-
tangular in plan, with an E-W axis. All of it is known but the NE part, which
was never exposed.
Only the S half of its E side was found, and only one course of this is pre-
served. This is the wall shown on Plan 17, 3. The S side is roughly parallel to
the Classical encasement, but it is not in a perfectly straight line. Apart from
two small gaps next to the SE corner, the line of the wall is preserved com-
pletely. At its SW corner it makes an acute angle turn toward the NE, follows
a fairly straight line for about 10 m., then makes an obtuse angle turn and
runs E, parallel to the N Classical encasement, to 1 on Plan 17. There it makes
a very wide angle turn, continuing toward the E along line 2, thus following
for about 5 m., a line parallel to the S side.
The outer sides of the Mycenaean bastion are constructed of massive
stones, often measuring over 1 m. Small stones and earth made up the corn-
215. Robert, Aus Kydathen p. 182, Lol- 218. Ibid. p. 139 and pl. H.
ling, Topographie p. 338, Miller, AJA 1893, 219. AM 1923, pp. 190-201, pls IV-V.
p. 486, Picard, L'Acropole I pp. 18-19, Ju- 220. Ephemeris 1937 f (Athens 1956),
deich, Top. p. 218. pp. 776-807. See also BCH, Chron. 1936, p.
216. Pelargikon p. 9. 455; 1937,p.443; 1938,p.448; 1939,p. 289,
217. Kavvadias - Kawerau pp. 129, 137 and Welter, AA 1939, pp. 1-22.
and pl. H.
116
THE THIRD PHASE
a
0 1 2 3 4 5 10
~Ht=lr--i~=lF-'1~==::EI~~~~~I M
8 ~ ~· 13 1,89
132,93 ·6 0 0 o0 G0~
Plan 17. The west bastion of the fortification and the area around it.
117
THE THIRD PHASE
pact interior of the wall. At about 4,50 m. E of the W outer wall and paral-
lel to it, a solidly built wall runs across the entire width of the bastion. It is
founded on fill and constructed of two rows of large stones. Its purpose was
evidently to retain the fill to the E of it, thus lessening the weight to be sup-
ported by the W side of the bastion. Another similar cross-wall east of the
first will not have been needed, since the rock rises toward the E and the
weight of the fill lessens correspondingly. The fill consists of stones, decreas-
ing in size toward the center of the bastion, and earth fill increasing corre-
spondingly toward the centre. Thus the centre of the tower, the space between
the outer walls, was full of earth mixed with small stones. 221
The greatest preserved height of the bastion is at the west side, where it
stands to 3,80 m. above the rock, which at that point is 135,45 m. above sea
level.
As noted, the E side of the bastion is not preserved entirely. The exterior
face of the section of the Cyclopean wall to the E of it and facing it, however,
is preserved for its full length without interruption well beyond the N side of
the bastion. This makes it clear that the bastion did not join the fortification
wall at that spot and that there was a narrow open space between them. The
wall is preserved to a much greater height than is the E side of the bastion.
Since the purpose of the bastion was to cover and to protect the fortification,
it cannot have been significantly lower than the wall itself. The only possible
conclusion is that it is preserved to less than its original height, which will have
coincided with that of the wall.
The W side of the bastion is the best preserved. Its construction in this
place is of special interest. Found in the fa~ade of one of the blocks of the
lowest course, about 1,50 m. from the NW corner, was a shallow rectangular
hollow measuring 0,30x0,22 and 0,15 m. deep (Plan 17, 4). There were traces
of fire and burned sherds on the rock beside the edge of the hollow,222 clear
indications of cult activity. S of this cutting is a large built niche in the foun-
dations of the bastion, the carefully levelled rock serving as its floor.
Balanos reports the following about the niche. 223 It is "about 5 m. long",
but its depth could not be ascertained because the stones of the interior had
221. Balanos, Ephemeris 1937 f , p. 788. 223. Ibid. pp. 790-791, figs 20-21.
222. Ibid. p. 791 and fig. 22.
118
THE THIRD PHASE
collapsed. It was found blocked by dry masonry, the surface of which is set
back from the fac;ade of the bastion by 0,15-0,20 m. Built in at one place with
the other stones was a poros pier that reached the ceiling of the niche. Both
pier and dry masonry were placed there when the Classical sheathing was
built. The pier evidently replaced an earlier little round column or stele, as
shown by the circular cutting in the rock on which it stood. The W face of the
support measured 0,60 m., its width toward the interior was 0,35 m. and its
height 1,25 m. Balanos' plan224 shows the position of the pier (see Plan 17, 5),
but not the niche. Nor does he provide the diameter of the round cutting
beneath the pier, which held the stele that preceded it. This, however, can be
calculated at 0,45-0,47 m. on the basis of his figs 20 and 21, and by analogy
with the 0,60 m. face of the pier. The measurements are of course only approxi-
mate, but since the restoration of the bastion has hidden the floor of the niche
there is no other way to make a calculation. Problematic also are the exact
positions of the niche in the W fac;ade of the bastion and of the pier within
the niche. Since the niche 225 is not included in Balanos' survey, and since its
224. Ibid. pl. 1. justify the discrepancy by assuming that the

225. For the niche, James Wright fol- measurements given by Balanos involve
lows Mark (The Sanctuary of Athena Nike both the real and the rejected niches plus
in Athens: Architectural Stages and Chronol- their surround, as he calls it, presumably
ogy, Hesperia , suppl. 26, Princeton 1993, pp. the boulders along their outline. In this he
13-14) who, based on some preliminary was betrayed by his imperfect knowledge of
drawings by Balanos, kept in the Archives modern Greek. Balanos' unequivocal state-
of the Archaeological Society at Athens ment reads, «To oA.Lxov µ'fixrn; "t'fi<; aQxaw-
and made while the work was in progress, "tEQac; "tUU"t'fl<; XOYX'fl<; Elvm 5 µ. JtEQGtou ,
assumes that what he calls Balanos' archival EvW "tO "t'fi<; "tOU JtEQL~A~µa"tO<; Elvm 3,135
drawing shows two niches side by side and µ.». By JtEQL~A.'flµa Balanos means through-
concentrates on that on the right hand (illu- out his report the Classical porns sheathing
strated by Balanos in his figure 20), which of the bastion, so what he says is that the
he accepts as being the only one. His meas- total width of this earlier (i.e. Mycenaean)
urements on the drawings led him to be- niche is ea. 5 m., while that of the one left
lieve that its width does not exceed 1.865 in the later sheathing is 1,135 m. Mark's
m. , which is far off the width of ea 5 m. interpretation and his resulting computa-
given by Balanos in his final and therefore tions, as also those of Wright, are based on
only authoritative account. Mark tries to a misunderstanding.
119
THE THIRD PHASE
,
I
I I
0 0,80t (.:
4 l-
I
- 4,65 --·
I
I
Plan 18. The niche in the W f a~ade of the bastion.
edges are hidden by the mortar that consolidated the W side of the Classical
bastion, we must resort to the following estimates.
SOUTH EDGE OF THE NICHE: The photograph reproduced as Balanos' fig.
20, taken from in front and therefore with no real optical distortion, shows
the pier and the S edge of the niche. A comparison of the width of the face
of the pier and the distance from its S edge to the S edge of the niche, sug-
gests that the distance is around 0,90 m.
NORTH EDGE OF THE NICHE: Preserved within the Classical encasement of
the bastion is the NW corner of the Mycenaean bastion. It is visible for its
full height, together with the part of the rock on which it stands. The cutting
can be seen at 4 on Plans 17 and 18. The modern mortar connecting the fa~ade
of the Cyclopean bastion with the interior of the restored Classical encase-
ment, makes the middle section where the niche was located inaccessible. The
mortar starts precisely 0,80 m. S of the cutting, with no sign in this space of
the beginning of the niche. But the S side of the last visible limestone block
S of the cutting is vertical, and the cement mortar follows immediately. This
may have been the last wall block before the niche. In any case, it is absolutely
certain that the N side of the niche is 0,80 m. S of cutting 4, if not more.
The S side of the niche, therefore, is at a point 0,90 m. south of the pier;
the N side, at a point not less that 0,80 m. S of the cutting (see Plan 18). The
distance between these two points, as determined and measured on the plan,
is somewhat over 4,60 m. Bearing in mind that our calculations can only be
120
THE THIRD PHASE
approximate, and allowing for the imprecision of Balanos' relevant measure-

ments as shown on his plans, this measurement is not far from the «approxi-
mately 5 m.» that he gives.
However imprecise the calculations may be, one thing is certain: the pier,
or rather the earlier little column, is not in the middle of the niche but close
to its S side, at about 1/ 4 of its total width. This means either that there was
only one roof support in the niche - the height of which can be derived from
the height of the Classical column at 1,25 m. - or that there was also a sec-
ond, similar one, placed symmetrically toward the N side. This second
arrangement is surely the most satisfactory and it is shown on Plan 18.
In addition to the niche, there are other traces on the W side of the bas-
tion. These were found not on the bastion itself, but on the rock on which the
bastion stands. They are at a lower level than the foundation of the bastion
and 2-2,50 m. W of the fac;ade. They consist of massive stones (Plan 17, 6, 7)
set on the rock and now partly concealed by the Classical encasement at the
height of the 18th course from its top. They are not at the same level. The
stones at 6 are at 133,85 m., those at 7 at 133,40 m. Thus they rise slightly
from S to N, following the configuration of the rock to the foot of the bas-
tion.
In removing the Classical bastion, at the 18th course and the level of these
enormous stones, Balanos made the following discovery. Whereas the space
between the Cyclopean bastion and the Classical facing elsewhere was filled
with poros stones, at precisely this height there was a very carefully laid layer
of hard Piraeus stone that continued E to the rock on which it is set. Beneath
this layer, the filling was again more or less rough. 226 The layer was not a
chance deposit. It belongs, moreover, to historical times. Yet the existence
and especially the preservation of the enormous stones at 6 and 7, show that
it was a later reworking of an already existing Mycenaean layer, which was
1,50-2 m. lower than the base of the bastion and ran from S to N. As Stevens
has already observed, 227 this is a Mycenaean approach to the foot of the bas-
tion. It was retained by a supporting wall of which the stones at 6 and 7 are
remnants.
226. Ephemeris 1937 r , p. 791. 227. Hesperia XV, 1946, p. 76 and fig. 2.
121
THE THIRD PHASE
Below these traces, almost 2 m. further down the rock, is a series of deep
cuttings (Plan 17, 8) for the convenience of those ascending. Long use has
deepened them more in the centre than at the sides and rounded out the
edges. As Bohn rightly observed, it is an ascent intended for animals. 228
The first to discover these traces was Beule229 when he removed the slabs
of the Mediaeval ascent that covered it. There is no doubt at all that the cut-
tings go back to Mycenaean times. In fact, they are much further down than
the lowest courses of the Classical Nike bastion, which, coarsely worked as
they are, were not meant to be seen and will have been buried. They are even
lower than the top of the Archaic terrace wall W of the Propylaia, which
formed a ramp leading to the Acropolis entrance and covered over the cut-
tings. Thus they are earlier than the Archaic ascent and the surface wear
shows that they saw centuries of use before they were buried.
As preserved today, they begin below stone 6 (Plan 17), at a level of 131,89
m., and they ascend toward the N, gradually diverging from the W side of the
bastion to an elevation of 132,93 m. Visible today are 10 cuttings. Kawerau
and Bohn230 record 11 and Bohn noted that the last ones tum eastward.
Weller accepted this, noting the beginning of a 12th cutting beneath the first
step of the stairway to the Propylaia. 231 The lOth cutting, measured with the
greatest possible accuracy, is somewhat further E than is shown on Bohn's
plan, but it does not agree at all with Weller's measurements. If the ascent
turned toward the E continuing along the N side of the bastion, as shown by
Bohn and Weller, the existing differences in elevation, would mean an incline
of 38,5%. This may be totally ruled out, especially for animals and certainly
for loaded draft animals.
The direction of the cuttings, as indicated by the lOth, the last one visible
now, and also by the 11 th as recorded by Kawerau, is obliquely NE, in the
general direction of the Agrippa base. As shown by the measurements of the
cuttings that are accessible, there is an incline of 18,25% from the lst to the
lOth. They run in a direction completely different from the ascent at 6-7 (Plan
228. Prop. p. 15. Prop. p. 15, pl. II.

229. L 'Acropole p. 85. 231. AJA 1904, p. 68, pl. I.
230. Kavvadias-Kawerau pl. H, Bohn,
122
THE THIRD PHASE
17) and their distancing from the bastion means that they are following a gen-
tler slope than did the other path. Thus we have traces of two separate ascents,
each with a different course and a different degree of incline.
The first stones of the Cyclopean wall visible to those approaching the
Propylaia, lie N of the bastion beneath the foundations of the Pinakotheke.
It was Stevens who discovered them. He collected sherds, all prehistoric, from
the mud mortar binding the stones to the rock. The stones he attributed to
the W leg of the wall. 232 A systematic clearing of the area revealed other traces
of the wall in addition to the stones.
At this place, the rock forms a brow that runs out from beneath the W
wall of the Pinakotheke, passes the Agrippa base to the W, which is founded
on the slope, and runs obliquely toward the SW corner of the exterior fac;ade
of the Propylaia (Plan 19). The part of the rock that lies between the central
entrance of the Propylaia and the temple of Athena Nike was cut at a num-
ber of places in historical times for the founding of various bases. Traces of
the Cyclopean wall, however, are clearly preserved between this entrance and
the Pinakotheke. To begin with, there are the stones in situ noted by Stevens
(Plan 19, 1-2). They face west and are built precisely on the brow of the rock.
They follow a slightly curving line for a length of some 5 m. from SE to NW.
Their size and arrangement show that they are part of an initial outer layer,
fairly low, made in order to provide a relatively horizontal surface on the slop-
ing rock that would be able to carry the enormous stones of the fortification.
A row of poros blocks, approximately 1 m. N of these stones, at 3 on Plan 19,
belong to a later foundation, probably Archaic.
Some 2-3 m. E of the row of poros blocks, the rock has been worked to
make a uniform but not exactly flat surface with rounded corners, thus creat-
ing an extensive area further up and at about the same level as the tops of
the poros blocks. In the S part of this characteristically Mycenaean working
of the rock, a small stone remains that closes a minor split in the rock (next
to 4 on Plan 19). As can be seen on the plan, the entire worked surface extends
to the brow of the rock and went through the spot where the poros blocks at
3 were later set. It gives us the line of the Mycenaean wall.
232. Hesperia XV, 1946, pp. 73-75, fig. 2.
123
THE THIRD PHASE
136,93
0 5 10 20
r-tr-t~i---i~=4r--i~=========l=I============~~==II M
Plan 19. Remains and traces of the west fortification wall.
There is a second, similar surface at a higher level just behind and NE of

this first worked surface. It rises step-like up the rock and it is around 1,50
m. wide (Plan 19, 5). Its position is roughly parallel to the first. The purpose
of these levels (Fig. 17) is evident. The rock at this place slopes gently but
continuously. These small artificial flat surfaces were absolutely necessary in
order to set a foundation firm enough to carry the stones of the fortification
wall. At the E end of the worked surface at 5, where the surface of the rock
is cracked and uneven, small stones and mud have been used to extend the
level area. A number of narrow cuttings in the flat surface at 4 were obviously
made for shifting or setting the blocks of the wall.
124
THE THIRD PHASE
Fig. 17. The cuttings on the rock W of the Propylaia, from the S.
Here, then, are traces of the fa~ade of a slightly curving section of the wall.
It can be followed for some 10 m., partly by stones still in situ and partly by
cuttings in the rock. There is also a valuable piece of evidence for the wall's
original width. The full width of the worked surfaces was certainly intended
to be used and the direction in which they face shows clearly that they were
made for the outer face of the wall. The rock to the E and above the second
cutting is itself fairly level. The width of the wall will evidently have equalled
the width of the traces as a whole, with the addition of another series of stones
for the inner face. The width of the traces from the W side of the stones in
situ at 1 on Plan 19 to the middle line of the cuttings in the rock at 5 is about
4 m. Accordingly, the width of the wall at this point should be as much as 5 m.,
hardly surprising since it is next to the main entrance of the fortification.
125
THE THIRD PHASE
' 0
I
Plan 20. The west fortification wall.
Wherever preserved, the Mycenaean wall adheres closely to the brow of

the rock. The significance of this is that it indicates the course of the wall to
both N and S. On the basis of this, the line of the wall has been drawn toward
the SE corner of the Propylaia (Plan 20, 1). The wall traces themselves are
shown at 2 on the plan. The line of the wall continues N of them, traversing
the location of the later poros blocks, running parallel to the W wall of the
Pinakotheke foundation. It is not precisely on the same line as the Classical
wall, since that wall, built in a straight line, has at many points gone beyond
the brow of the rock, thus hiding the course of the Mycenaean wall, which lies
further in. 233
233. Leake, Topography p. 313, observ- gikon. It was not a remainder of the Pelar-
ing that the wall of the Pinakotheke founda- gikon but, as we shall see, it is built along
tion is at an angle to the Pinakotheke, in- the same line.
terpreted it as a remainder of the Pelar-
126
THE THIRD PHASE
These conclusions help in the interpretation of the next remnants of the

Mycenaean wall, preserved within the Pinakotheke (Plan 20, 3). Here the exca-
vation of 1889 revealed large stones lying in disorder directly on the rock. 234
Preserved to the E of this pile of stones was a clean Mycenaean fill of at least
1 m. in height. Likewise founded on the rock was a corner wall, its lower part
constructed of small stones, its top of rough slabs. It is the remnant of a build-
ing belonging to the Mycenaean Acropolis. 235 Because of their size and
because they are directly on the rock, the fallen stones can only have come
from the Mycenaean fortification wall, the only structure that could have
retained such a high fill on the inner side. 236
These massive stones, lying as Mardonios' Persians left them when they
destroyed the wall, no longer preserve the line of the inner face. There is how-
ever the corner wall, the W leg of which was evidently built next to and par-
allel to the fortification wall. This is indicative. Its construction and the preser-
vation of the rough slabs at the top show that it had not been built directly
against the fortification wall. Yet the fortification wall cannot have been far
from the place where the massive stones were found. The inner face of the
wall will thus have been parallel to the house wall, following a course 0,10-
0,20 m. W of it toward the N.
The outer face, as we saw, cannot have been outside the Classical founda-
tion of the Pinakotheke. Since it follows the configuration of the rock, it
can only be somewhat further in. The distance between the inner face, as
determined from the house wall, and the exterior face, as determined by the
rock itself, is almost precisely 5 m. This gives us the width of the wall in this
place and supports our conclusion about its width further south.
234. Kavvadias-Kawerau p. 59. p. 73). More recently, Bundgaard (Mnesi-

235. Ibid. pp. 41-43, 59, pl. B, and Kav- kles pp. 47-48), came up with the peculiar
vadias, Deltion 1889, pp. 105-106. view that the stones were the remains of a
236. These observations were first in- terrace wall. This idea he based on the fact
terpreted and analysed by Heberdey (Olh that the enormous stones did not form an
1910, pp. 2-3), in a reply to Koster, who interior face (clearly because of the Per-
placed the wall further east. This was ac- sians' destruction of the wall).
cepted also by Stevens (Hesperia XV, 1946,
127
THE THIRD PHASE
There is a triangular space just beyond the NW corner of the Pinakotheke,

at the point where the end of the Classical N fortification wall meets it (Plan
20, 4). Here the rock is not only sheer, uneven and full of cracks, but there
are no traces or signs of the rock having been worked, as might have been
expected considering its formation. The Mycenaean wall therefore could not
have continued along the same line without changing direction. It did not go
across the space at 4. It took the only course possible, along the line followed
later by the Classical wall. Just after 3 it turns slightly E, runs beneath the
foundation of the Pinakotheke, then turns again and runs N for some 5-6 m.
as far as about point S on Plan 20. After 5, the rock formation compels the
wall to make an approximately right angled turn toward the E. Here ends the
W wall and the N wall begins.
THE NORTH SIDE UP TO THE NORTH FOUNTAIN
After turning eastward at the NE corner of the Pinakotheke, the fortifica-

tion wall follows the brow of the rock with its outer face to the N. It has left
no visible trace of the first meters of its course. Yet the brow of the rock con-
tinues to provide evidence and ample information is to be found inside the
Classical wall as well. Preserved are a number of constructions, dating to his-
torical times before the Persian destruction, that had been more or less
adapted to the still existing Cyclopean wall. They provide good evidence about
its position. The assignment of these constructions is easy enough because a
clear terminus ante quern is provided by the building referred to by Kawerau
as the northwest building, numbered 2 on his plan. 237 Details of its construc-
tion show that it was built at the same time as the Classical wall, to which it
is joined. Its foundations lie on top of all the earlier constructions in the area.
237. Kavvadias-Kawerau p. 67, pl. B. Its vens, in his reconstruction of the Acropolis,
exact use is unknown. The various inter- characterises it as a service building (Hes-
pretations are given by Judeich, Top. p. 246 peria, Suppl. III, fig. 1).
and n. 1, without himself identifying it. Ste-
128
THE THIRD PHASE
.:-_·:::::: ~
,-
2 ...;;:_:,..':::
- --
:,• -- --71~" ~--:..~-,
I I
I 1
~
~
-
I
I L r
;Ml
I~ I I
L , rj I
/ I 1 : t 1 '' L __ ...J L • • • ....J L_.J
,'
/ / I
I
I I
I I I 1 I
Ll_____ ~
I/
- - - - - -- .: - -f.1_ - _
-- - - - - _-_-_-: ::.-'_; I I
0 5 10 20 30 40
Plan 21. The north fortification wall from the Pinakotheke to the NW descent.
These are as follows:

I. The large, Archaic reservoir (Plan 21, 1), probably built at the time of Pei-
sistratos, 238 and restored on the drawing according to Kawerau's observa-
tions.239
II. The Archaic drain (Plan 21, 2) near the NW corner of the reservoir was
made after the reservoir, but was in use before the construction of the
Classical fortification wall. 240
III. The oblique wall, 3 on Plan 21, precedes the northwest building, but per-
haps not by much. 241
238. Kylon's supporters, besieged in the besieged. The reservoir will thus have been
Acropolis, were obliged to surrender because built between those two dates, probably by
there was no water (Thuc. I 126); therefore Peisistratos himself, who in any case had
the reservoir was not in existence in 632 established himself in the Acropolis.
B.C. Quite otherwise, the Peisitratids were 239. Kavvadias-Kawerau p. 65.
adequately supplied with food and water 240. Ibid. p. 67.
(Herod. V 64 ), when they were in turn 241. Ibid. p. 63.
129
THE THIRD PHASE
IV. Wall 4a on Plan 21, which 3 overlies, and 4~, which because of construc-
tion and plan clearly belongs with 4a.
To these should be added also wall 5, Plan 21, on the N part of which
rested the Classical wall.
As Kawerau rightly observes, 242 the way in which the reservoir (1) is built,
presupposes a deep fill around it to hold the rather thin walls, which other-
wise could not have withstood the water-pressure. A fill of this sort could exist
only if retained in turn by a strong wall along the brow of the rock, in this
case the Cyclopean wall. The wall therefore ran N of the reservoir. A more
precise indication of the line it followed is given by the drain (Plan 21, 2). The
course of the drain, indeed, is determined neither by the reservoir nor by the
Classical wall, which is of course later. Apart from the last section that pro-
jects outside the wall, it could not possibly have been built underneath the
foundations of the Late Helladic wall. That is clear from its careful work-
manship, which implies ample room, and from the fact that it is covered by
relatively thin slabs. These would not have been needed and in any case could
hardly have supported the weight of the wall on top of it. The course of the
drain runs parallel to the brow of the rock. Between the N side of the drain
and the edge of the brow there is a level space of somewhat more than 4 m.,
a space corresponding, that is, to the width of the wall. All these features taken
together define the course of the initial part of the N wall as it is shown on
Plan 21. After a turn toward the E, the outer line of the wall follows the brow
of the rock, taking in also the place where a buttress was set later on. Its inter-
ior face is parallel to the drain as far as the W chamber of the Archaic reser-
voir. A small section of the wall (Plan 21, 6) that is preserved outside the Clas-
sical wall shows the continuation of its course.
Just E of the cave of Apollo (Plan 21, 8) and above the cave of Pan (Plan
21, 7), the brow of the rock juts sharply out to the N. The large buttress of
the N wall (Plan 22, 1), built during Mediaeval times, 243 is founded on the E
242. Kavvadias-Kawerau p. 65. is smaller and not entirely symmetrical with

243. See ibid. pl. B, on which only the the base. Plans 21 and 22 show the base
top of the buttress is marked. Because it with its actual measurements.
becomes smaller in size as it rises, its area
130
THE THIRD PHASE
edge of this projection. Some stones of the N Cyclopean wall are still pre-
served in situ on the rest of this projecting surface, to the E of the buttress.
They were first recorded by Kavvadias on his plan of the area of the caves.
Koster too had observed them. He took them for the beginning of the
Pelargikon, which would have continued in the direction of the Areopagus.
His interpretation, quite rightly, was not accepted. 244 Such an arrangement
would have meant that for 30 m. this leg of the Pelargikon would have had
to descend to a level some 15 m. lower, and it would have passed right through
the cave of Pan. This in itself rules out Koster's interpretation.
In fact, this is a section of a massive wall of Cyclopean construction. The
wall is preserved for a stretch of 3 m. along the brow of the rock facing N and
running from the NE corner of the buttress at an angle to the Classical wall.
It comprises a series of enormous stones built in 2-3 courses with a relatively
even vertical outer surface (Plan 22, 2-2; Figs 18 and 19). There is no doubt
whatsoever that the stones belong to the Cyclopean wall, which faithfully fol-
lows the line of the rock. Left and right of this series of stones, toward the
buttress and toward the Classical wall, the rock has been roughly levelled for
bedding the Mycenaean wall (Plan 22, 3-3). A considerable number of stones
from the interior of the wall are preserved on this graduated cutting of the
rock, which is much rougher on the interior of line 2-2 (Plan 22, 4-4; Fig. 18)
than along the line prepared for the outer face. While some of the stones are
quite large, they are smaller than those of the outer face and they are entirely
unworked. Some lie directly on the rock, others are held in place by smaller
stones. In the joins of most is an argillaceous yellowish mortar that yielded
sherds supporting earlier conclusions about the date of the Mycenaean forti-
fication. 245 Cleaning showed that the stones, except for a few next to the but-
tress, had remained undisturbed, thus providing secure evidence.
Behind the stones and running beneath the foundations of the Classical
wall for a length of somewhat over 3 m., are other stones on which the Clas-
sical wall was built (Plan 22, 5). These stones were, in all probability, not
moved but simply trimmed along the outer side so they would conform with
the smooth face of the wall. In other cases, the Classical poros stones have
244. See supra, n. 127, 128 and 129. 245. See Appendix II, group 5.
131
TIIE THIRD PHASE
0 2 3 4 5 10
Plan 22. Remains of the north fortification wall beside the big Mediaeval buttress.
been trimmed roughly so as to conform to the uneven shape of the Cyclopean

stones.
Preserved in front of the exact middle of the buttress, at 6 on Plan 22, is
another massive stone, isolated and at an angle to the buttress. It will have
belonged to the Mycenaean wall but it has clearly been moved since small
stones and lime mortar have been used to secure it to the rock. Byzantine
sherds were found next to it on the rock.
Here, then, is not the beginning of the Pelargikon, but part of the Myce-
naean fortification wall. It was built precisely along the brow of the rock, a
line ignored by the straight Classical wall. Since the place was difficult of
access, the Cyclopean wall is preserved as it was left by those who built the
Classical wall.
Let us now return to Plan 21. We have followed the Mycenaean wall to
the point where a recess in the rock forms the cave of Apollo (Plan 21, 8) .
132
THE THIRD PHASE
Fig. 18. The stones of the Cyclopean wall next to the Mediaeval buttress, from
above.
Beyond that point its course to the E is defined by two factors: the orienta-
tion of the rock above the caves of Apollo and Pan and, especially, the pre-
served section of the exterior face of the wall to the E of the buttress (Plan
21, 6). These two factors show that it curved toward the N following the curve
of the rock, the sequel of which brings it to the N leg of wall 4a. To judge by
its construction, this wall, as also 4P, is probably Mycenaean, but from the end
of the period. It is in any case very old. It is thick and made of small unworked
133
THE THIRD PHASE
Fig. 19. Successive courses of the Cyclopean wall beside the Mediaeval buttress,
from the W
stones. 246 It cannot possibly be part of the fortification wall. It follows a dif-
ferent course and, still more significant, it is not founded directly on the rock
but on a fill some 0,90 m. high,247 the formation of which presupposes the
existence of the fortification wall to the N of it. Thus it was built after the
wall, and indeed some time afterwards. In all likelihood the N leg of 4a was
supported against the inner side of the fortification. Since it was built after
the wall, it provides evidence for both the course and the width of the wall.
246. Kavvadias-Kawerau p. 63. 247. Ibid. pl. I, lower left.
134
THE THIRD PHASE
The NE corner of 4a is no more than 3 m. from the brow of the rock. This
means that the wall, which is around 4 m. wide as far as the drain at 2 on Plan
21, gradually narrows to a width of 3 m. at the NE corner of 4a. Indeed the
rock at this place is sheer and because of the caves inaccessible, so that a par-
ticularly heavy fortification was not needed.
The oblique poros wall, 3 on Plan 21, also supports these conclusions about
the inner line and consequently the thickness of the Cyclopean wall. The
oblique wall is later than 4a since it rests on top of it. It is earlier than the
Classical wall since the northwest building was founded over its S leg. Thus it
is virtually certain that it was built before the Cyclopean wall was destroyed
and consequently was adjusted to it. If the line of its short N leg is extended
westward, it runs parallel to the inner line of the Cyclopean wall as we have
reconstructed it.
Walls 4P and 4a, built in the same fashion and with the same arrangement,
clearly belong together, but to what sort of complex is not clear. A hypothet-
ical arrangement is shown on Plan 21, but this is not the only possibility.
A section of the wall is preserved a few metres to the E of complex 4a-4p,
at 5 on Plan 21. There is no question whatsoever about this. It is constructed
of enormous stones, founded directly on the rock248 at the exact place where
the wall had to pass. It is notable, however, that the stones facing S, while
quite large, are neither as large nor as evenly laid as is usual for the wall faces.
The N face has been hidden by the Classical wall. The S is exposed, but it is
uneven and constructed of relatively small stones. The massive stones of the
S face clearly were removed at some time, perhaps during the Turkish dom-
ination, leaving only the interior fill. To restore it to its original appearance,
at least one row of larger stones, 0,50-0,60 m. in width, would have to be added
to both faces, thus bringing the width of the wall to about 5 m. The proxim-
ity of the wall at this point to an entrance, where the NW descent begins,
explains the difference in width between this and the 4 and 3 metres width of
the wall to the W.
In fact, the stones begin to make a turn toward the N just at the point
where the preserved part of wall 5 ends at the E. The Classical wall follows
248. Ibid. pl. I, lower right.
135
THE THIRD PHASE
the same course, a course imposed by the rock itself. Since at this place the
Classical wall runs along the brow of the rock, the outer faces of both walls
can only have coincided. If the Cyclopean wall did not follow this course, it
would not concur with either the previous section to the W or the beginning
of the Pelargikon (see below). Thus, for a stretch of at least 5-6 m., the wall
runs N.
In Classical times there was a stairway here, enclosed within the Themis-
toclean wall. It made a right angle turn and led from the beginning of the NW
descent to the top of the rock (Plan 21, 9). There must have been some sim-
ilar arrangement through the Mycenaean fortification.
As it is today, the descent begins at a point S of the place where the rock
projects at 1 on Plan 23. This point (Plan 23, 2) is at an elevation of 137,14
m. The corresponding point inside the fortification wall, both Mycenaean and
Classical, is at a level of 143,40 m. The Classical stairway reached a level of
144,35 m. within the wall. Since no trace survives, the arrangement of the
beginning of the Mycenaean descent, which must have passed through the
wall, can be calculated only from the differences in elevation.
One thing is certain. The steep slope requires a stairway. This rules out
the existence of a canonical gateway, as it would have occupied space needed
for the stairs. In any case this was a secondary approach, no doubt for a lim-
ited amount of traffic, and a canonical gateway would have been unnecessary.
Without positive evidence, a number of solutions may be conjectured for
the plan of this stepped path. Generally speaking there are two possibilities.
1. A straight stairway, cutting transversely through the fortification wall. The
difference in elevation between point 2 on Plan 23 and the corresponding
point inside the wall is 6,26 m. Based on the width of the wall as calcu-
lated, the length of the stairway would have to be 5 m. In this case the dif-
ference could be covered by 31 steps with treads 0,16 m. wide and risers
0,20 m. high, or by 25 steps with treads of 0,20 m. and risers of 0,25 m. 249
2. A stairway formed like a Z, beginning at 2 in Plan 23 and rising to 3, with
a course like that of the Classical stairway. This has a vertical rise of
249. These calculations, as those fol- for various combinations between the two
lowing, are only indicative. There is room extremes.
136
TIIE TIIIRD PHASE
0 5 10 20 30 40
Plan 23. The north fortification wall from the NW descent to the House of the
Arrephoroi.
7,21 m. for a stairway 9,80 m. long. It implies a stairway of 36 steps with

treads 0,27 m. wide and risers 0,20 m. high.
I believe this second solution to be the most probable and have shown it
on Plan 23. It has various advantages. It makes an easier ascent than all the
others, its defense is simpler and it agrees more with the arrangement of Clas-
sical times, which, in the natural course of things, would have followed the
earlier plan. Similar solutions are not unknown in Mycenaean fortifications. 250
After this short tum to the N, the wall again runs E, continuing along the
brow of the rock. The inner side has left no trace. Part of the exterior line,
250. See the S entrance of the second period of the Tiryns acropolis.
137
THE THIRD PHASE
Fig. 20. Stones of the Cyclopean wall on the edge of the rock outside the Classi-
cal fortification wall, W of the North Fountain.
however, is preserved on the rock and up to now it has gone unnoticed. A

limestone block projects from beneath the poros blocks of the Classical wall
at the point where that wall turns E, beneath the corner of its foundation,
which is now destroyed (Plan 23, 4). This was one of the stones used in the
building of the Cyclopean wall. Directly E are two other similar stones on a
line that can be followed further still. At this point the brow of the rock makes
a wide curve out to the N. The Classical wall, built as it is in a straight line,
ignores it.
Preserved at the same level on the narrow space that remained outside the
Classical wall, is a series of large stones founded on the rock. They follow the
line of the rock (Fig. 20; Plan 24) until the curve straightens out at the base
of the Classical wall. Most of these stones are large, but some small stones fill
138
THE THIRD PHASE
0 5 10 20
M
Plan 24. Remains of the north fortification wall next to the House of the
Arrephoroi.
in the spaces. Only the lowest row is preserved. All the space available for a
wall foundation is covered. The stones do not appear to be trimmed, but they
are arranged so that their relatively even sides face outwards. The rock where
they are set was cut very little, only to the extent absolutely necessary for a
firm foundation. Directly behind these stones, practically flush with the Clas-
sical foundation, the rock rises abruptly so that its highest level coincides
approximately with the top of the series of stones. No fill is preserved between
the two, as was shown in cleaning the area, and nothing remained from the
second course.
This series of stones is part of the outer face of the Cyclopean wall, which
continues for 11,50 m. toward the E (Plan 23, 4-5). From point 5 the brow of
the rock runs in toward the Classical wall so that at 6, where the Classical wall
is adjacent to the east wall of the House of the Arrephoroi, the Mycenaean
line no longer projects beyond the Classical wall. Just E of 6, another curve
begins which brings it out somewhat beyond the Themistoclean wall.
Thus the exterior line of the Cyclopean fortification wall is clear between
4 and 6, first from the stones in situ, and second from the brow of the rock.
The interior line has of course to be parallel to the exterior, so that its course
is no problem. The problem lies, rather, with the width of the wall. The solu-
139
THE THIRD PHASE
tion is provided by the beginning of the descent to the N Fountain, which is

preserved beside the NW corner of the House of the Arrephoroi.
Shown in broken lines at 7 on Plan 23, is the still preserved Mediaeval
descent to the interior of the fissure between the rock and the big piece that
broke off from it, as revealed by Kavvadias. 251 Ten wide steps lead from the
level of the Mediaeval (and modern) fill to a wide landing cut into the sur-
face of the rock. This was the beginning of the Mycenaean descent to the
Fountain (Plan 23, 8) as well. The poros foundations of the House of the
Arrephoroi rest in part on top of this. Here, at the beginning of th.e descent,
the rim of the main mass of rock has been cut back to form a ste·p facing N.
By analogy at least with the Perseia spring at Mycenae, this first step of the
descent should already be beneath the wall. In fact it is, for the following sim-
ple reason: the wall and, indeed, its inner face, cannot possibly have been
founded over the fill in the fissure. It will have rested on the rock, either on
the main mass or on the part of the rock that pulled away. We can exclude
the part that pulled away because, as can be seen from the plan, the wall would
have been only 2 m. wide if not less. The wall must therefore have spanned
the fissure and been founded on the rock itself. In order to be securely
founded, however, a width of at least two blocks, about 1 m., would have had
to rest on the edge of the rock. This brings it precisely to the point where it
covers the beginning of the descent, as we suggested, and the wall in this place
will have been some 4 m. wide. The S face of the wall has been drawn on the
plan in agreement with this conclusion.
THE NORTH FOUNTAIN HOUSE
The mass of rock that pulled away from the main body to rest at an angle
against it, left a deep fissure, varying in width from 1 to 3 metres. Water from
the rock collected in the recess of the fissure, and since it had no outlet, most
of it was collected there. Realising this, the inhabitants of the Acropolis dug
deeply between the two walls of the rock and built a stairway in the fissure
251. Kavvadias, Ephemeris 1897, p. 28.
140
THE THIRD PHASE
leading down to the level at which the water settled (Fig. 21). When the foun-
tain house went out of use, only the upper part of the descent continued to
be used. In Classical times it led to the cave of Aglauros and the Arrephoroi
went down this way to the sanctuary of Eros and Aphrodite. 252 Later on, and
indeed down to the time of the War of Independence, it appears to have
served as a concealed sally port from the Acropolis fort. Kavvadias cleared
the first nine steps of this later descent. 253 Not realising that they continued
on down, he proceeded no further. The fountain itself was discovered by
Broneer during the systematic excavation to which we are indebted for the
information on this subject. 254
The Mycenaean excavation within the fissure descended some 34,50 m.
from the rim of the rock to a level of 109,05 m. In winter the water level
reaches 112,96 m., four metres above the level of Klepsydra and five metres
above that of the Asklepieion spring.255
After the first step, which has already been discussed, the Mycenaean
descent to the Fountain House begins. It is divided into 8 flights running alter-
nately E-W and W-E. The first flight comprised 25 wooden steps running from
E to W, anchored in hollows cut in the S wall of the fissure. It ended at a
landing from which the second flight of the descent continued in the opposite
direction by means of 40 steps of generally similar construction. This con-
tinued to a level of 130,40 m., corresponding roughly to the level of the cave of
Aglauros. These first two flights are the ones that continued to be used dur-
ing historical times. No trace remains of the last steps of the second and first
steps of the third flights. Perhaps they rested on the fill in the fissure rather
than on the rock itself. Be that as it may, between the last preserved steps of
the second flight and the first preserved steps of the third there is a gap of
some 4 metres. The next part of the third flight, running westward, has 9 steps
ending at a landing. From this the next, the fourth flight begins. The 3 steps
of this flight consist of slabs of grayish-blue marble built with yellowish clay
on a support of small stones and plenty of mud plaster, strengthened and held
252. Broneer, Hesperia I, 1932, p. 52; AJA 1938, pp. 445-450, Hesperia VIII, 1939,
VIII, 1939, pp. 322 and 428. pp. 317-433, pls XI-XIII.
253. See supra n. 251. 255. Broneer, Hesperia VIII, 1939, p.
254. Hesperia VII, 1938, pp. 168-170, 429, n. 194.
141
0
WATt.ALtVtL Ave; . 20 . 'Ql 8 .

Fig. 21. The North
Fountain within the
natural cleft of the
Acropolis rock. Sec-
tion looking E (Bro-
neer, Hesperia VIII,
1939, pl. XII).
THE THIRD PHASE
Fig. 22. The descent to the North Fountain, model (Broneer, Hesperia VIII, 1939,
fig. 17).
in place by wooden beams framing it along the sides (Fig. 22). The fifth flight,
built like the two preceding ones, leads to a landing about 1,50 m. wide,
formed of two stone slabs on a wooden support that spanned the fissure so
that from here on the stairway continues down anchored in the N wall of the
rock. The landing itself comprises the sixth flight and it is followed by the 5
steps of the seventh, running from W to E. The 4 steps of the eighth and last
flight follow, in the opposite direction. These last three flights have stone
steps, firmly set into appropriate cuttings in the wall of the rock, which at this
depth is no longer hard limestone but a much softer schist. The last step of
the eighth flight is at an elevation of 118,16 m. Beneath this a cylindrical well
had been dug with a diameter of about 2 m., its walls held by wooden sup-
ports. The well ended in a hive-shaped cistern measuring about 4 m. in dia-
meter, with a concavity in the centre of the floor (Fig. 21) to serve as a settling
basin. The water was drawn by lowering a container secured by a rope. There
143
THE THIRD PHASE
was no need to descend any further than the landing of the fifth flight, as
water could easily be drawn from here.
Visitors to the fountain house will of course have used artificial lighting.
Kylikes were found with interiors blackened by carbon from the burning wick.
One of these home-made lamps lay on the last step of the fifth flight. 256
The fountain house was constructed during the final years of the LH IIIB
period and was used for only a short time. After this it seems to have fallen
into disrepair and, the danger that brought about its construction evidently
having passed, it was never rebuilt. On the basis of the sherds retrieved from
the substructure of the steps, which date their construction, and from sherds
that fell into the fissure during its use, together with what was discarded there
after its abandonment at the beginning of LH IIIC times, when the fountain
house served as a dump, Broneer concluded that its period of use did not
exceed 25 years. The material collected supports his view. 257 As we shall see,
the construction of both fountain house and fortification indeed go back to
about that time, and they appear to have been dictated by the same need.
THE NORTH SIDE FROM THE NORTH FOUNTAIN TO THE NE

ASCENT
The formation of the brow of the rock provides the only existing evidence
for the course of the wall E of the beginning of the descent to the Nort Foun-
tain and for a stretch of some 30 m. Between the E wall of the House of the
Arrephoroi and a point about on a line with the NW comer of the N porch
of the Erechtheion (Plan 25, 1-2), the wall makes a curve to the N that is
almost imperceptible in relation to the Classical wall. Yet the curve is suffi-
cient to bring the outer face of the Cyclopean wall outside the corresponding
line of the Themistoclean wall, but close to it and roughly parallel. Just E of
point 2 the rock juts sharply out to the N. In accordance with what we have
256. Broneer, Hesperia VIII, 1939, p. Daniel, AJA 1940, pp. 558-559, Furumark,
377, fig. 58b. OpArch 1944, pp. 197, 231, Mountjoy pp.
257. Ibid. pp. 335, 346, 349, 395, 417, 43-44.
423, AJA 1948, p. 112, Antiquity 1956, p. 13,
144
THE THIRD PHASE
0 5 10 20 30 40 50
Plan 25. The north fortification wall from the House of the Arrephoroi to the area
NE of the Erechtheion.
already observed, the Cyclopean wall will have followed the same line. In fact,
three stones of the wall are still in situ at the W end of this protuberance. This
is evident from the way they lie on the rock and from the fact that two are
adjacent, their outer faces forming together a line corresponding to the edge
of the rock. The stones are not especially large and the outer surface is not a
regular face. Thus they are part of the fill rather than the outer side of the
wall and the actual face must have been a little to the N of them.
The curving projection of the rock continues to point 3 on Plan 25, where
the corner of the Classical wall rests at the very edge. Here the outer faces of
the two walls coincide. On the inner side of the corner there is evidence for
the inner line of the Mycenaean wall. Preserved at point 4 were two of the
little Middle Helladic cist graves, 258 the slabs of which were set directly on the
rock. These graves could not possibly have been preserved if the wall, like-
wise founded on the rock, had been built over them. Since the graves sur-
vived, even with their cover slabs intact and their walls upright, the wall will
258. See supra, p. 54 and Plan 2, 7.
145
THE THIRD PHASE
have run between them and the brow of the rock. We can thus determine not
only the line of the S face but also the width of the wall. The distance between
the easternmost grave and the brow of the rock at 3 allows the wall at most
a width of 3 m. Accordingly the S face has been drawn on Plan 25 to show
the width decreasing from 4 m. at point 1 to 3 m. at point 4. After 3, the rock
turns slightly toward the SE, forming two sheer projections that drop off
steeply, full of cracks and without any sign whatsoever of cutting or other
working. The wall cannot possibly have stood on this and it must have fol-
lowed more or less the line of the Themistoclean wall.
At 5 on Plan 25 we again pick up its traces. Here, the excavators of the
Acropolis noted three colossal stones that formed a sort of corner. 259 They
attributed them to the palace, together with the remnants of a wall of the
Turkish period to the N of them.260 Stones of such great size (2x 1,30 m.,
1,20x2 m.), however, cannot have come from any construction other than the
fortification wall. Their size and their weight alone saved them from being
flung down when the wall was destroyed, but we may well ask whether all
three are in their initial location.
For the following reasons this may be ruled out.
First: All three together occupy a space 4 m. wide, a space that could equal
the entire width of the wall. It is most unlikely that the builders of the wall
would sacrifice so large a stone, difficult to find and difficult to move, just to
include it as fill. In every fortification wall preserved in all the Mycenaean
citadels, such large stones are reserved for the wall faces, while the interior is
filled with smaller stones.
Second: Even if we accept this to be so, the wall, which was 4 m. wide
6-7 m. to the west, narrowing to 3 m. afterwards, would at this point (5) have
been inexplicably and unnecessarily wide at the cost of valuable building ma-
terial.
Third, and most important: It is quite clear from Kawerau's plan that the
westernmost and largest stone has fallen in such a way that the edge toward
the east rests slightly on the middle stone. It lies fallen at an angle to the
ground, in a position inconsistent with a place in the face of the wall.
259. Kavvadias-Kawerau p. 85, pl. r , 35. wall 27.

260. See supra, p. 60, Plan 3, next to
146
THE THIRD PHASE
This stone at least is clearly not in its original place. It was in all likelihood
placed lengthwise on top of the other two stones, which, given their meas-
urements, it will have covered. When the wall was torn down, the stone
appears to have been shifted and overturned, but it was evidently too difficult
to move it any further. The other two stones, however, appear to be set in
regular fashion, their W faces on one and the same line, coinciding with the
line that would have been the natural course of the wall. Thus here too the
inner face of the wall is preserved. With the line of the outer face defined by
the rock, the approximate width here is 3 m., just as in the previous section.
After 5 on Plan 25, the wall again runs eastward. At this point we encounter
the section of the Classical wall that is founded on half finished and broken
marble and poros column drums. The drums here are set on the brow of the
rock so that the N face of the Cyclopean wall must have coincided at least
with the first 7 (Plan 25, x, A., µ , v, ;, o, :1t ). At drum :Jt there is evidence for
the line of its S face.
Here (Plan 25, 6) the N end of the W wall of the LH I building excavated
by Holland was found. 261 This wall, poorly built, ends at the N somewhat over
3 m. from the outer face of the wall opposite it. Clearly a construction as in-
secure as this one, already abandoned long before the fortification wall was
built, will have been destroyed completely when the foundations were set. This
is precisely what happened, since the truncated end of the wall cannot be
explained in any other way. In any case, the fortification cannot have stood
without obliterating this wall. Therefore the wall ran just N of the preserved
part of wall 6, continuing to be 3 m. wide.
Some 8-9 m. to the E, at point 7 on Plan 25, is the W supporting wall of
terrace I with the beginning of its continuation northwards. For the Cyclopean
wall to continue eastwards, it had to be adapted to the N supporting wall of
the terrace. It evidently ran on top of it, thus gaining around 1 m. in width.
In any case, E of 7 at point 8, the inner face of the wall is preserved, and it
coincides with the terrace wall. 262 E of drum x the rock again swings out to
the N, away from the line of the Classical wall. The Cyclopean wall, to be
sure, follows the brow. This is not simply hypothetical. If the line of the Clas-
261. See supra, pp. 73ff, Plan 5. 262. See supra, p. 76, Plan 7.
147
THE THIRD PHASE
sical wall had coincided with that of the Cyclopean wall, the Cyclopean wall
would have been less than 3 m. wide at point 8 (where its inner face is pre-
served). This is highly unlikely given its massive construction and its proxim-
ity here to an especially vulnerable point, the NE ascent. Thus, here as else-
where, we can see that the artisans of the time founded their wall as close as
possible to the brow of the rock. The distance from the rim of the brow to
the preserved S face of the wall at 8 gives a width of 3,80-4 m. This width was
attainable because the wall was founded on the earlier terrace and it was ne-
cessary because of the proximity of the NE ascent.
In order to reach this width uniformly, without comers or additions on the
inner side - both unlikely and superfluous - it was necessary to start widen-
ing the wall at 6 on Plan 25, continuing gradually eastward to the NW corner
of terrace I. This is shown by the fact that the E wall of the LH I building, as
preserved, is shorter than the W wall. It explains also the somewhat uneven
formation of the inner foundation of the fortification wall at 8. The lower
courses, with uneven faces, belong to the old terrace wall; the upper courses
are part of the regular structure of the fortification wall.
Yet another remnant of this time is preserved in the same area: the wall
that was built on top of the S wall of the LH I building (9 on Plan 25). It is
only partially preserved and runs in a straight line from the SW corner of the
earlier building to the W supporting wall of terrace I, on which it rests.
Approximately 0,55 m. in width, it is narrower than the earlier wall on which
it is built. In places it is preserved to a height of 0,20 m. above it. At the point
where it abuts the terrace wall, it is carelessly built with small stones. Else-
where it is carefully constructed with both faces forming even surfaces. 263
Whether it continued further W than the wall of the earlier building or fol-
lowed its turn to the N, is unknown.
The purpose of the wall at 9 is not entirely clear. Its construction created
a long narrow space between the fortification wall and the corner of the ter-
race. Its narrow plan and generally poor construction, especially towards the
E end, means that it cannot have risen as high as the top of the fortification.
Thus it cannot have been a staircase. For these reasons and also because the
263. Holland, AJA 1924, pp. 151-153, 156-157.
148
THE THIRD PHASE
0 5 10 20
SHtEH:t:EH~====~I~~~~~I M
Plan 26. The north fortification wall in the area of the NE ascent.
inner face is even, it cannot have been a supporting wall for a stairway to the
top of the wall or even to the top of the terrace. It will most likely have been
the S side of a roofed space that served as a guard-house or storeroom.
At the point where the inner line of the wall coincides with the earlier ter-
race wall (Plans 25, 8, and 26, 1), at column drums 1-~ (Plans 25 and 26), the
brow of the rock curves out so that the edge protrudes some 1,60 m. N of
drum ~' 2,30 m. beyond drum b and 2 m. beyond y. The curve forms the bound-
ary of the space available for the foundations of the Mycenaean wall and fixes,
as we saw, the width of the wall at around 4 m. The rock swings abruptly S
at the point where drum a is incorporated. It drops off precipitously with a
considerable difference in altitude, breaking up into a series of smaller rock
masses (Plan 26, 2). As a result, this part of the wall has to follow the line of
the E face of terrace I. Yet it cannot possibly have been restricted to the width
of the terrace wall, which would have been quite insufficient; and while the
149
THE THIRD PHASE
height of the terrace wall cannot be determined precisely, it will not have been
that required by the wall. The wall therefore faced the same way as did the
terrace wall, but it was higher and wider. The extra width will have been
inwards since the formation of the rock rules out any extension outwards. In
other words, the wall stood on the E side of the terrace, which it covered.
How wide was the wall?
Following the line of the terrace wall, we come to the landing of the NE
ascent. Here the terrace wall makes a right-angled turn toward the W (Plan
27, 1). Opposite and parallel to it, terrace wall II runs in the same direction
(Plan 27, 2), so that the ascent continues westward between these two walls.
When the fortification wall was built, this pathway, as we saw, was blocked by
three cross-walls, 3, 4 and 5 on Plan 27. The three walls are parallel to each
other. Wall 3 is ea. 2 m. wide, runs from one terrace to the other and com-
pletely blocks the pathway. It is preserved today to a height of around 1,80 m
above the surface of the rock and the top as preserved is uneven. It cannot
have been the threshold of a gate or other such opening. Rather it was built
expressly to block the pathway, and it formed part of the exterior face of the
fortification (see Fig. 7). After a space of about 1 m. this wall is followed by
wall 4, which is about 1-1,20 m. wide. It stops a little before reaching terrace
II, but the end is irregular and the addition of another stone would have
brought it to the terrace. Another metre to the W is the last cross-wall, 5. To
the south it ends up with a large elongated stone, set in the manner of a door
jamb across the width of the wall, 1,40 m. from the N face of terrace II. Here
then is a series of three parallel walls separated by two long narrow spaces,
the westernmost with an opening 1,40 m. wide that leads in from the W (that
is, from inside the fortification). Each 5 m. long and 1 m. wide, these spaces
are obviously not rooms. The easternmost in any case would have been blind
and inaccessible. They can only have formed a stairwell, enclosed in the wall,
which evidently led from the level of the rock to the top of the fortification.
The entire system is, in fact, very similar to the stairwell between the Lion
Gate and the Granary at Mycenae. 264 In my opinion there can be no doubt
as to its use.
264. See Wace, BSA 25, 1921-1923, pp. 17 f., pl. I, n° 8, and Mycenae p. 54, fig. 3, n° 8.
150
THE THIRD PHASE
0 5 10
~H~r---l~=lr---!==l~~l==:==:==:==:~IM
Plan 27. The a"angement of the fortification wall at the point where the NE
ascent ends.
There is no evidence for the form and arrangement of the steps. They will
in all probability have been wooden, with the possible exception of the first
few or only the first, which lay on the damp ground. In fact, the space between
walls 5 and 4, on a line with the end of 5, is taken up by one large stone and
three smaller ones. Kawerau's plan, 265 unfortunately, does not show whether
265. Kavadias-Kawerau pl. f.
151
THE THIRD PHASE
these stones are slabs and in situ, in which case they will have belonged to the
first step. Nor are they mentioned in the text, so they could just as well be
stones fallen from one of the walls.
Be that as it may, with no other evidence, we may suppose that there was
a wooden stairway here. It will have begun on a line with the S end of wall 5,
risen toward the N, then turned in the opposite direction between walls 4 and
3. There it will either have ended or it will have turned again toward the N
over the lowest flight of stairs to end on top of the wall. This second arrange-
ment appears more likely, not only because it brings the end of the stairway
closer to the inner face of the wall where it would be safer for defense, but
also because the addition of a third flight conforms better with the probable
height of the wall. In fact, if we calculate the measurements of the steps to
conform as far as possible with known Mycenaean examples, with treads 0,30
m. in depth and risers 0,20 m. high, a stairway of two flights would go up about
5,40 m., whereas the second arrangement with three flights would rise 8,60 m.
Given a level at the beginning of the stairway of 148,40 m., the top of the wall
(based on three flights) would be at 157 m. At the terminus of the ascent, the
level is 147,57 m. At its outer face the wall must have been at least 9,43 m.
high, which seems likely enough. These calculations are necessarily arbitrary,
but they cannot be very wide off the mark.
There is no doubt that there was a stairway in this place framed by thick
walls. The construction of the stairwell shows that it must have ended within
the wall. The terminus, as I believe I have demonstrated, must have been at
the N end of the space, between walls 4 and 5. For this reason the inner line
of the wall will have coincided with the W face of wall 5 and have turned S
just after the E end of line 1. The wall will thus have varied in thickness from
4,50 to 6 m., depending on the outer face.
Another factor must be noted here. When the wall was destroyed, every
trace of it disappeared and only the terrace walls, the walls of the stairwell
and the stones of the inner face (1 on Plan 26) remained. That is, the walls
founded on the rock remained - and these only fragmentarily. The destruc-
tion on this part of the Acropolis was the work of the Persians alone. This is
evident because when the Themistoclean wall was built it was given an entirely
new line toward the E of 2 and the area was filled in to a much higher level
so that the ruins were buried to a great depth. It is difficult to believe that
the Athenians would have gone to all the trouble of methodically removing
152
THE THIRD PHASE
whatever the Persians had left in an area that was to be covered over and
where remnants of the Mycenaean fortification would not have hampered
them at all since their new wall was being built elsewhere. So it happened that
when the area was covered over at the time when the Classical wall was built,
it remained in exactly the same condition as when Mardonios and his cohorts
left it in ruins.
The outer fortification walls together with the inner walls 4 and 5 were left
in place by Mardonios and his demolition gangs. All the others they destroyed.
How did they get rid of ~he remainders of the interior walls? Surely not by
lifting them over the remaining outer walls in order to throw them down the
slopes. In this case it would have been far simpler to tear down the outer side
first and then dispatch the rest below; in which case, of course, 4 and 5 would
not have been left.
Whatever walls remained, even as low ruins of a few structures, have, as
we noted, a common characteristic: they are founded on the rock. The situa-
tion can easily be explained by the fact that apart from the outer side, which
stood on the earlier supporting walls, the interior of the wall was built on the
hard packed floor of the terraces rather than on the rock. The Persians
destroyed all the upper part of the wall. When they reached the earth fill of
the terraces, they left it, with part of the former terrace wall and the lower
part of the fortification wall itself. There was no reason to complete the
destruction of the terrace walls since they were of no value to the fortifica-
tion. Thus it is clear that the entire inner face of the wall was founded on the
terraces. This is why at the very spot on the Acropolis where you might well
have expected to find traces, there are none at all.
From the S end of 5 the inner line of the fortification wall must have
continued S to about the N face of terrace III. There it will have turned E
(Plan 26, 6). Traces of the outer N face of the wall remain at the point of
this turn.
At right angles to the E supporting wall of terrace II, a row of stones (Plan
28, 1) runs in a continuous line from E to W. Its purpose was evidently to
block off the opening between terraces II and IV, so that the fortification
would be continuous. For approximately the first 3 m. toward the W, the line
comprises enormous stones, of the sort used for the face of the wall. After
the last one of these gigantic stones at the E (Plan 28, la), the line juts out,
then continues eastward with much smaller stones as far as the NW corner of
153
THE THIRD PHASE
Plan 28. Trace of the north fortification wall over the end of the NE ascent.
terrace IV. 266 Considerably lower down because of the slope of the rock, at 2
on Plan 28, just N of stone la, there is a crack that has been blocked by two
small stones not recorded by Kawerau. About 4 m. further east on the same
line, four more stones in succession (Plan 28, 3) fill in a similar break in the
surface of the rock. 267 Next to these, further east and somewhat higher, begin
the first stones of the N face of the wall preserved in situ, which had been
built against the N supporting wall of terrace IV.
North of the small stones E of la, between points 2, 3 and 4 on Plan 28,
the rock falls off abruptly and it has been cut back carefully to form a
266. These small stones, recorded by by error of the lithographer, they were not
Kawerau on his pl. f , no longer exist. coloured, so that while they are still in
267. The stones at 3 on Plan 28 were place they were never noticed.
recorded by Kawerau (pl. f) , but, probably
154
THE THIRD PHASE
sequence of three virtually horizontal levels. 268 This is characteristically Myce-

naean work, with surfaces roughly levelled and with obtuse angles and
rounded corners (Plan 28, 5). The rock was prepared in this way for setting
the foundation stones of the wall, thus adding 2,50-3 m. to the width of the
N supporting wall of terrace IV. The lowest of these three levelled surfaces,
which will have held the enormous stones of the face of the wall, was also the
widest, varying between 1 and 1,50 m. The other two prepared surfaces above
it are considerably narrower, varying in width between 0,50 and 1 m. The cut-
tings (Fig. 23) were absolutely necessary given the steep incline of the rock;
the wall could not have been founded otherwise. Their greatest interest, how-
ever, lies in their W boundary, which gives us the line of the outer face with
great accuracy.
Just N of wall 1, the rock falls off abruptly to the level of the fissure of the
approach. Here, no evidence for any working of the rock has survived. This
begins only at the corner of la, runs somewhat obliquely towards 2, then turns
E following a line to exactly above 3 and then towards 4. Making approxi-
mately a right angle, it covers a line formed by the small stones E of la and
then by the face of terrace IV, leaving uncovered the large stones up to la.
Thus the line of the fa~ade of the wall is clear: at a right angle to the E wall
of terrace II, it runs for about 3 m. to la. Thence it makes a right-angled turn
N toward 2, then runs E in a straight line just above 3 to 4. From here on the
wall is preserved unbroken up to the point where it meets the Classical wall
(Plan 26, 7).
The inner face, which we followed to point 6 on Plan 26, will have turned
along the line of the N supporting wall of terrace III and continued eastward,
cutting across the space between terraces III and IV, about 5 m. S of the outer
face of the wall. A width of 5 m. will have been retained for the same reasons
as held for the previous metres of its course. In this way about 1/3 of it will
have stood over the N end of terrace IV.
At the juncture of the outer face of the Cyclopean wall with the later wall,
at 7 on Plan 26, the last stone is not in the same strict E-W line, but makes
268. Kawerau records the place of this Holland (AJA 1924, pl. VII) copies this
cutting without great precision (pl. r), and without comment.
155
THE THIRD PHASE
Fig. 23. The stepped cuttings in the rock for the foundations of the fortification
wall S of the end of the NE ascent, from the W
a wide angled turn toward the NE. This is explained by the formation of the
brow of the rock here. It curves outward and the curve is reflected to some
extent by the later fortification wall; all the more reason for the Cyclopean
wall to follow it. The angle of the last stone of the outer face at the E demon-
strates this. The brow of the rock, while jutting out some 2,50 m., is not even,
so that it was impossible to place the wall exactly at the edge. The configura-
tion of the rock allows it only 1 m. beyond the Classical wall.
The inner face, following a line parallel to the outer face and retaining the
width of 5 m., brings it precisely NE of the supporting wall of terrace IV, which
it follows. It is most unlikely to have rested on top of the terrace in this place,
since the width would have been increased to an unnecessary 6 m. Moreover,
156
THE THIRD PHASE
as we have seen, the terrace wall is constructed of increasingly smaller stones

because the rock rises here. It would have made a poor foundation for the
wall, which will most probably have been founded just in front of it.
With the wall now built as we have suggested, the NE ascent went out of
use and was abandoned. On the slopes below, a settlement of small and hum-
ble houses came into being, some of which invaded and covered over the
ascending pathway. The houses, which will be discussed below, and finds from
that area gave their discoverer, Broneer, 269 additional material enabling him
to date the wall more closely in connection with the pottery from the North
Fountain.
THE NORTH SIDE FROM THE NE ASCENT TO THE BELVEDERE
After the point where the Cyclopean wall meets the Classical wall above
the NE ascent, the inner face runs parallel to the N supporting wall of ter-
race IV. Some 11-12 m. east of the end of that terrace wall, Kawerau records
a number of large stones (Plan 29, 1).270 He provides no other information
about these and they are not mentioned in the text. They are set right in front
of the terrace, but their size and position show that they do not belong to the
terrace wall. They are part of the inner face of the Cyclopean wall. Their rel-
ative disorder along the S side makes sense since they belong to the inner
foundation of the fortification wall, which was added to the terrace and there-
fore hidden by it. The position of the stones in relation to the brow of the
rock shows that the wall, which had reached a width of 5 m. over the fissure
of the ascent, begins to narrow just after that. At this point it has already been
reduced to 4 m., a more than sufficient width since the cliff is particularly
steep here. As we shall see, the width decreases still further as the wall pro-
ceeds eastwards. The NE corner of terrace IV (Plan 29, 2) provides the next
indication of its course. The fact that it is preserved, as we noted, means that
the wall was not built on top of it, and it therefore has to lie between that
269. Hesperia II, 1933, pp. 352-355; IV, 270. Kavvadias-Kawerau pl. !l..
1935, pp. 111-113, pl. I.
157
THE THIRD PHASE
IV
,,.-
___
1:;4,00
v
0 5 10 20 40 80 __J.,-
M
Plan 29. The north fortification wall from the NE ascent to the Belvedere.
and the brow of the rock. The available space at this point is 3,20 m., and this
will have been the width of the wall.
East of point 2 there is no trace of the wall for a stretch of over 40 m. The
only existing feature is, as usual, the rock, which here too should determine
the wall's course. The Classical wall likewise ran along this line. As will be
evident from the next trace, the Cyclopean wall can hardly have taken any
other course.
At point 3 on Plan 29, the brow of the rock juts out a little beyond the line
of the later fortification wall, forming a small triangular flat space. On top of
this, in situ, are two large stones of unequal size, the faces of which follow the
line of the rock. The lower course of the later wall stands on the inner edges
of the stones, showing that they were there before it was built. Their size and
accommodation to the rock proclaim them remnants of the Cyclopean wall.
They show also that the wall here made a turn toward the SE to follow a line
that was quite unrelated to the further course of the Mediaeval extension of
the fortification, which runs eastward in disregard of the rock. From here on,
traces of the Mycenaean wall are to be sought within the visible walls, be they
Classical or Mediaeval.
The next remnants clearly attest the divergent course of the Mycenaean
wall. The two enormous stones at 3 belong to the outer face of the wall. The
inner face is found again at 4, showing that it had exactly the same orienta-
tion.
158
THE THIRD PHASE
Indeed, at this place 271 Kawerau recorded part of a construction that can
only be the Mycenaean wall. A series of large stones forming a straight and
regular face toward the SW runs for a stretch of about 3 m. They are paral-
lel to the face of the enormous stones at 3. Northeast of them, behind the
inner face, haphazardly placed smaller stones remain in situ from the interior
filling of the wall. These do not extend as far as the outer N face, which in
any case is not preserved. They extend far enough, however, to give a clear
idea of the line of the outer face.
There is yet another piece of evidence. Up to point 3 on Plan 29, the later
wall, despite repeated repairs, was basically the Classical wall. From about point
3 on, the Mediaeval extension takes a new line, ending in the Belvedere tower.
(Plan 29, 6). Where exactly the Classical wall ends and the Mediaeval begins is
not clearly discernible. This is because the Classical wall is badly preserved and
has undergone extensive repairs and rebuilding later on. Yet beneath the top
as visible today, there is a short stretch of a construction that is unquestionably
Classical (Plan 29, 5, hatched lines). It is the continuation of the immediately
preceding stretch, ·which is W of the corner at 3. It runs obliquely to the course
of the later wall, along exactly the same line as that shown by the remains of
the Cyclopean wall at 3 and 4. This earlier series of poros stones marks the
original course of the Classical wall,272 and shows that, as at most places on the
N side, here too the Classical wall follows more or less the same course as did
the Cyclopean, allowing always for its rectilinear construction.
At point 2 on Plan 29, the wall was 3,20 m. wide at most. As far as point
3, it will have continued with this same width, perhaps decreasing to 3 m., for
the sloping rock formation here did not require more massive fortification.
The remains of the wall at 3 and 4, however, show that here the width begins
to increase again. The southeastward extension of the line indicated by the
massive stones at 3, in relation to the line of the inner face at 4, shows that
the width here has already increased to 3,40-3,50 m. Moreover there is good
reason to believe that it continues to increase in width as it runs eastward.
271. Kavvadias-Kawerau pl. !l., 50. 272. See ibid. pl. !l., 50.
159
TIIE THIRD PHASE
From point 3 on, the rock continues to be steep, rough and broken into
fissures, but it no longer forms such a notable brow as it does along the N
side. While it is difficult to climb in places, it is accessible. A very strong for-
tification was therefore essential. The point where the Mediaeval tower was
built, where the Belvedere stands today (Plan 29, 6), was not excavated by
Kavvadias, nor was it ever explored after that. Thus we do not know the exact
configuration of the rock in this place, or whether remains of the Mycenaean
fortification have survived. There is no firm evidence for the course of the
wall from 4 to 7-7 on Plan 29. Its course can only be conjectured on the basis
of the configuration of the rock, where possible, on the probable course of
the Classical wall273 and on the one and only secure fact: its continuation
appears again at 7-7, as we shall see below. This fixes its line, its course to the
S and its width, which at this point reaches some 5,40 m. On the basis of its
appearance at 4 and at 7-7, it is shown as curved on Plan 29. This is arbitrary
to some extent, but it cannot be very far off the mark.
THE EAST SIDE
The E side of the rock, between the Belvedere and the Acropolis Museum,
is quite uneven and slopes considerably from N (elevation 153,93 m.) to S
(148,88 m.). Along the line a-a-a on Plan 30, the rock falls off to the E, form-
ing a sort of brow. At the S end of this area, one of the best preserved and
longest stretches of the Mycenaean wall is visible today (Plan 30, 1). It was
discovered in the great excavation of the Acropolis. 274 This section has large
and carefully set stones at the faces, with the usual smaller stones as filling in
between. It is preserved for a length of around 15 m. It is massive and impres-
sive and it gives a good idea of the spectacle the rock will have presented,
crowned by so imposing a fortification (Fig. 24).
As preserved, it forms a wide angle, with the shorter leg (la) running NW,
the longer part (lP) SE. The width at la is as much as 5 m.; at lP it is nar-
273. See Stevens, Classical Buildings pls 274. Kawadias-Kawerau pp. 39, 95, pl. E.
1-111.
160
THE THIRD PHASE
Fig. 24. The outer face of the SE fortification wall, from the E (phot. DAI n° 80).
rower, never over 3,50-4 m. 275 Its continuation toward the SE was cut by the
Classical wall, which stood on top of it. Cuttings in the rock near the NW cor-
ner of la were made for setting the next stones (Plan 30, 2). Here Kawerau
noted a series of stones in situ, which no doubt will have existed when he exca-
vated but they have been lost since then. No other remains of its course
toward the NW have survived. There are, however, quite a few traces on the
rock. At a number of places where the surface was uneven, projecting pieces
275. On his plan E, Kawerau has the removed along the outer face, so that the
width as less than 2,50 m. This is evidently wall appears to be narrower than it actually
an error, as appears from the repeated is as is evident from its foundation . Kawe-
measurements I made. The reason for this rau evidently measured it across the top,
mistake is evidently that a line of stones giving this width to the entire wall.
preserved at the uppermost level has been
161
THE THIRD PHASE
MUSEUM
0 5 10 20 30 -i..
Plan 30. The east fortification wall.
of the rock were removed. This was coarse work, probably done with a stone
hammer, certainly not with a chisel or other such cutting tool. Small irregu-
lar and practically level surfaces were formed (Plan 30, 3, 4, 5, 6, 7) on which
the stones of the lowest course were set. These worked places on the rock
coincide roughly with the faces of the wall. Together with the configuration
of the rock they give us both the course and the width of the wall, which
remains about 5 m. wide as far as the S side of the Belvedere (Plan 30, 7), to
the point marked 7-7 on Plan 29. In addition to removing projections at two
places, 5 and 7, they have cut down into the rock a little. The cuttings show
the well-known characteristics of Mycenaean rock-working: rounded corners,
irregular edges and rough surfaces.
162
THE THIRD PHASE
Other traces as well are evident in this same area. They are E of the line
of the wall and at a level of about 1 m. lower. At 8 on Plan 30, there is a long,
narrow trench-like cutting with a graduated width of 0,83-0,90 m. It is care-
fully worked with a point, traces of which are clearly preserved on the rock.
While the work has been done with care, it has not the quality of Classical
workmanship. 276 Beside the SW end of the trench the rock has been levelled
(point 9 on Plan 30). At each end of 8 parts of walls still remain within the
hollow of the trench, which was therefore cut as a foundation trench. The con-
struction of the walls is typical of the period of Turkish domination, with var-
ious kinds of stones, bits of marble and plenty of lime mortar (Fig. 25). On
Plan 30 at 10, just S of 9 and at exactly the same level, there was a thick layer
of mortar covering a hollow in the rock. This was found filled with many
sherds of different periods and origins, mostly from the later years of the
Turkish domination, and the traces are thus the remnants of a Turkish instal-
lation. They are within the later wall and may be connected with a gun
emplacement known to have been in this area. 277
THESOUTHEASTCORNER
At the SE corner the wall makes a closed, almost elliptical curve that is
imposed by the rock formation. Two roughly parallel sections are preserved.
The rest of it is hidden and possibly destroyed by the Classical wall, which was
built on top of it.
The preserved NE section has already been discussed. This stops at I on
Plan 31, where it was cut off by the Classical wall. The next remaining section
of the Cyclopean wall is at II on Plan 31. It too lies within the Kimonian wall.
The top of the curve between them is usually rendered as projecting some-
276. These cuttings are recorded also by lieves them to be steps of the Classical
Kawerau (Kavvadias-Kawerau p. 101, pl. E, period, cut into the rock and leading from
n° 52), who interprets them as traces of the the E level part of the Acropolis to the ter-
continuation of the Cyclopean wall to the race N of the sanctuary of Pandion.
N. Stevens (Hesperia XV, 1946, p. 25) be- 277. Travlos, n OAEOO. fig. 138.
163
THE THIRD PHASE
Fig. 25. The E side of the rock, with the cuttings and walls of the Turkish
domination.
what beyond the corner of the later fortification. 278 At this place, however,
the rock slopes and is uneven. It shows not the slightest trace of the prepa-
ration that would have been necessary in order to found the Cyclopean wall.
278. First proposed by Kawerau (Kavvadias-Kawerau pl. A).
164
THE THIRD PHASE
The line of the outer face will therefore have been within the area covered
by the corner of the Classical wall.
The section of the wall at n on Plan 31 is not preserved to its full width.
Only the inner face has survived. The stones of the outer face are missing so
that, to reconstruct the line, another row of large stones should be added
along the outer side. The width of the preserved stones varies between 0,60
and 0,80 m. and this section of the wall preserves a width of 4,60 m. With the
addition of stones to the outer face, the width will have been 5,20 m.; it is
shown accordingly on the plan.
When the excavation of 1887 was carried out, the Museum had already
been built and explorations were therefore restricted to the interior of the gal-
leries. More remnants of the inner face of the Cyclopean wall came to light,
running in the same NW direction as the section at II on Plan 31, and con-
tinuing to III where it makes a wide turn toward the W.
The area defined by the elliptical curve of the wall, contained remnants of
different periods. The most important of these is the sanctuary of Pandion,279
not shown on the plan, however, as it is much later. In addition to this there
are various other walls along the inner face of the NW leg of the wall (Plan
31, 1-16), which must be investigated.
The complex between r and 14 was built after the Persian Wars with ma-
terial collected from various ruined buildings. Among the pieces utilised was
the Moschophoros base. NW of the complex, parallel to the inner face of the
wall, was wall 1, Plan 31. The upper part of it was built of raw brick; this was
on a carefully constructed substructure of thin flat stones laid in horizontal
courses (Fig. 26). The inner side initially had a coating of clay and traces of
a tamped clay floor are preserved at the NW corner. 280 Its construction shows
that this too belongs to historical times, but it precedes the sanctuary of
Pandion, the poros stones of which lie on top of it. The corner wall 2 (Fig.
26) is founded on a deep fill. Only a small part of it remains, but it is con-
structed in the same fashion and has the same width (0,50 m.) as 1. This holds
also for wall 3, which lies NW of 2 and in exactly the same line. Like 2 it forms
279. See Stevens, Hesperia XV, 1946, 280. Kavvadias-Kawerau p. 99, pl. E .
pp. 22-25 .
165
~
tTl
~-',18
~
......
:::0
ti
"tl
~9 ~> II ~
C/)
0 tTl
_____;..--
0 5 10 20 30 40 50
~;;;;;;;c:::;;;;;;i:::=:;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;1:::=::=::=::=::=::=::=::=:::::-=;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;i:================:::::1===============:=====! M
Plan 31. The southeast comer.

THE THIRD PHASE
Fig. 26. Walls inside the SE comer of the Mycenaean fortification (phot. DAI n° 51).
a corner and as wall 4 it continues toward the NE corner of 1. Thus walls 1,

2, 3 and 4 enclose an area that is roughly square. It belongs to historical times
and is probably contemporary with the complex mentioned above. Contem-
porary as well, if not slightly later, is wall 5, which is next to the SW corner
and built against 3; likewise wall 6, added later to 5, and partly to 3. These
walls indicate the existence of two other spaces that were added to the first
and used perhaps as workshops while the Kimonian wall was being built. This
would explain their careless construction.
WalJ 7, part of a thick wall 1 m. wide and built of large stones, is difficult
to place chronologically. Since it has cut off the SW end of wall 8, it is cer-
tainly later than that. It precedes 5, since it in turn was cut by the extension
of that wall, no longer in existence.
The other walls in the area (8-16 on Plan 31) are all earlier than these and,
as we shall see, they go back to Mycenaean times.
Among them is wall 8, preserved as a line of quite large stones. The SW
side forms the face. The uneven line of the other side shows that the stones
167
THE THIRD PHASE
that formed the NE face have been removed, perhaps for use as building
material. Its course toward the SE is cut off by wall 7. It may not, however,
have continued much beyond this point since if it made a right angled tum to
the NE and continued along this course, it will have met wall 9, built in sim-
ilar fashion and likewise destroyed. Walls 8 and 9 must belong late in the
Mycenaean period. Not only are they earlier than 7, which in turn precedes
5, but they are oriented in precise agreement with the inner face of the shorter
leg of the Cyclopean wall. Wall 9 is contemporary with wall 10. Together they
form a sharply acute angle as a result of the change in course of the fortifi-
cation wall behind them; 10 has been oriented toward the long leg and was
built at the same time as wall 9. This is evident from the fact that the corner
where they meet is bonded to some extent, being closed by small stones that
are common to both walls. If 10 is projected southward for 2 m. it meets wall
11, which projects from beneath the NW face of 4 and is built in the same
style as 10, 9 and 8. If wall 10-11 continues in the same direction for about
0,70 m., then makes a right angled tum to the SE, it coincides almost imme-
diately with the remains of wall 12. Both faces of this wall are preserved. It is
built in the same fashion as the previous walls, with two rows of relatively
large stones. Further on, the line coincides with that of wall 13, constructed
in similar fashion and with the same orientation. Wall 13, as can be seen in
both Plan 31 and Fig. 27, continues for some distance. At its SE end, it may
turn N, if wall 14 - on which stood the wall that incorporated the Moscho-
phoros base - is the continuation. While this is not clear from Kawerau's plan,
it is very probable, for wall 14 in any case appears to be unrelated to the later
complex, and it is considerably lower (Fig. 27).
Walls 10, 11, 12, 13 and 14 are of equal thickness (0,40-0,50 m.) and, with
the possible exception of 14, they all appear to belong to the same building.
The area they enclose is large and it is not a single unit. It is divided into two
approximately equal parts by wall 15 (Figs 26 and 27), which is similar to these
other walls and at right angles to the fortification wall and to walls 12-13. Wall
15 is preserved for a length of about 1,70 m. Its course toward the fortifica-
tion wall is cut off by wall 1, so it is earlier. Its projection SW meets 12-13,
thus cutting the area in two.
Their construction, their relationship to the later walls of historical times
and their orientation in relation to the fortification wall, all suggest that walls
8 and 9 are Mycenaean. The same applies to wall 10, which is contemporary
168
THE THIRD PHASE
Fig. 2Z The inner side of the SE fortification wall and walls 1, 15, 13 and 14 of
Plan 31 (phot. DAI n° 54).
with 9, and therefore also to 11, 12, 13, 15 and probably also 14 as well. The
NE wall of this building, assuming it existed and that it was not supported
directly against the fortification wall itself, will have followed the same line as
1. It may have been destroyed when 1 was built.
Thus there are three enclosed spaces, A, B and r on Plan 31, built in front
of the wall so that the tum of the wall affects the orientation of one (A), sep-
arating it from the others. There is also another wall, 16 (Fig. 28). It is at an
angle to the others and may be earlier. On Kawerau's plan it is shown run-
ning in a straight line. 281 The last stone at its SW end, however, appears to
281. Kavvadias-Kawerau pl. E.
169
THE THIRD PHASE
Fig. 28. Interior of the SE comer of the Mycenaean fortification wall, from the
S, showing walls 16 and 6 and tomb 18 of Plan 31 (phot. DAI n° 55).
turn, forming a corner toward the NW. In Fig. 28 the wall can be seen con-
tinuing in that direction, with two large isolated blocks like those at the cor-
ner lying in a line more or less parallel to wall 6; these were not recorded by
Kawerau. The whole is founded within a hollow formed by the rock, on the
N side of which wall 6 is built. A fourth space, A, of which two sides only are
just discernible, is thus added to the three.
The next building preserved in this area is denoted by number 17 on Plan
31. Built against the inner face of the fortification wall, it has a regular and
well preserved corner at the NE, whereas along its W side there are two dis-
orderly rows of stones about 1,50 m. apart from each other, rather than a wall.
The use of the building, as well as its specific form, is uncertain. If the stones
at the N belong to the building and have simply been disturbed, it could be
170
THE THIRD PHASE
interpreted as a guard-room with a staircase leading to the top of the wall.

Yet the stones may just as likely have fallen from the top of the wall and be
quite unconnected with the room. These details are not elucidated in the pub-
lication of the Acropolis excavation.
There remain the graves, one of which is at 18 on Plan 31 (Fig. 28), and
another three in a cluster at 19. They are all cist graves, their sides lined with
small slabs and with similar slabs for covers. Three of them are of small dimen-
sions, probably for the burials of children. The fourth alone contained the
skeleton of an adult, lying extended. This was the only grave containing funer-
ary gifts; it had a tall, narrow, unpainted pithoid amphoriskos of late Myce-
naean times. 282
If these graves, as is quite possible, are connected with the houses we have
been discussing, the houses will have been contemporary. This agrees well
enough with the fact that they were built after the wall and it places them at
the very end of the Mycenaean period.
THE SOUTH SIDE
We have followed the course of the wall as far as III on Plan 31. It con-
tinues in a westerly direction for a few more metres, then makes an oblique
turn toward the SW at 1 on Plan 32. Although preserved only in broken
stretches the inner line is clear from the sections discovered within the
Museum. Traces of the outer face were not found, but the enormous stones
with which it was built bespeak a massive construction, high and therefore of
considerable thickness. The exterior face has been drawn on the plan 4 m.
outside the interior face. Yet the actual width may have been greater still.
282. This grave, as the others, was pub- by Kawerau, is provided by Wolters (Graef-
lished without details by Kawadias-Kawerau Langlotz I p. XXXIV, fig. 5). Furumark
pp. 39, 95. More information, together with (MP p. 36, fig. 8, n° 58) places the pottery
a detailed sketch of the graves at 19, drawn shape in late LH IIIC times.
171
THE THIRD PHASE
This new oblique course of the wall continues to where it approaches the
Kimonian wall. At 2 on Plan 32, we find the next traces of the wall, which now
follows a westward course. This too is a section of the inner face, preseived
for the most part in a single row of stones, which do not follow a straight line
but form two successive legs set obliquely to each other. In the concavity
formed by one of these walls, Kawerau shows a pile of regularly placed stones,
notably smaller than those of the Cyclopean wall, without mentioning ....them
in the text. Whether the stones were set there in Mycenaean times as · a sort
of reinforcement for the wall or a foundation for some building related to the
wall, or whether they belong to a later time is unknown. Their relation to the
neighbouring mud-brick wall, which as we shall see is Mycenaean, is not
recorded. This would have enabled us to tell whether that wall was added to
the stones or the stones were placed at the edge of the wall. As it is, both pur-
pose and date of the pile of stones remain unclear.
At point 4 on Plan 32, Kawerau 283 reports finding stones fallen from the
Cyclopean wall and covered over by the mud-brick wall. This is puzzling
because, to begin with, it means that the wall that covered the stones of the
fortification wall must be considerably later and built after the destruction of
the wall. Yet the mud-brick wall is Mycenaean, as we said, and it was built
after the fortification wall. This is evident from its line, which follows and is
adjusted to the anomalies of the fortification wall. Yet it cannot be so late as
to allow for the destruction of the wall in the meantime.
This strange connection between the mud-brick wall and the stones fallen
from the fortification wall could be explained as a minor repair of the wall
during Mycenaean times, after which a number of stones were left where they
were and were covered over by the bricks. A more probable explanation is
that the stones were placed between the mud-brick wall and the fortification
wall on purpose, either to help drain the water that would have collected in
the narrow space between the two constructions, or simply to reinforce the
foundations of the mud-brick wall.
283. Kavvadias-Kawerau p. 103.
172
Plan 32. The south forti-
fication wall, up to the SW
comer of the Parthenon,
including the remains pre-
seTVed beside· it.
Some 5 m. W of 4, a
few stones of the inner
face of the wall are still
-~~, .
. preserved running in a
I.... ;5.
. .
141 ,65 --- ·: :.
5a :".-. ··
y, straight line toward the
- ~-
IJ't- W. Built against it is a

small square construc-
tion, measuring 1,20 x
0,80 m., with narrow
stone walls forming the
E, N and W sides with
the fortification wall pro-
viding the S boundary.
~8 The construction is later

than the fortification
wall, and later also than
141 ,95
the Mycenaean mud-brick
z wall to the E, against
0 9 which it is likewise built.
z
UI
Yet it is founded on the
:I: rock at the same level as
l- the wall 284 and therefore
a:
<{.
it cannot be very much
0..
284. Kavvadias-Kawerau pl.

0 section y-6.
0 5 10 20 30 40
a:E;t::)!!l:====l=i!!!!ii!!i!!!!!!!!!!!!!!!!ii!!!!!it:=======liiiiiiii;;;;;;;o;;;;;;;;;m;-M
THE THIRD PHASE
later. It appears to have been built before much fill had accumulated next to
the base of the wall. Its use is unknown.
The fact that there are graves and walls at 6 on Plan 32, which are of course
within the wall, shows that at least to this point the course of the wall con-
tinued unchanged. Not for any great distance, however. Recorded at 8 is a
wall of small stones with a regular face on the N side, which because of its
construction cannot possibly be part of the wall. Since it is not mentioned in
Kavvadias' text, and since the level on which it was built cannot be determined
from the plan, it is impossible to determine whether it is Mycenaean or not.
There are two stones of the outer face of the fortification wall at 9, against
which, after the wall was destroyed, a kiln was built with the entrance above
them. 285 The stones show that while the general direction of the wall remained
the same, the line has been shifted further north than it was at points 4 and
6. The wall will therefore have turned toward the NW either between 6 and
8 or between 8 and 9. The morphology of the rock suggests the first point
because at that place the gradient is gentler. The wall has thus been drawn to
show the turn at 7. Wall 8, being outside the fortification, must be consider-
ably later. Yet only the precise dating of this wall could have resolved the
problem.
From 9 on the course of the wall is clear, for it is preserved in a continu-
ous line up to and beyond the SW corner of the Parthenon. The outer side is
preserved intact for its entire length, while the inner face is preserved at two
places on either side of the corner of the crepidoma of the temple. Here,
therefore, the thickness of the wall can be calculated exactly at 4 m. For the
rest of its course, the face has been destroyed by the polygonal terrace wall S
of the Parthenon, which was built parallel to the line of the wall and on top
of it, and because, as Kolbe showed, 286 stones were removed for use in the
foundation of the SW corner of the crepidoma of the temple. The stones
285. Kavvadias-Kawerau p. 119. p. 344, Research and Progress 1940, pp. 254-
286. AA 1939, p. 235, FuF 1939, p. 394, 255.
Bericht VJ Intern. Kongr. Archiiologie 1940,
174
THE THIRD PHASE
removed from this part may have been those found in a pile at 11 on Plan
32. 287 It would have been natural enough for the Persians while destroying the
wall, to have rolled them downhill, but they would not have remained there
on the slope. The presence of the Kimonian wall, however, and the poros sup-
porting wall with its corner built inside it, would have prevented the builders
of the Parthenon from shoving the big stones off the rock.
The wall stops at 13 on Plan 32, where its continuation is cut off by the
construction of the stairway W of the Parthenon and by the Chalkotheke.
The area was inhabited, indeed closely inhabited, as is evident from the
various Mycenaean remains preserved within the Cyclopean wall. At 3 on Plan
32 two long walls were found, the northernmost of which appears to curve at
the W end. It may well be the later of the two, since their relative position
makes it unlikely that they existed at the same time. The southernmost is most
certainly Mycenaean. It is founded directly on the rock and it is constructed
of sun-baked brick built on a foundation of stone 1 m. high. The whole thing
had been covered over by the Persian destruction level. A hoard of bronze
weapons, vessels and implements with traces of wood still in the handle attach-
ments were found between this wall and the inner face of the fortification
wall, at the point marked by a cross on the plan. These had been hidden on
purpose, some in the empty space between the two constructions, some among
the foundation stones of the wall. The sherds found with the bronzes are all
Mycenaean, the latest sherd belonging to LH IIIC early times. This has a rep-
resentation of a human figure, preserving the lower part together with the
legs. It is painted in brown directly on the yellowish ground of the clay. 288
287. Kavvadias-Kawerau p. 119. plain axes, one chisel, one rasp, one knife,
288. Ibid. pp. 37-103, AM 1888, p. 107, one sickle, one tool similar to a wide chisel,
BCH 1888, pp. 244-245. The find was pub- one object resembling a half-finished dirk,
lished systematically in detail by Montelius, 2 round mirrors and one two-handled bowl.
VHAM 1889, pp. 49-60, figs 1-25, and La The entire assemblage is very similar to
Grece pp. 155-156. Two of the sherds shown what Mylonas found on the Mycenae acro-
by Montelius are published also by Graef- polis, likewise within the fortification wall
Langlotz I as nos 202 and 222. The bronzes (Ergon 1959, p. 99, fig. 104).
are: one dirk, one spear, 9 double axes, 5
175
THE THIRD PHASE
Found in addition to these sherds, there appears to have been a stirrup jar
that was not described. 289 At point 5 is a complex of walls, defining irregular
spaces and founded on a layer of fill varying in height because of the slope of
the rock. These walls are built in the same fashion and the areas between them
have floors of small irregular stones covered with a layer of clay. 290 The floors
are all at the same level and the walls therefore belong to the same building.
In addition to this complex of walls, Kawerau notes three others, one W and
two E of 5, without providing any information. No conclusion about their
chronology or use is possible.
Three more sections of walls are preserved at 6. Their construction sug-
gests that they belong to Mycenaean buildings, constructed probably against
the wall. There are also four cist graves of children, the walls lined with small
slabs. The westernmost appears to have been surrounded by a wall built of
small stones, only part of which is preserved. Found in one of the graves were
the bones of a child. The burial was accompanied by sherds and offerings com-
prising two pebbles and a one-handled skyphos, painted brown inside and out
except for the base. 291 These graves are similar to and contemporary with
those found at the SE corner, and they belong to the final years of the
advanced LH IIIC phase.
The excavators' comments on the various fills are informative for the devel-
opment of this area and the succession of buildings here. It appears that
slightly S of the Parthenon, on a line with the isolated section of wall to the
N of 6, unassigned and not reported, the rock was exposed down to Classical
times. Further south, the Mycenaean wall caused the build-up of fill in the
course of time. As shown by the height of its stone foundation, the wall where
the bronzes were hidden was built before the fill exceeded a height of 1 m.
Later on, the fill rose to some 2 m. next to the wall. Kawerau's section of the
area makes it clear that the walls next to 5 on Plan 32 were built over this fill
289. It is mentioned only in the BCH (where the vase is illustrated and the peb-
1888, p. 245. bles are mentioned), Kavvadias, Deltion
290. Kavvadias-Kawerau p. 103, pls E 1888, p. 83, and Graef-Langlotz I n° 176,
and E> section y-6. where the vase is published.
291. Ibid. pp. 37, 121, AM 1888, p. 228
176
THE THIRD PHASE
and that the graves were dug into it. Thus, by the end of Mycenaean times,
the fill had reached this height. The southernmost wall at 3 must be slightly
earlier. 292
Finally, the wall at 10 on Plan 32 is later than the destruction of the Cyclo-
pean wall. It is built against the outer face just as is the kiln at 9, with which
it might be connected. The lower part of the wall is constructed in polygonal
fashion and from Kawerau's plan it appears to have regularly alternating large
and small stones. The upper part was composed of crude bricks, some of which
were found fallen into the inner corner. Kawerau attributes the stones that
are piled up to the W to this same construction. Yet, to judge by their size,
they are more likely to have been part of the filling of the fortification wall. 293
Thus wall 10, the kiln at 9 and probably wall 8 will have been built and in use
between 479 B.C., when the Mycenaean wall was destroyed, and 465-460,
when the Kimonian wall was erected.
After the SW corner of the Parthenon, the wall begins to widen from 4 m.
to around 5,50 m. It stops, however, at the point where it was cut off by the
stairway W of the temple and the foundation of the Chalkotheke (Plan 33, 1).
Its remains are encountered again some 75 m. to the W, approximately on a
line with the E fa<;ade of the Propylaia. Here the N, interior face of the wall
is preserved inside and parallel to the Kimonian wall (Plan 33, 2). The face is
preserved for a course of around 10 m., following the westward course of the
S wall. It then turns abruptly N, forming an almost acute-angled corner and
then runs straight (Plan 33, 3), with both faces still preserved and a width of
6 m. The SE corner of the S wing of the Propylaia touches the W face of this
leg of the wall. Further N it stops, cut off by the pre-Mnesiklean propylon (4,
Plan 33). Thus the SW corner of the fortification circuit is preserved, visible
today for its full length. What is preserved of the S leg is short and incom-
plete, the W practically complete. Missing is the outer top of the corner and
the S face of the short leg. The wide leg facing W remained visible through-
out antiquity. Because it was not buried after the catastrophe of 479 B.C., but
continued to be used as part of the Brauroneion peribolos, it underwent fre-
292. Kavvadias-Kawerau pp. 37, 103, pl. 43, Montelius, VHAM 1888, p. 51.
0 , section y-6, Kavvadias, Deltion 1888, p. 293. Kavvadias-Kawerau p. 119.
177
THE THIRD PHASE
quent repairs and additions even up to Mediaeval times. That it was repaired
not long after its destruction is evident from the late 5th century sherd found
by Botticher within the wall itself, 0,50 m. below the top as preserved. 294
Relevant to this is the question of its original height. Dorpfeld 295 main-
tained that at the time the Propylaia was built, the wall was still standing to
a height of about 10 m.; this because right up to the roof, the corners of the
marble blocks of the Propylaia that touch the wall have been cut back so as
to fit against the face of the wall without damaging it. As many other obser-
vations of Dorpfeld's, this became generally accepted. Yet, in my opinion, it
does not fit the facts. To begin with, it is quite unbelievable that the Persians,
who destroyed the fortifications of the Acropolis so thoroughly, would have
left the wall intact precisely at the entrance. While this opinion could be chal-
lenged, there is positive evidence, as White observed and explained in reply
to Dorpfeld. 296
The wall today is preserved to a height of 3,45 m. above the surface of the
rock. Up to that level, the corner of the Propylaia has been cut back 0,89-0,90
m., in order to adjust it to the surface of the Cyclopean blocks. From this
point upwards it continues to be trimmed back, but only around 0,40 m., so
that the edge of the lowest of these blocks rests on the top stone of the wall.
This is explainable only if we assume that the outer face of the wall was not
the same for its full height, but from the height of 3,45 m. upwards it was
recessed around 0,50 m. from the surface of the lower part. Such recessing of
the face of the wall is completely unknown, not only on the Acropolis of
Athens, but in any of the Mycenaean fortifications elsewhere, and such a
reconstruction is unacceptable. There is other evidence as well. A cutting is
preserved on the top surface of the uppermost block of the Cyclopean wall at
the NW preserved end of this leg, where it meets the corner of the pre-Mne-
siklean propylon. It runs in a straight line (Plan 33, 4a) continuing the line of
the fac;ade of the propylon. This means that when the propylon was built, the
294. Akropolis p. 60. 296. Ephemeris 1894, pp. 50-51, and

295. AM 1885, p. 139. See also Judeich, D'Oodge, Acropolis, Append. II, pp. 361-
Top. p. 115 and n. 2. 368.
178
THE THIRD PHASE
I \
~ ,.,
Plan 33. The south fortification wall from the SW comer of the Parthenon to the
Propylaia.
poros blocks of its S corner rested in part on top of the Cyclopean wall, which
was preserved at this point to a level even lower than at the previous place.
Thus, as White demonstrated, after the Persian destruction the Cyclopean
wall here was preserved to the height of this cutting and further south to the
level at which the corner of the Propylaia rested on top of it; that is, the height
it has today. The cutting back of the corner of the Propylaia above that point
on up to the roof of the building shows that after the Persian destruction and
before 437 B.C., a high wall was built on top of the Mycenaean fortification
wall, recessed, and thus much less thick than the Mycenaean wall. Indeed it
has now lost its characteristics as a fortification wall, so that it appears to have
been simply part of the Brauroneion peribolos. It was against this wall rather
than the Mycenaean wall that the corner of the Propylaia rested from the
height of 3,45 m. up. 297
Despite all these vicissitudes, the wall continued to retain its Cyclopean
character unchanged, with the enormous stones of its fa~ade founded directly
on the rock, small stones wedged into the few spaces between them, a corn-
297. Similar, but somewhat vague, views dias-Kawerau pp. 137, 139, pl. H).
were expressed also by Kawerau (Kavva-
179
THE THIRD PHASE
Fig. 29. The outer, W face of the fortification wall that is preserved S of the Propy-
laia (phot. by P. Mylonas).
pact fill and its great imposing mass (Fig. 29). The outer face is preserved for
its full length unbroken and only at the S end have its corner stones been
replaced for a width of 0,50-0,60 m. by the corner of the later wall. The inner
face is less carefully constructed with smaller stones. This is especially notice-
able in the shorter S leg, which includes not only Acropolis stone but also
some stones from the Pnyx and pieces of schist as well. Up to about the level
of the present surface, the lower courses are more carelessly laid than those
above. The fill that hid them, therefore, at that time was at about the same
level as today.
The inner side of the corner of this section of the Mycenaean fortification
wall is preserved undamaged, but not the outer side. To begin with, the
Kimonian wall was built in its place, and later on, especially during the Turk-
ish domination, various additions and changes were made. Yet there is evi-
180
THE THIRD PHASE
dence that leg 3 continued about 1 m. beyond the point to which it is pre-
served today, and that the corner projected about 0,50 m. S of the Classical
wall. At precisely this point, the S wall of the Classical bastion of Athena Nike
meets the S leg of the Kimonian wall, together forming an obtuse angle. This
corner continued, in a way, the line of the Cyclopean wall. It comprises a num-
ber of courses of partly destroyed poros stones on which stands the wall in its
present form. These courses must therefore be earlier (Plan 33, 5). The poros
stones show an apparent effort to repair on the spot the destroyed corner of
the walL before erecting the Kimonian wall. Thus they provide evidence for
its original plan, which they followed more faithfully than did the Kimonian
wall. The rock, in any case, prohibits its continuation further S. For the
restored drawing of the wall (Plan 33), this line has been taken into consid-
eration. It gives the beginning of the S leg a width of about 4,40 m., far more
plausible than that obtained by adapting it to the line of the later wall.
Between the ends of the stretches preserved at 1 and 2 on Plan 33, no trace
of the wall has survived. The surface of the rock was cut down and levelled
over the entire area during the construction of the Brauroneion and the
Chalkotheke and the only evidence remaining is the configuration of the rock
itself.
As is evident from the contour lines, the rock slopes gradually toward the
W in the W half of the area where the Brauroneion stood, and falls off steeply
to the S. While there is no slope toward the W in the Chalkotheke area, the
rock makes a precipitous drop to the S. Here there is no natural brow and
the declivity is such as to exclude any possibility that the Mycenaean wall stood
in this place. The stretch of wall beside the SW corner of the Parthenon there-
fore marks the beginning of a turn toward the NW so as to circumvent the
declivity. On the basis of what we know, this may be taken as sufficient evi-
dence.
At point 1 the wall is founded at around 149 m. The curve it makes shows
that it is headed uphill where the rock is flatter. At the elevation of 150-151
m., the southward slope is already gentler and quite suitable for the wall foun-
dation. The wall therefore will have continued its curving course about as far
as the NW comer of the Chalkotheke. From here on it will have met the gentle
westward slope, which takes it obliquely toward point 2 where it is founded
at 145 m. The course shown on Plan 33 reflects these observations and while
it is not the only possible solution, it is the most likely.
181
THE THIRD PHASE
In the corner 2-3-4, at 6 on Plan 33, Kawerau records an angular wall,

opening to the N. The wall, carelessly built of different sorts and sizes of
stones, including marble and pieces of roof-tile joined with lime plaster, is a
foundation of the Turkish period. Inside the angle are traces of a carelessly
laid slab flooring. Finally, preserved at 7 on Plan 33 are two of the Middle
Helladic children's graves already discussed. 298
THE ARRANGEMENT OF THE WEST ENTRANCE
The course of the wall encompassing the top of the Acropolis has now been
examined and its line re-established as accurately as the evidence permits.
Another matter remains to be investigated, that of the form of the main
entrance of the fortification at the W.
It is perfectly clear that the entrance is opposite the S wing of the Propy-
laia. This is the only really accessible part of the rock. Moreover there is also
the bastion, which would otherwise be both unnecessary and inexplicable.
The discovery of the bastion placed the problem on a new basis and the
arrangement of the entrance was now understood to depend on a main ques-
tion: the correct relationship of the bastion to the wall. The position of the
gate and the course of the approach depend on how these two are associated.
Let us look at the situation as it is.
The bastion stands on the spur of the rock SW of the Propylaia (Plan 34,
1). Except for the NE corner, its form and position are known. Opposite it to
the E is preserved the massive leg of the Cyclopean wall at point 2 on Plan
34. It runs in a straight line, parallel to the E side of the bastion. N of it and
slightly oblique in relation to it, is the curving section of the wall at 3. Between
this and the N side of the bastion there is thus a space, which in form and
dimensions resembles the entrance to the acropolis of Mycenae, in front of
the Lion Gate. 299
298. See supra, p. 54, and Pl. 2, 5. m., the available width of the entranceway
299. At Mycenae the length of the side (from the base of the rock to the bastion
of the bastion toward the entrance is 14,80 opposite) in front of the gate 7 m. and at
182
THE THIRD PHASE
Since the bastion came to light, four specific solutions to the problem have
been proposed. Stevens300 and I. T. Hi11 301 leave the bastion isolated and
entirely unconnected to the wall. This clearly is impossible. A bastion that
does not communicate with the fortification is completely useless. In the case
of an enemy attack it is even dangerous, since it is impossible to defend and
if it fell into enemy hands it could serve as a base for storming the entrance.
Welter302 and Travlos 303 connect it with wall 2 on Plan 34 accepting the idea
of a guard-house in the space between the bastion and the wall. This arrange-
ment, however, presupposes a wall running from the SW corner of 2 to the
SE corner of the bastion, closing off the space between. Such a wall, forming
part of the outer face of the fortification, would have to be strongly built, com-
parable to the short wall blocking the opening between terrace walls I and II
(see Plans 26, 3 and 27, 3). Like this wall, it too would have to be some 2 m.
thick. Moreover, both the corner of 2 and the corresponding corner of the
bastion (which, we may note, are not even aligned) should have had traces of
this wall on the first two metres at the S, and the rock itself should have had
similar marks where the foundations stood. Yet there are no traces on either
one. The natural surface of the rock has survived without any tooling at all,
despite being uneven, steep and full of small cracks at this spot. Moreover,
both the E side of the bastion, as far as it is preserved, and especially the W
face of the wall opposite, which is preserved to 1,10 m. from its outer corner,
have continuous and unbroken faces showing that there was no cross-wall con-
necting them anywhere. This is clear on the plan, and clearer still from the
appearance of the surface of wall 2.
We may conclude that the bastion is not connected with 2. It therefore has
to be connected with 3, running eastward to the line of its NE corner. The main
gate of the fortification must have been precisely at this spot (Plan 34, 4).
the beginning of the bastion 6 m. The cor- 303. IloA.EoC>. fig. 7. In pl. I of the same,
responding measurements at Athens are 14 he shows another arrangement, connecting
m., 7,75 m. and 6,30 m. the bastion with wall 3 on Plan 34, with a
300. Hesperia XV, 1946, pp. 73-106, wall closing the passage which runs along
fig. 2. the S side of the bastion, cuiving afterwards
301. Athens fig. 3. toward the N between that and wall 2. The
302. AA 1939, fig. 4. rock formation rules out this course.
183
THE THIRD PHASE
The main gate, but not the only gate. For if wall 3 ended at the inner line
of this entrance and the empty space between it and wall 2 were open, then
both bastion and the gate at 4 would be superfluous, the existence of wall 2
meaningless and the access to the Acropolis would have been free and unhin-
dered. The rock to the S of the bastion is indeed steep and difficult, but it
can be climbed if need be. There will therefore have been a second gate at 5
on Plan 34, at right angles to the first and evidently of smaller dimensions.
The width of this second gate is given with fair accuracy by the width of the
space between the bastion and the wall at 2. Its depth can be calculated from
the eastern limit of the north face of wall 3 in relation to the length of the
preserved outer face of wall 2, which cannot possibly have turned toward the
W further south than the last preserved stone at the N end.
This arrangement, as I believe I have shown, is the only one possible. All
others are excluded by the evidence. The plan has great tactical advantages
for the defenders. Indeed any attackers reaching the space between the bas-
tion and wall 3 would, as at Mycenae, be exposed on both sides to the defend-
ers' missiles, but they had space in which to manreuvre and to retreat again
to their base of assault. Those who happened to climb the rock and get as far
as the narrow space between the bastion and wall 2 would find themselves in
a trap from which there was no escape; any attempt to retreat down the pre-
cipitous rock would be tantamount to suicide as they would be attacked from
all sides.
Thus there had to be two gates, planned as described. There is no sure
evidence for their dimensions or for their exact appearance. For this we must
resort to similar constructions in the well-known citadels.
The space in front of gate 4 is similar enough in plan to that at Mycenae
to justify reconstructing the gate on the basis of the Lion Gate. Its opening
would be 3 m. wide and around 4,40 m. in depth, coinciding with the thick-
ness of the wall it penetrates. The depth is unlikely to have been any greater
since its interior line is defined by the east side of the bastion, where pre-
served. The outer side, to the west, must have been founded on the uneven
rock surface. The height of the gate should be somewhat greater than its
width, thus 3,10-3,20 m., allowing for an additional relieving triangle above, a
necessity given the weight of such a construction. By analogy with that at
Mycenae, the triangle, if any, and the superstructure would have measured
184
THE THIRD PHASE
142,45
5 I
• , 142,40
L-
141 ,40
4
r - --, ........ ... ... .. .. .
I 1
I
. ....... . . .
I ) .. ... . . . .
I
I
~ >
) )
138,36
0 5 10 20
~H~H~Hl====~~~~~~I M
Plan 34. The southwest comer of the fortification.

185
THE THIRD PHASE
3,80-4 m. at the base (in order to avoid shear-force above the lintel, to the
extent possible) and been equal in height or slightly higher than the gate.
The inner gate, narrower than the outer, will have had correspondingly
smaller dimensions, determined on the basis of its proportions and the width
of the opening. This cannot have exceeded 2,50 m. The depth, as calculated,
will have been as much as 3,50 m.
The rock slopes from E to W at the location of the gates. The thresholds,
as in all gates, will have to be horizontal. Thus their level is determined by
the altitude of the rock at the eastern end of each threshold, which gives us
the lowest level possible for each one.
Along the east side of the bastion and so also along the inner side of the
outer gate, the level of the rock is at 141,40 m., rising to 142,40 m. beside the
foundation of wall 2. Thus the threshold of the outer gate will have been at
141,40 m., that of the inner gate at least at 142,40 m.
The height of the wall cannot be calculated with precision, since no relevant
information has sutvived. Yet on the basis of our calculations for the
gates, which must be within the range of probability, we can estimate the fol-
lowing.
The Lion Gate, the only one to have survived, has the following measure-
ments: width 2,95 m., height 3,10 m., thickness of lintel 0,80 m. and height of
the relieving triangle above the lintel 3,30 m. These proportions must hold as
well for the inner gate of the Acropolis of Athens. Since this stands on a level
1 m. higher than the outer gate, the height of the wall will have been affected,
for it had to be the same for the entire entrance complex. Moreover, because
the space is narrow, there is less room for the opening, which can be calcu-
lated more accurately at around 2,50 m.
Thus, in accordance with the proportions of the gate at Mycenae, the
dimensions of the inner gate may be calculated as follows: width of opening
2,50 m., height 2,50 m., thickness of lintel 0,65 m. and height of relieving tri-
angle 2,60 m. This gives a total height of 5,75 m. from the threshold to the
top of the triangle. If another four courses, the fewest possible, are added
above the triangle, the total height of the wall will have been around 9 m.,
bringing the top of the fortification to a level of 151,40 m. This means that
leg 2 on Plan 34, is 9 m. high and the W side of the bastion, founded at 135,45
m., is 16 m. high. If, on the other hand, we accept that there was no super-
186
THE THIRD PHASE
structure above the lintel (as at Gla), the total height of the entrance would
be ea. 3,20 m., bringing the top of the whole structure to 144,60 m. and the
W fa~ade of the bastion to a height of ea. 9 m.
Preserved below the W side of the bastion, as we have seen, are traces of
two separate pathways, the lower of which was designated for animals. Both
form and arrangement of these two pathways can be reconstructed approxi-
mately, based on the traces preserved and on the slope to be ascended.
The first of these pathways is in the form of a ramp rising along the foot
of the bastion. The preserved ramp, with a set width of around 1,50 m. , begins
below the SW corner of the bastion at the 133,40 m. level (Plan 35, 1). Pre-
served at a level of 135,45 m. on the surface of the rock at the NW corner of
the bastion (Plan 35, 2), is the shallow four-sided cutting around which cult
traces were found, related evidently to the niche in front of which there must
have been an open space. The cutting must thus have been uncovered and
therefore higher up than the level of the ramp. This provides a limit to the
height that could have been reached by the ramp from its beginning to the
NW corner of the bastion (Plan 35, 3) which is at around 135,20 m. Thus the
first leg of this pathway (Plan 35, 1-3) rises toward the N with an incline of
14,5%, which is a more or less easy climb.
The slope changes after the corner. From 135,20 m. , the approach must
reach a level of 141,40 m., that is the height of the threshold of the outer gate.
We may assume that the threshold itself is in the form of a step, low enough
to be negotiated by animals entering the Acropolis, at most 0,10 in height.
The ascent would thus have had to rise from a level of 135,20 m. at the NW
corner of the bastion to a level of 141,30 m. in front of the threshold of the
first gate. This is a rise of 6,10 m. over a length of 14 m., which means an
incline of 43,5% . Such an incline requires a stairway. Shown as an example
on Plan 35 is a stairway of 20 steps with risers 0,20 m. high, treads 0,60 m.
and an incline of 13,33% . They ascend to a wide landing (Plan 35, 4) in front
of the threshold, which occupies the full space in front of the gate and equals
the risers in height, with an incline of 13,6% . The actual arrangement of the
steps of the Mycenaean approach will, to be sure, have been much more
uneven, perhaps more difficult to climb. The solution illustrated is simply for
the purpose of showing the limits within which the stairway will have been
constructed.
187
THE THIRD PHASE
0 5 10 20
M
Plan 35. The affangement of the west entrance.
188
THE THIRD PHASE
Of the second ascent cuttings are still preserved in the surface of the rock
(Plan 35, 5-6) that rise from 131,89 to 132,93 m. over a length of 5,70 m. Since
these are unquestionably part of the authentic Mycenaean approach, we can
accept the 18,25% incline given by this leg as a basis for calculating the rest
of the ascent.
Further on toward the same direction but with a few turns the course of
the ascent reaches the SE corner of the base of the Agrippa monument (7 on
Plan 35), where the rock is at a level of 136,93 m. Thus over a course of 22
m. a 4 m. rise is gained, with an incline of 18,20%. From this point it would
have been possible to continue along the wall with an incline of 17% to the
threshold of the outer gate. The pathway from here on will have needed arti-
ficial fill at a number of points, especially just before the gate, but nothing
more complicated.
There are similar problems with the approach between the two gates, since
there is a difference of 1 m. in level. The distance between the two gates is
3,60 m. This means an incline of 28%, which requires steps. This is not diffi-
cult for people, but this final stretch had to be negotiable also by animals, and
pack animals at that. Thus the steps must have been very low. The rock itself
presents further difficulties in that it slopes from E to W, at right angles to
the axis of the stairway. Indeed if the steps ran parallel to the fa~ade of the
inner gate, they would end at the E before they had covered the full width
between the outer gate and the wall to the E of it, becoming progressively
narrower, with the lower steps so narrow as to be practically useless.
This difficulty is overcome if the steps are arranged in a fan-like curve, so
that the lower step is almost parallel to the threshold of the outer gate and
the top step parallel to the fa~ade of the inner gate (Plan 35, 8). The ascent
thus climbed a slope of 14%, with 6 steps of risers 0,06 m. high and treads
about 0,50 m. wide at the beginning and 1 m. wide at the end, to reach the
inner gate, with a threshold around 0,06 m. above the top and final step.
While lacking concrete evidence other than the existing levels, this solu-
tion, shown on the plan, is the most logical and it conforms to known Myce-
naean practice. 304
304. James C. Wright, Hesperia 63, 1994, struction of the W entrance of the Acro-
pp. 324-360, pls 77-80, questions this recon- polis. Based on a combination of the inter-
189
THE THIRD PHASE
THE PALACE
The palace complex of the kings of Athens305 stood on the terraces at the
top of the rock, protected by the strong and dominating fortification we have
described. No remains of the palace are preserved in situ. Yet there are a few
architectural members that in all certainty belong to it.
pretations put forward by Bundgaard, Mne- the natural bedrock of the area at the end
sikles (Copenhagen 1957), W. B. Dinsmoor of the Late Bronze Age, perhaps as a monu-
Jr. (The Propylaia to the Athenian Acropolis, mentalization of the entrance." He also
1, The Predecessors, Princeton 1980) and I. agrees with W. Dinsmoor Jr, in assuming
Mark (The Sanctuary of Athena Nike in that the piece of the Cyclopean wall dis-
Athens: Architectural Stages and Chronology, covered by Stevens in front of the Propy-
Hesperia, suppl. 26, Princeton 1993) and on laia and the disorderly pile of stones in the
assumptions of his own, he explains away basement of the Pinakotheke represent the
the cuttings on the rock brought to light by remnants of a low extended terrace (his figs
Beule as being Mediaeval although Beule 8, 9), which, however, would have buried
reports explicitly that he found them cov- the Mycenaean house walls next to the
ered successively by a sandy layer, slabs of stone pile and would have provided an
the Archaic ramp and the Mediaeval as- ideal assault platform to those attacking the
cent. According to Wright's thinking the gate. As to the north and eastward continua-
cross-wall inside the compact fill of the bastion of the fortification wall above the caves
tion is unlikely to have served to break the in the Pelargikon, he ignores completely
load against its W fa~ade but formed rather the configuration of the rock, scrupulously
the side of a room built within the bastion adhered to here and everywhere by the
and entered from the ground level at the Mycenaean builders. In view of all this I do
rear of the construction. The piece of wall not think that I need to make any changes
at 3, Plan 17, 1 m. thick, perfectly aligned in my version of the arrangement of the
with the Cyclopean wall to its E and the Mycenaean W entrance to the Acropolis.
fa~ade of the bastion to its W and built, What happened to it in later times is of no
according to the excavators, with unworked concern to this study.
boulders ('n'tav6A.L8m) he believes to be 305. It may be that the palace on the
neither part of the bastion nor Mycenaean. Acropolis was not the only royal dwelling,
In fact Wright considers the whole bastion, for the Athenians of historical times, on the
which he lowers to the level of the rock to basis of old traditions, referred to a build-
its E and treats as a platform bearing an ing near the Ilissos that was known as the
upright room-like structure on its W half, house of Aigeus (Plut. Thes. 12, 3 and see
as nothing more than "a formalization of Judeich, Top. p. 387). The house is near the
190
THE THIRD PHASE
The first is a column base that was discovered during the excavation of
1887, lying "not in its original position, but on the flat space east of the
Erechtheion," 306 somewhere above the E parts of terraces II and Ill Today it
rests E of the Erechtheion beside the Classical fortification wall. It consists of
a fairly large block of Acropolis limestone, the upper side of which is quite
flat and has been carefully smoothed for the reception of a wooden column.
No attempt was made, however, to obliterate the many natural cracks and
anomalies of the stone. Since there are no tool marks, the stone was probably
smoothed with sand (Fig. 30) .
•
Only the tipper part of the stone has been worked. It is a low cylindrical
projection, 0,11 m. high, in one piece with the rest of the block. The side of
the cylinder is not exactly straight but slightly conical. The circle is irregular
and the diameter ranges between 0,76 and 0,81 m. The edges have been
rounded and that part of the base has been smoothed in the same way as the
flat surface at the top. The lower part of the base, rugged in shape, has been
left unworked and it is uneven in thickness too (Fig. 31). This will have been
hidden by the floor of the place where the column stood. The form of the
lower surface shows clearly that it was not set directly on the rock or on some
other hard surface, but that it was sunk at least some 0,32-0,35 m. in fill, per-
haps stabilised by some light substructure as well. In material, construction,
form and dimensions, the piece is similar to comparable bases from other
Mycenaean palaces.
Two large sandstone slabs rest against the stones of the fortification wall,
not far from the base. Their origin and place of finding are totally unknown.
They are not mentioned either in the Acropolis excavations or by other explor-
ers of the rock and to the best of my knowledge they have gone unnoticed
sanctuary of Apollo Delphinios next to the installations in the Ilissos area (see Thuc. II
Gates of Aigeus, identified by Travlos 15, Plato Kritias 112, b-d; see also Broneer
(lloA.eoo. p. 53) as the gate which he exca- in AJA 1941, p. 92) and is beyond the scope
vated N of the Olympieion. The entire of the present study.
question is connected with that of the exist- 306. Kavvadias-Kawerau p. 89. See also
ence of the two sanctuaries and other p. 83.
191
THE THIRD PHASE
Figs 30-31. The Mycenaean column base; from above (fig. 30) and side
view (fig. 31).
until now. Since they had been kept together (one of the slabs is right next to
the base), they must have been found in the same area as the base. Given the
size of the slabs, moreover, they are most unlikely to have been moved far
from where they were found. One of them may possibly be the slab shown in
Fig. 11, lower right, lying on the foundation of the ancient stairway north of
the Erechtheion. Fragments of similar slabs were used by Kavvadias as build-
ing material for the walls he erected around the NE approach. It is thus very
likely that their original position was somewhere near the base.
They are both made of the same stone and worked in the same fashion;
they are undoubtedly contemporary, belonging to the same construction.
While the sides of the slabs are for the most part flat and carefully cut, there
is a clear distinction between the large surfaces. On each slab, one of these
surfaces is carefully levelled and smoothed, whereas the other is uneven and
more coarsely worked. One side was clearly meant to be seen, the other not.
The slabs therefore were either orthostates applied to a wall, or part of a slab
pavmg or, most probably, steps. Careful examination of them provides an
answer.
192
THE THIRD PHASE
Figs 32-33. The first slab: the upper, visible surface (fig. 32) and the hidden
surface (fig. 33).
The first slab is not preserved for its full length. Both ends have broken
off and are lost. Its maximum preserved length is 1,01 m., width 0,69 m., and
thickness 0,18 m. Preserved on the smoothed surface that was meant to be
seen, along one of the long sides is a narrow band, 0,10 m. wide. It is
imperceptibly higher than the rest of the surface but the difference is so small
that it was clearly not part of the original cutting. The band had simply suf-
fered less wear than the rest of the surface evidently because it was covered.
The rest of this surface is evenly worn; even the edge has been rounded
through wear, especially in the middle (Fig. 32). The side that was hidden (Fig.
33) shows clear traces of having been worked with a narrow drove. 307 Similar
toolmarks, dating the slab unquestionably to Mycenaean times, are preserved
on the surfaces of the regular blocks of the fa~ade of the tomb of the Lions
at Mycenae.
307. The tool was not unknown in My- of bronze, it could only be used on very soft
cenaean times (Orlandos, 'YA.ixa Ooµfjr; II stone.
p. 125, n. 1 and fig. 62). Since it was made
193
THE THIRD PHASE
Figs 34-35. The second slab: the upper surface (fig. 34) and the lower surface
(fig. 35).
The other slab is better preserved. One end has broken off unevenly, while
the sides and corners have suffered only minor damage. It is preserved to a
length of 1,29 m., a width of 0,68 m. and it is 0,17 m. thick. Here too the
smoothed surface has a slightly raised, protected band, in this case 0,14 m.
wide. The edges of the corners have traces of damage in places, probably
incurred while excavating or moving it, rather than from use, as with the other
slab (Fig. 34). The initially invisible part of the surface is fairly uneven and it
has not been smoothed as was the other. Moreover it has a narrow irregular
cutting, 0,01-0,02 m. deep, close to and along the preserved narrow side. The
masons appear to have found that the surface on which they were setting it
projected slightly at this point and so they tried to adjust it (Fig. 35). That
194
THE THIRD PHASE
they trimmed the slab rather than the support suggests that the stone of the
slab was less hard.
In addition to these two slabs there are, as we mentioned, fragments
of others that were built into the walls erected by Kawadias E of the
Erechtheion. There are five of these altogether, at least two of which preserve
worked surfaces. Another, the largest of all, is built into the corner of the wall
that covers the N supporting wall of terrrace IV, precisely over the big Myce-
naean block of the NW corner of the supporting wall (see also Fig. 44). The
piece comes from the corner of a similar slab. It is preserved on each side of
the corner for a length of 0,55 and 0,56 m. and it has retained its original
thickness of 0,17 m. Thus at least three fragmentary slabs of this sort survived
and were found in the Acropolis excavation. Their use remains to be seen.
The wear along the edge of the first slab, attributable to use, shows that
they were not orthostates. This sort of rounding of the upper or lower edge
of a slab to be set upright makes no sense. Therefore it was placed horizont-
ally, either one next to the other with the smoothed surface on top as a paving
for a courtyard or similar area, or else one was placed above the other to form
treads in a stairway. The second interpretation is the most likely and the nar-
row protected band along the edge is clearly the part of the tread that was
covered by the next step. The wear along the edge of the first slab is evidently
due to use.
The slabs are therefore steps, from a Mycenaean building to judge by the
material, the stonecutting and their proportions. 308
Two ways of using similar steps are known in Mycenaean palaces. One is
at Tiryns, where the entrance to the ante-chamber of the megaron is stepped
with two low steps taking up the entire width. The first, lowest step is a slab
of similar sandstone. The second, of limestone, is placed over the edge of the
first. 309 The second method is seen in the stairway of the palace at Mycenae,
308. The possibility that they were used for a door pivot to support this interpreta-
as thresholds has not been examined since tion.
the material would not be suitable for that 309. See Dorpfeld in Schliemann, Tiryns
use, nor are there any jamb traces or hole p. 239, fig. 114.
195
THE THIRD PHASE
where the steps preserved are likewise made of sandstone,310 and in the well
preserved stairway of House M at Mycenae. 311
For the first method, a presupposition is that the exposed part of the upper
surfaces of the slabs, since they are part of the same stairway, must have
exactly the same width. This is not the case with the slabs from the Acropolis
at Athens. On one, the hidden band is 0,10 m. wide and the free surface of
the tread 0,59 m., whereas on the other the width of the hidden band is 0,14
m. and the exposed surface 0,54 m. The treads are thus of different widths
and cannot have been set consecutively.
Accordingly, the slabs belong to a stairway of the palace complex, perhaps
to the main stairway, which evidently will have been somewhere E of the
Erechtheion. It may have led from the rock itself to the top of terrace III. It
is also possible that the stairway was within the palace, facilitating communi-
cation between the storeys, as at the palace of Pylos. 312 The relatively good
condition of the slabs would suggest this.
The conclusion that emerges from all the above is that there was indeed a
palace, that it stood in the area later occupied by the Archaic temple and the
Erechtheion, on terrace III, and that it had at least one grand stairway with
steps of sandstone. Anything else is pure conjecture.
310. Wace, BSA 25, 1921-23, p. 159 and pls A, K.

pl. II. 312. Blegen, AJA 1955, p. 35.
311. Mylonas, Praktika 1963, p. 102,
196
THE PELARGIKON
Below the top of the Acropolis rock was yet another part of the Myce-
naean fortification. Despite occasional efforts to protect it from continuous
encroachment and exploitation, it had remained outside the Classical fortifi-
cation walls, and was abandoned, deserted and fallen into disrepair with the
passage of time. The Pelargikon. There are few archaeological areas the
boundaries and extent of which have been as much disputed by various schol-
ars as the Pelargikon. The different opinions about its exact location are prac-
tically equal in number to the writers who have studied the question.
THE SOURCES AND THE BIBLIOGRAPHY
The ancient sources referring to the subject, listed in Appendix I to the

present study (n° 8 20-32), have been collected and published by various
writers. 313 Of all these, ten are really useful, those listed as numbers 20-29.
The others simply mention the Pelargikon without giving any information.
To begin with, a number of conclusions can be drawn from a study of the
sources. It is reported as being below the Acropolis, 314 on the rocks315 below
the N wall and next to the Erechtheion, 316 within it was the cave of Pan, 317 it
was situated on the Panathenaic Way between the Eleusinion and the Pythion,
at the level where the course of the Panathenaic ship carrying the peplos
ended, 318 and it was separate from the Asklepieion, the Areopagus, the tomb
313. Jahn-Michaelis, Atx Ath. p. 79, n. Av. 836.

to I. 14, Wide, Ausonia 1912, pp. 195-197, 316. Luc. Pesc. 47, 48.
Judeich, Top. p. 52, n. 5. On the relevant 317. Cratinos fr. 321 (Edmonds), and
traditions see Frazer, Paus. II p. 359. Wilamowitz, Hermes XIV, p. 183, Luc. Bis
314. Thuc. II 17. Acc. 9.
315. Aristoph. Av. 836, Schol. Aristoph. 318. Philostr. Vit. Soph. II, a, V.
197
THE THIRD PHASE
of Talas, the Anakeion, the Pythion and the Eleusinion. 319 As for its form and
condition, it was deserted, not enclosed and therefore habitable in the hour
of need by numerous families,320 full of stones, earth and weeds321 and it con-
tained sanctuaries and altars. 322
The only location that combines all these characteristics is the rough little
level space where the caves are located, at the NW of the rock between the
NW descent and the Klepsydra. No one has doubted that it was included in
the Pelargikon. All but Leake 323 thought it was only part of a far more exten-
sive area. The interpretations proposed vary considerably in their details, but
as a whole they fit into a few categories. In the first, the Pelargikon comprises
an area that surrounds the entire Acropolis rock (Fig. 36) and it is either for-
tified or not. 324 According to the second, it takes in the Wand S slopes of the
Acropolis (Fig. 37) including the springs of the Asklepieion and the Klepsydra
and it is surrounded by a strong fortification wall. 325 Others too locate it on
the W slope of the Acropolis, including the Klepsydra within its boundaries, 326
while some confine it to the NW slope as a protection for the Klepsydra. 327
319. Luc. Pesc. 42, Thuc. II 17 (where Acropolis pp. 24-25, 30, Dorpfeld, AM 1911,
he distinguishes it specifically from the p. 72, Walter, Athen, Akropolis p. 15, Kera-
Eleusinion). mopoullos, PraktAkAth 1932, pp. 110-124,
320. Thuc. II 17. Ephemeris 1934/1935, pp. 89-92, 95-96, 98,
321. Dittenberger, Sylloge 3 n° 83, Luc. 105, 110-111, 113 (where he considers it to
Pesc. 48, Pollux Onom. VIII 101. be later), Dorpfeld, Alt Ath. I pp. 3, 9, 105,
322. Dittenberger, Sylloge 3 n° 83. Travlos, Ephemeris 1939/1941, pp. 59-62,
323. Topography pp. 309-315. IloA.EoO. pp. 21-22, 25-26, 204.
324. Curtius, SBBerlin 1884, pp. 499- 326. Beule, L 'Acropole p. 82, Dorpfeld,
512, Botticher, Akropolis pp. 56-61, Curtius, AM 1886, p. 163, Lolling, Topographie pp.
Stadtgesch. p. 47, Wide, Ausonia 1912, pp. 337-338, Koster, Pelargikon pp. 16-28, Kav-
177-197, where he compares it to the vadias, IIeoiaroetx~ /1.exmoA.oyia pp. 300-
Roman pomerium. 301, figs 361-363, Pfuhl, BerlPhilolWoch
325. Unger, Sitzungsb. Akad. Miinchen 1911, pp. 299-307, Schede, Die Burg von
1874, pp. 303-306, Harrison, M. and M. pp. Athen p. 10, Keramopoullos, Deltion 1929,
41, 330, 350, 464, 536-538, Miller, AJA pp. 77-90, Picard, L 'Acropole I pp. 11, 18,
1893, pp. 484-488, Dorpfeld, RhM 1896, pp. Judeich, Top. pp. 116-117, 218, Berard,
131-132, Tsountas, /1xe6noA.t~ pp. 8-9, P. Stud. Rob. pp. 140-141, 150.
Kastriotes, MvrJµEta rwv /1.0rJVWV p. 28, 327. Frazer, Paus. II pp. 355-359, Bro-
Harrison, P.A. pp. 30, 32-35, D'Oodge, neer, AJA 1948, p. 112, Antiquity 1956, p. 10.
198
THE THIRD PHASE
Wilamowitz, who did not separate it from the rest of the fortification system,
extended it from the Areopagus to the gate of Hadrian. 328
A Pelargikon that extends beyond the boundaries of the NW slope of the
rock is not supported by the ancient sources. This view is based mainly on a
faulty conception of the topography of ancient Athens, on a number of ideas,
perhaps of value in themselves but with insufficient evidence, and on incor-
rect dating of existing material. The basic arguments of this school of thought
are as follows:
1) The ancient sources329 make it clear that the Pelargikon was near the
Eleusinion. The exact position of the Eleusinion, however, which was on the
Panathenaic Way NW of the Acropolis, has only recently been determined. 330
Prior to its discovery and identification, it had been located at various times
on the E part of the rock, 331 on the NE, 332 the W, 333 the SW, above the Odeion
of Herodes Attikos, 334 S of the Areopagus, 335 S of the Acropolis, 336 NW of
it,337 on the Pnyx,338 next to the Asklepieion, 339 W of the Odeion of Herodes
Attikos, 340 and finally E of the Areopagus or W of the Odeion of Herodes
Attikos. 341 Thus• scholars adapted their ideas about the Pelargikon to their
views of the subject.
328. Burg u. Stadt pp. 97-172. 335. Milchhofer, Topogr. p. 198, Lolling,
329. Dittenberger, Sylloge 3 n° 83, Phi- Topographie p. 317, n. 1, Weizsacker, Jahrb.
lostr. Vit. Soph. II, i, V. fii.r Philo/. 1887, pp. 618 ff.
330. Vanderpool, Hesperia XVIII, 1949, 336. Dorpfeld, AM 1894, pp. 508 ff. ,
pp. 128-137, Travlos, floJ..t:oO. p. 33, Thom- Pauly-Wissowa, RE V (1905), 2335 (Wachs-
pson, Hesperia XXXIX, 1960, pp. 334-338, muth).
fig. 3. 337. Judeich, Jahrb. fii.r Philo/. 1890,. p.
331. Leake, Topography pp. 296-301, 775, who changed his mind however, and
Botticher, Phil. Suppl. III, 1867, pp. 289 ff. Foucart, MemAclnscr 1900, p. 106.
332. Wachsmuth, Die Stadt Athen pp. 338. Svoronos, 'EQµT}vda uov µvT]µdwv
301 ff. , Wilamowitz, Aus Kydathen (1880) p. 'tOU 'EA.EUOLVLax.ou µuo"tT}QLax.ou X'UX.AOU ,
128. JoumlntemArcheolNumism 1901, pp. 420 ff.
333. Rangabes, Bulletino dell'Instituto di 339. Versakes, Ephemeris 1912, pp. 48 ff.
Corrispondenza Archeologica 1850, p. 136. 340. Judeich, Top. pp. 287-289.
334. Unger, Sitzungsb. Akad. Munchen 341. Keramopoullos, PraktAkAth 1932,
1874, pp. 282 ff. pp. 119-120, Ephemeris 1934/1935, p. 113.
199
~
A k ropol1s
N
~
0
0 §
"'
~
tTJ
~
·~
~
I~
~
~
Fig. 36. The Pelargikon as the area around the Acropolis, according to Botticher (Akropolis p. 58, fig. 7).
.
, ,,.;r -
,~~,' !
,'/ '
,','
, ,
I '
.
I
I
r·---------
i Altar?
,,,,,'.,
,,
Hieron of GE. ,,,,
& Demeter
Chloe
a ,.....
' ... -_, , ,
£ \_.. ,,,"'S
G [ ~
\"' tTJ
...
"I'
JN /\ 0 0 /
Tomb of
Kalos
~
__ .-- ______________ _
'(
--------------. ~
N ':\
G!.. ---· §
........
e
0 , , ,""', ...
':~ ......
....... ;\r , , , ,.
"""" , ,' , , ,"
~~=fa----
=:Jr.Pandemos ,..
I•
,
,, , .'
-- ""d
~tTJ
/
.: .' ,
---c::.-=---::.:-::::-::J
0 Or.ofl 1 ~ E
Mon~ of Themis
!iippolytos
J •••••••.•••••••••••
--- - ... . --- ·· · ------- '=.
-
Ro11te of Pausan ias ~
A .Te mpie of Ash /epios
B. Alto r
C. Krene PLAN OF
D.Portico
E . Precinct Wall SOUTH SI DE OF ACROPOLIS -------- ---- -----
H. Priests' Dwellin!JS ?
G . Boundnry Wall of Asklepieion ............... -
Metres ---- ··------- . · -
N. Foundations of Ni kins Monument n 10 :?O 30 40 .'\O too zc;o
Fig. 37. The Pelargikon as the area Wand S of the Acropolis, according to Harrison (M. and M. fig. 35).
THE THIRD PHASE
2) Thucydides342 tells us that the polis of Athens was initially on the Acrop-
olis itself and in the area toward the S. From this was formulated the syllo-
gism that since the city was S of the Acropolis, and since as other ancient
cities it will have had a fortification wall, the Pelargikon was this fortification
and it will therefore have been to the S. 343 The syllogism is based on the con-
viction that the Mycenaean cities were walled in the same way as were those
of historical times. This, as we now know, is not the case. The cities and set-
tlements of Mycenaean times were not walled and, as K6ster344 rightly
observed, what Thucydides is talking about has nothing to do with the
Pelargikon.
3) The fortress of the Acropolis cannot have been without a water supply.
Yet there is no water on the rock. Therefore the fortified section below the
rock, the Pelargikon, must have included the springs of the Asklepieion and
Klepsydra, or only Klepsydra, 345 the rock-cut steps of which may be compared
to those of the Perseia at Mycenae.346
This was the most serious of all the arguments, and it was determinative
in locating a Pelargikon below the slopes of the rock. Yet not only the sources
but the facts themselves show it to be wrong, especially after the discovery of
the North Fountain.
To begin with, the ancient sources tell us that in historical times the Acrop-
olis had no spring. Kylon and his fellow conspirators, Aristion later on and
finally the Turks all surrendered because of thirst. The Peisitratids, likewise
besieged in the Acropolis, had supplied themselves with water, which would
have been quite unnecessary had there been a spring. Research has shown
that the vein of Klepsydra had not yet been found in Mycenaean times and
342. II 15. Miller, AJA 1893, p. 488, Dorpfeld, RhM

343. Dorpfeld, RhM 1896, pp. 131-132, 1896, p. 131, Tsountas, 'Axe6.7wA.t~ p. 9, Har-
Harrison, M. and M. pp. 537-538, P.A. pp. rison, P.A. p. 35, D'Oodge, Acropolis p. 24,
30-31, Keramopoullos, PraktAkAth 1932, pp. Picard, L 'Acropole I p. 11, Judeich, Top.
114-115, Ephemeris 1934/1935, p. 89, where, pp. 117, 191, Keramopoullos, Ephemeris
however, he considers it to be Archaic. 1934/1935, p. 90, Dorpfeld, Alt Ath. I p. 3,
344. Pelargikon p. 41. Berard, Stud. Rob. p. 141, Travlos, lloA.t:ob.
345. Botticher, Akropolis p. 59, Lolling, p. 26.
Topographie p. 338, Curtius, Stadtgesch. p. 47, 346. Curtius, Stadtgesch. p. 47.
202
THE THIRD PHASE
was still unknown. Later on, not only was it left outside the fortification walls
of the Acropolis but there was no communication with it. This was arranged
only in Roman times, to which the stairway belongs. It was included within
the fortification for the first and only time during the War of Independence,
in the bastion of Androutsos. 347
During the Mycenaean period, when the Pelargikon was built, the inhabit-
ants of the Acropolis knew that the area had water, since they continued to
dig wells there. Yet the very existence of the wells is an indication that the
source itself had not yet been discovered. The only traces of Mycenaean activ-
ity in its immediate vicinity are two rectangular excavations, contemporary
with the N spring, that served as dumps. 348 For the same reason the North
Fountain was constructed with great effort and care. It would indeed have been
superfluous if there had been a source of water within the Acropolis walls.
4) Various walls found between the Asklepieion and the Odeion of

Herodes Attikos were thought to be Mycenaean and were associated with the
Pelargikon (Fig. 38), as its remnants. 349 These walls, however, are Archaic
rather than Mycenaean, a fact that soon became evident. Yet so firmly
implanted was the idea that the Pelargikon included this area, that even when
it became clear that the walls were later, they were explained as successors of
an earlier Mycenaean wall,350 or else the Pelargikon as a whole was consid-
ered as Archaic. 351
347. Parsons, Hesperia XII, 1943, pp. not a trace of a Mycenaean wall has been
191-267. found in the vicinity. Moreover, it would
348. Ibid. p. 206. have been strange indeed for the Athenians
349. Botticher, Akropolis p. 57, Miller, of the Archaic period, while respecting the
AJA 1893, pp. 485-486, fig. 1, Harrison, PA. fortification of the rock as a whole, to build
p. 35, D'Oodge, Acropolis p. 25, Judeich, anew along this line of the wall, carefully
Top. p. 116-117, Keramopoullos, PraktAkAth removing every trace of an earlier wall,
1932, pp. 114-115, Ephemeris 1934/1935, pp. assuming that it ever existed. By analogy
95-96, 98, Dorpfeld, Alt Ath. I pp. 3, 105, with the rest of the wall, they would have
Travlos, Ephemeris 1939/1941, pp. 59-62. preserved what there was, making use of it
350. D'Oodge, Acropolis p. 25, Travlos, for its greater extent.
lloAEoO. pp. 25-26 and n. 1. There is no spe- 351. Keramopoullos, PraktAkAth 1932,
cific evidence to support this view, which is pp. 114-115, Ephemeris 1934/1935, pp. 89, 98.
completely counter to fact. To begin with,
203
THE THIRD PHASE
'
\
\
----~
- -·--·- - -- -·--- - ·-
Fig. 38. The Pelargikon on the basis of the Archaic walls SW of the Acropolis,
according to Miller (AJA 1893, p. 489, fig. 1).
5) Cleidemos (mid-4th century B.C.), and Cleidemos alone, reports that

the builders of the old wall .. . TIEpH~~aAov ... EVVECxTIUAOV TO neAacry1K6v. 352
This is the only case in which the Pelargikon is equated with the Enneapylon.
Moreover, the Enneapylon is nowhere else referred to by this name. Pole-
mon353 alone, mentioning the sanctuary of the hero Hesychos, says that it is
near the Kyloneion, EKTOS Twv £vvEa TivAwv. It was generally accepted by
recent scholars that there was in fact an Enneapylon and efforts were con-
centrated on locating the nine gates in a logical series with a semblance of
reality. Basically, three theories were formulated. According to the first,354 the
gates were arranged along the length of the wall at the top of the rock. The
352. Bekker, Anecd. Gr. I p. 419, l. 27, 354. Unger, Sitzungsb. Akad. Munchen
Souida, s.v. Cx1TE5a and T)TIE5i~ov. 1874, pp. 330-333, Belger, BerlPhilolWoch
353. Muller, Fr. Hist. Gr. III p. 131, fr. 1894, pp. 46 ff.
49. See also Schol. Oed. Col. 489.
204
THE THIRD PHASE
second355 held that the gates belonged to a hypothetical lower wall, which
enclosed the upper wall and ran around the slopes of the rock. According to
the third and most prevailing theory, 356 the gates were placed at various points
in an extended fortification, concurring with the Pelargikon, and located gen-
erally toward the W or SW of the rock. In addition, with the exception of Kera-
mopoullos,357 all subscribed to the view that the Enneapylon was Mycenaean,
most believing that the number of gates recorded was correct. 358
Thus the Acropolis fortification must have had nine gates. Yet we have
seen that the wall enclosing the top of the rock had far fewer. Therefore the
rest, or all nine, must have been in the extension of the fortification (see Fig.
3) below the rock, which implies a Pelargikon covering a very large area
indeed, if the gates are to be arranged in a practical way. 359
To begin with, it is very doubtful that the term "Enneapylon" is to be taken
literally in the sense that there were actually nine gates leading to the Acrop-
olis. 360 Secondly, such a construction, extensive in area, would be redundant
and against all principles of the art of fortification. It would take the fortifi-
cation dangerously close to the Areopagus, which by its very nature was the
perfect place from which to launch any kind of attack. The wall of the Myce-
naean Acropolis served its purpose well without this extension of doubtful use.
Moreover, it does not agree with Mycenaean norms of fortification. We have
only to imagine Mycenae or Tiryns with such an enormous extension to realise
355. E. Curtius, Gesammelte Abhand- pp. 11-12, Travlos, IloA.EoO. p. 26 (and n.

lungen I p. 442, Stadtgesch. p. 47, fig. 13, 2), p. 204, fig. 7, pl. I.
Botticher, Akropolis p. 57, fig. 7. 357. Ephemeris 1934/1935, pp. 88-89.
356. Beule, L 'Acropole p. 82, Bohn, 358. On this score, doubts are raised by
Prop. p. 3, Harrison, M. and M. p. 41, Mil- Tsountas ('Axe6.7wA.t~ p. 9), Picard (L 'Acro-
ler, AJA 1893, p. 484, fig. 1, Tsountas, pole I p. 11) and Broneer (Antiquity 1956,
'Axe6.7roA.t~ pp. 8-9, Harrison, P.A. p. 32, p. 12).
Koster, Pelargikon pp. 22-23, Frazer, Paus. 359. Botticher, Akropolis p. 57, Lolling,
II p. 356, Schede, Die Burg van Athen p. 10, Topographie p. 338, Miller, AJA 1893, p.
Lolling, Topographie p. 338, Walter, Athen, 484, Harrison, PA. p. 32, Travlos, IloA. EoO.
Akropolis p. 15, Judeich, Top. pp. 115-116, pp. 25-26.
Keramopoullos, Ephemeris 1934/1935, pp. 360. See P. Kastriotes, Mvr;µEia rmv
88-89, Welter, AA 1939, pp. 8-9, Berard, 'AOr;vmv p. 29, and Broneer, Antiquity 1956,
Stud. Rob. p. 140, Broneer, Antiquity 1956, p. 12.
205
THE THIRD PHASE
how foreign this sort of arrangement would be to the Athenian fortification

system. An arrangement of this sort is ruled out entirely by actual Mycenaean
practice361 and specifically by the circumstances prevailing in the Athenian
Acropolis. As we have seen, when the Acropolis wall was built, no gates were
added and the existing NE approach was even given up. It is quite clear that,
as Keramopoullos362 surmised, the Enneapylon was not Mycenaean. The con-
siderably later, isolated information given by Cleidemos refers not to this old
wall, which in any case left no trace at all, but to later constructions that were
not for purposes of fortification but belonged to the Archaic organisation of
the west approach, many remains of which were found and attributed to the
Pelargikon.
This later organisation lies beyond the limits of the present study. It would
take us into another era, a time when the situation in Athens had radically
changed. The Acropolis was no longer the fortified seat of an hereditary
dynast and the refuge of his subjects in time of danger. The administrative
centre now lay outside the fortress, which had itself been dedicated to the
gods. Just when this change occurred cannot be determined with precision.
Yet the sherds from the fill of the N Fountain suggest a number of hypotheses.
Indeed, as the latest sherds demonstrate, the spring continued in use as a
dump, while the Acropolis itself was inhabited as always down into protogeo-
metric times. 363 At this same time, or slightly earlier, the centre of activity
will have begun its move down to the lower city, as is evident from the Sub-
mycenaean graves in the Kerameikos. 364 Certainly by the beginning of Geo-
metric times, the transformation of the Acropolis from an administrative to a
cult centre had already been completed. This will no doubt have played a part
in the fortunes of the fortification wall, which will have lost some of its sig-
361. The citadel of Mycenae has two x~ 'EA.A.a~ p. 57), has four (Ergon 1958, fig.
gates and an additional small sally port. 48).
Tiryns has three successive gates in the east 362. Ephemeris 1934/1935, pp. 88-89.
side and two secondary ones in the west. 363. Broneer, Hesperia VIII, 1939, p.
Glas, with its huge area and the protection 427, Antiquity 1956, p. 13.
afforded by neighbouring communities (see 364. Kraiker - Kubler, Kerameikos I (Ber-
Marinatos - Hirmer, KQ~T'YJ x a i M v xrJva"i- lin 1939).
206
THE THIRD PHASE
nificance in terms of protection. Perhaps also the Panathenaia, held from then
on, will have played a part, requiring an easier and more accommodating
approach for the crowds.
During his restoration of the temple and bastion of Athena Nike, Balanos
discovered a small rectangular eschara (sacrificial hearth) beneath the Classi-
cal temple in the fill within the bastion. Cut out of the rock in the middle of
the eschara was a rounded hollow co~taining many small clay figurines, 365 dat-
able to Submycenaean times. 366 The eschara succeeded the W niche of the
bastion, which had by now been abandoned as a focus of cult practice. Indeed
some of the votives may have been tranferred here from the earlier niche.
Later, in Archaic times, a little shrine was built around the venerable eschara.
This was the predecessor of the Classical temple and to the east of it stood
the altar. 367 Contemporary with this construction and at about the same level,
an Archaic polygonal crowning wall was added around the walls of the bas-
tion. The finding of two pieces of Karra stone in this crowning wall appear to
date it to Peisistratid times, thus contemporary with the altar. 368
The level of the eschara makes it clear that at some point, probably early
in the Geometric period, the upper part of the bastion was taken down to
somewhat lower than the Classical level. The bastion was thus transformed
into a terrace where the chthonic cult of the goddess with the pomegranate
was now established, the W niche having been given up. 369 At the same time,
the entrance, now without a bastion and thus without the outer gate, com-
prised only the inner gate which may have been widened. Subsequently, in
Archaic times, the top of the truncated bastion, which was uneven because
365. Balanos, Ephemeris 1937 r , p. 785. finds from the household shrine at Asine
366. The figurines have occasionally (see Asine p. 299, fig. 206, lower right, and
been described as Archaic (Balanos, Ephe- pp. 308-309, fig. 212).
meris 1937 r, p. 785, BCH 1936, p. 455), 367. Balanos, Ephemeris 1937 r, pp. 785
also as Mycenaean (BCH 1938, pl. L, B) ff. , Oikonomos, Ephemeris 1939/1941, p. 105.
and Submycenaean (BCH 1939, p. 289). 368. Welter, AM 1923, p. 193, AA 1939,
They have been dated as Submycenaean by pp. 9, 14, Oikonomos, Ephemeris 1939/1941,
Professor Marinatos, who has seen them p. 105.
and describes them as similar to two of the 369. Ibid. pp. 97-110.
207
THE THIRD PHASE
the upper courses had been removed, was smoothed and embellished by the
addition of a polygonal crowning wall, which was level on top. 370 Here was
built the shrine and altar, completing the transformation of the bastion into
a temenos.
These changes in the plan of the entrance entailed, of course, corre-
sponding changes in the arrangement of the approach. The Mycenaean stair-
way was no longer sufficient for the crowds that now ascended in the Pana-
thenaic procession. A wide and ample approach had to be constructed. A basic
part of it is undoubtedly the big polygonal supporting wall, the most impor-
tant piece of which is preserved today between the Propylaia and the Beule
Gate. 371 The extension toward the W was discovered and identified correctly
by Keramopoullos. 372 The wall, connected by some with the Pelargikon,373 is
in fact the supporting wall of an Archaic ramp leading to the Acropolis
entrance, built perhaps during the time of Peisistratos. Comparable, if not
contemporary, are the remains of various similar walls W of the Beule Gate
and on the SW slope of the Acropolis. Although he connects them with the
Pelargikon, Keramopoullos identifies them as Archaic.374
Stated simply without going into detail, it is clear that during Geometric
and Archaic times the entrance to the Acropolis was changed in form and that
various walls were built to the SW and W of it. These served no defensive
purpose whatsoever. They were designed, instead, to facilitate access to the
Rock. Taken as a whole, these outworks form a complicated complex of ter-
races and approaches that have yet to be studied. They appear to compose
what was known as the Enneapylon. 375 It may even be that the ancient Greeks
of Classical times themselves, and Cleidemos even more so, connected the
370. Welter, AM 1923, p. 193. 374. Ephemeris 1934/1935, pp. 87-105,

371. Bohn, Prop. p. 15 and pl. II, Kav- pl. 1.
vadias-Kawerau p. 129, pl. H. 375. This complex, which continued to
372. Deltion 1929, p. 74, n. 3, Ephemeris function almost to his time, is what Hero-
1934/1935, p. 87, pl. 1. dotos (VIII 53) is referring to in speaking
373. Beule, L 'Acropole p. 84, Bohn, Prop. of the "gates" and the "ascent" to the
p. 15, Dorpfeld, AM 1885, pl. II, Miller, AJA Acropolis.
1893, p. 486.
208
THE THIRD PHASE
walls, by then in ruins and partly buried, with still earlier Mycenaean times;
whence the attribution of the Enneapylon to the Pelargikon.
6) During the Peloponnesian War the Pelargikon was used as a place where
refugees could live. Despite the oracle, moreover, it was in continuous use by
various people as a source of earth and stone and even herbs. Still others
placed altars and performed cult rites in the area, quite arbitrarily. This con-
tinued to such an extent that laws had to be imposed time and again. 376 Thus
it must have been extensive, its addition increasing the area of the Acropolis
substantially. 377 These arguments do not hold up. Thucydides says that the
area was inhabited as were other vacant parts of the city and whatever sanc-
tuaries and heroa had been left unfenced. This in no way suggests that thou-
sands of refugees were settled in the Pelargikon, nor does it give any signifi-
cant information about its size. Likewise the poaching of various materials is
quite unrelated to the size of the place where this was being done. The pri-
vate altars and cult places established there without control will have occu-
pied only a few square metres precisely because they were private affairs, with-
out the assistance of the city, and thus illegal.
Let us recapitulate. The entire discussion about the Pelargikon, as found

in the existing bibliography, is a series of efforts to find arguments support-
ing opinions that are presented as if they were self-evident truths with the
force of dogma. Keramopoullos was the most systematic and careful re-
searcher of the subject. He recognised the material as later. Yet since he
could not bring himself to relinquish the idea that the Pelargikon extended W
and S of the Acropolis, he concluded that the construction in its entirety was
Archaic. He thus left the main question about the location and extent of the
Mycenaean Pelargikon unanswered. Koster began, and to an extent contin-
ued, on the basis of concrete observations. Yet he resorts to conclusions that
are entirely arbitrary and his plan shows the boundaries of an extensive area
376. Thuc. II 17, Dittenberger, Sylloge 3 ramopoullos, Ephemeris 1934/1935, pp. 110-
n° 83, Pollux Onom. VIII 101. · 111, where he corrects the «KaTa irAeov»
377. Botticher, Akropolis p. 57, Curtius, of Pollux to «KaTa ir11Aov» and suggests
Stadtgesch. p. 47, where it is considered a that the place was used as a source of clay.
pasture, Koster, Pelargikon pp. 16-17, Ke-
209
THE THIRD PHASE
(see Fig. 1), with the sole argument that it must have received additions.
Research, however, has shown that there was no reason for the Mycenaean
fortification to include the KJepsydra, which was at that time unknown. More-
over the exact position of the Eleusinion has now been discovered. Clearly,
there is absolutely no evidence that the Pelargikon was any larger than stated
by the ancient sources.
The construction of a complicated and extensive work of fortification to the
W or NW of the Acropolis (to say nothing of the S side), within which would
be the bastion and the main entrance, would be entirely counter both to what
is known about the form of Mycenaean fortification systems and to basic rules
of the art of designing fortifications. For this would bring the first line not only
closer to the Areopagus but, far worse, to a lower level than the hill, placing
the defenders of the fort in a hopelessly disadvantageous position.
Rather than resorting to theories, it is better to see to what extent the var-
ious ancient sources that locate the Pelargikon in the area of the caves agree
with the results of exploration on the spot.
THE REMAINS OF THE PELARGIKON IN SITU
Rising above the point at which the NW descent begins, is the rock that
fissured off from the main rock mass, as we have mentioned above. It forms
a sort of natural fortification wall projecting toward the W and protecting the
first metres of the descent (Plan 36, 1, and see Fig. 4). The top of this rock
has been roughly worked to form a flat surface of about 1,50x 1,50 m., which
bears all the hallmarks of Mycenaean workmanship (Fig. 39). Further down,
at various places on its S face, there are traces of similar cuttings, some hori-
zontal and some obliquely vertical (Plan 36, 1). Within the fissure between
the main rock and the projecting split-off, a number of large stones are
wedged in at different levels. They are not shown on the plan because the
rock conceals them. A series of cuttings have been made in the main rock
mass, near the base of the projecting split-off and just to the S of it. They are
more or less straight and cut with great care, parallel to each other and
stepped, forming repeated small horizontal levels. In one of the most regular
of these cuttings, three slabs are preserved in situ in a row (Plan 36, 2 and Fig.
40) . On the same line, further east (Plan 36, 3), lies another, larger stone, this
210
THE THIRD PHASE
too, in situ. W of all these, at 4, are three more stones, one on top of the other
and in line with the previous ones. Both the stones and the cuttings around
them were noticed by Koster. 378 Since he did not see the tooling at the top
or the cuttings in the S wall of the split-off rock, he interpreted them as isol-
ated stones belonging to the supporting wall of the start of the descent. This,
however, can be ruled out, quite apart from whether or not there are cuttings
on the projecting split-off. Although on the plan the larger stone appears to
be next to the previous ones, the differences in level are actually so great that
such a supporting wall would have to have been very thick in order to reach
as far as the stones. In terms of the construction it was supporting, this would
have been excessive. The stone at 5 on Plan 36, as already noted, belongs to
this construction (see also Plan 15, 2).
West of these traces, running obliquely in relation to the brow of the plateau
of the Rock, is a large area, the surface of which has been smoothed by cut-
ting projections where needed and by filling in existing hollows with small
stones (Plan 36, 6). The surface gives every indication of having been prepared
for the foundation of a fortification wall, resembling what we have seen at
other places on the Acropolis. The S line of this worked surface leads to a
point some 13 m. W of the end of the high projection of the rock. From this
point on toward the W, the rock has not simply been smoothed. Since it rises
toward the S, it has been cut to some depth, thus forming a deep cutting that
is straight and runs obliquely in relation to the edge of the rock, which it meets
some 6 m. W of its commencement. The cutting (Plan 36, 7), which increases
steadily in depth toward the W, has fairly even sides, rounded corners and
surfaces that are not particularly smooth. It has, in other words, the charac-
teristics associated with Mycenaean workmanship as we know it (Fig. 41).
The course of the wall to this point is clear: the interior face, that to the
S, follows the line leading from stone 3 to the cutting at 2 and the stones at
4. From there on, it coincides with the S boundary of the area 6 and the cut-
ting at 7. At this point it descends toward the base of the rock and continues
along it, as we shall see. The N line of the wall, founded on the worked sur-
face at 1 on the top of the protruding rock that had split off, runs W leaving
the N side of the protruding rock outside. It continues as far as the main mass
378. Pelargikon pp. 12-13.
211
0
~
0 5 10
M
Plan 36. The Pelargikon.

THE THIRD PHASE
Fig. 39. The top and S face of the projection of the rock, from above.
Fig. 40. Cuttings and stones of the Pelargikon in situ, from the W (Plan 36, 2).
213
THE THIRD PHASE
Fig. 41. The oblique cutting at 7 on Plan 36, from the W Below right, later
working of the rock.
of the rock and follows the N line of the area at 6 to the point where that dis-
appears at the brow. Here it too descends to the foot of the rock, and con-
tinues toward the W, parallel to the inner line. Given these two lines, the
width of the wall at this place varies between 4,80 and 4 m.
A pile of enormous stones rests at the foot of the rock NW of 7, at 8 on
Plan 36, which cannot possibly be the remains of the lower courses of the for-
tification wall. Kavvadias likewise had observed this,379 and thought that they
belonged to a sloping ramp of some sort connecting the plateau of the caves
379. Ephemeris 1897, p. 28, pl. l~ .
214
THE THIRD PHASE
with the lower city. Yet neither one nor the other holds. In the course of clear-
ing them, it became evident that they were not founded on the rock but lay
on fill containing sherds from the time of the Turkish occupation.
A fairly large piece of the wall is preserved just W of 8, almost touching
the rock and separated from it by a narrow fissure, still containing a number
of stones (Plan 36, 9). On its N side, some 3,50 m. below its top, there is a
stretch of the rock that is worked horizontally to varying widths. It has been
worked smooth, without corners and there is no trace of tooling. The line con-
tinues that of the worked rock surface at 7. This level unquestionably gives us
the inner line of the wall, which at this point curves around this piece of the
rock to continue toward the W. No trace remains of the outer line of the wall
on the soft, crumbly schist at the base of the rock and it is therefore impos-
sible to determine the width of the wall. The outer line can only be recon-
structed as parallel to the inner, estimating a width of at least 3-3,50 m. While
there is no need at this point for the security of a very wide wall, the width
must be sufficient to allow its defenders to circulate.
Above 9, partly on the top of the section of the rock separated by the fis-
sure and partly on the main mass, there are two long, narrow cuttings run-
ning E-W (Plan 36, 10), which are parallel neither to the Mycenaean remains
nor to each other. 380 While they have clearly been made for wall foundations,
they are later. Their surfaces are smooth, they have regular corners and the
sides of the cuttings are straight (see Fig. 41, lower right). Moreover they pre-
serve clear traces of the metal tools used in the cutting. The difference
between these and the Mycenaean worked surfaces at 7 is clear and easily
understood. Thus there was a wall at some time in this place, Archaic or per-
haps later, but in any case not Mycenaean.
As we have seen, the fortification wall along the line 7-8-9 ran from the
brow down to the base of the rock. The reason is clear enough from the plan
but even clearer on the ground. From this point westward begin the caves of
Pan (A on Plan 36), Apollo (B) and Zeus Olympios (r). 381 The distance
380. The cuttings were noted by Kavva- 381. See Kavvadias, Ephemeris 1897, pp.
dias (Ephemeris 1897, pl. 1, next to B) and 1-24, and Keramopoullos, Deltion 1929, p.
by Parsons (Hesperia XII, 1943, pp. 227-228 89.
and fig. 36, 3).
215
THE THIRD PHASE
between the various cave mouths themselves and to the brow of the rock is
so small that it would have been impossible to build a fortification wall in this
place without blocking the space completely and making any sort of circula-
tion impossible. Yet since the area had to be fortified, the only way was to set
the wall along the base of the rock. Apart from the fact that the wall could
not have run along the top of the rock, there is other evidence for the course
of the wall further W that is connected with its exterior line.
A circular well that was full of sherds 382 was revealed in the excavations of
the American School of Classical Studies some 12 m. W of 9, and about 5 m.
N of the vertical wall of the rock, at 11 on Plan 36. From bottom to top of
the well the sherds are closely similar. Except for a few earlier pieces, they
belong to the latest years of the LH IIIB and the first years of the LH IIIC
periods. The fill at the top had been disturbed and mixed in with the earlier
fill were a few fragments of a Roman lamp. Clearly the well was abandoned
and filled in practically to the top by pottery of the period in which, as we
shall see, the fortification wall had just been constructed and the N spring put
into use. The disturbance of the upper layers of the well and the presence of
Roman sherds show that the mouth had remained open and uncovered for a
long time, perhaps centuries.
Two rectangular pits containing pottery of the same period, LH IIIB-C,
were found further W, at 12 on Plan 36, beneath the Classical slab pavement
identified by the American excavators as the courtyard of the Pythion. 383
The first conclusion to be drawn is that the inhabitants of the area aban-
doned their wells when the N spring came into use. This appears to coincide
approximately with the time when the little houses built on the pathway of
382. The well, together with its con- excavation as Section OA, n° 19 and the
tents, is still unpublished. The Director of contents are characterised as Submyce-
the Agora Excavations, H. Thompson, per- naean. In the present publication, I have
mitted me to consult the excavation day recorded my personal observations after
books referring to the well and to examine examination of the sherds from the well.
its contents. The architect, J. Travlos, pla- 383. Shear, Hesperia X, 1941, p. 7 and
ced at my disposal his plans of the area. fig. 7, Parsons, Hesperia XII, 1943, p. 206.
The well is reported in the day book of the
216
THE THIRD PHASE
the NE approach were abandoned. 384 This means either that the wells fell into
disuse because the opening of the N Fountain made them redundant, or that
both wells and houses were abandoned for reasons of security. If the wells
were inside the fortification wall, there would have been no reason to aban-
don them. Quite otherwise, every effort would have been made to keep them
in repair and, most likely, to open others. Thus, the well at 11 and the pits at
12 on Plan 36 were outside the wall, which cannot have run N of them. If the
wall had been set on top of them, they would have been filled with the stones
used in preparing the ground for the foundations. Neither stones nor compact
filling were found and, as we have seen, the disturbance at the top of the well
at 11, together with the homogeneity of the fill just below, show that the well
had remained visible. Therefore the outer line of the wall ran S of the pits,
between them and the vertical face of the rock. This gives us the approximate
width of the wall, which, for reasons of security, cannot have come close to
the edge. The width of 3-3,50 m. that we noted further E will have been
retained for these reasons.
Just to the SW of 12 is the Klepsydra. At the spring site where the ancient
fountain house stood, apart from the various additions, there are two small
rocks. They are next to each other at the foot of the NW plateau. In antiquity
they were joined, forming a little cave in which the spring flowed. At some
time in the lst century B.C., the roof of the cave collapsed leaving the rocks
separated.
During the period we are discussing, there was only the small, low cave
(Plan 36, 13). It appears not to have been completely connected to the pre-
cipitous rock behind it, but separated from it by a narrow irregular fissure. A
number of boulders are preserved today in the lower part of the fissure,
well wedged into their places and incorporated in the wall of the time of the
War of Independence. We cannot know with certainty if these stones were
there from the beginning and the wall was simply laid over them, or if they
were found elsewhere and brought in as building material. Be that as it may,
the wall, as we have followed it to 12, runs directly over the fissure and the
little cave. This is evident from its next trace, which is just above and SW of
384. Broneer, Hesperia II, 1933, pp. 365-367, 372, figs 37-45.
217
THE THIRD PHASE
Fig. 42. The rock above Klepsydra, from the E. Discernible in the middle of the
photograph is the cutting 14 of Plan 36.
Klepsydra, on the edge of the plateau at 14 on Plan 36. The trace consists of
a curved cutting, sloping and oriented obliquely toward the S (Fig. 42). It
follows a line on the rock conforming with the Wend of cave A. 385 To the W
of the cave the rock rises abruptly forming a narrow spine from NW to SE.
It then drops off precipitously to the W, at the point where the Roman steps
385. Stepped cuttings at various places fore not shown on the plan (see Kavvadias,
around and at the mouth of the cave are Ephemeris 1897, pl. 1, al , a2, a3).
clearly later, probably Classical and there-
218
THE THIRD PHASE
were cut connecting Klepsydra with the W entrance to the Acropolis. The rock
configuration is such as to prevent any access to the area west of the Propy-
laia from cave A. The cutting at 14 is at a low level, at the beginning of the
spine of the rock. It is unquestionably Mycenaean work, precisely of the sort
needed for founding the wall on the abrupt and slippery edge of the rock. It
coincides with the line of the interior face. Up to this point the wall was built
along the foot of the rock and adapted to the rock face. It now turns SE and
climbs up the rock again, making use of the high, narrow spine, which in a
way forms its natural continuation, just as the jutting rock at 1 (Plan 36) was
utilised. Indeed this is verified by the cuttings on the sharp peak of the spine
at 15, Plan 36. The cuttings, short and not very deep, with rounded edges and
corners and surfaces summarily smoothed, are stepped precisely in order to
hold the big stones of the fortification wall so that they are securely founded
on the rock. They are just where you would expect to find such cuttings. They
show the course of the wall, which is in any case the only course possible given
the rock formation. The Pelargikon continues in this direction, rising continu-
ously and, after a course of a few metres, meeting the stretch of the Cyclo-
pean wall that corresponds to the NW corner of the Pinakotheke, as we have
already seen.
West of this last section of the wall and further down, the rock juts out at
an angle, descending toward the NW. On this projection there are four stepped
cuttings that run SE in a direction parallel to it (Plan 36, 16). The surface of
the rock above the first cutting was trimmed in Classical times for the founda-
tion of the fortification wall W of the Pinakotheke. Below the lowest cut-
ting, the rock formation has been changed considerably in connection with the
construction of the Klepsydra and even more so for the foundations of the bas-
tion of Androutsos. As a result of all this, the continuation of the steps upwards
and downwards no longer exists. The approach is clearly pre-Classical since it
is cut off at the SE by the Classical fortification wall. Both workmanship and
proportions of the rock-cut steps are moreover indubitably Mycenaean. There
can be no doubt that these are steps and they were in use for a long time as
can be seen from the surface of the treads, which has been worn smooth. There
appears to have been an access here leading from the level of the wells to the
west entrance of the Acropolis, along the Cyclopean wall.
Thus we have the full extent of the Pelargikon, based on the evidence pro-
vided by the area itself. The first section of the course of the wall, represent-
219
THE THIRD PHASE
ing about 1/3 of the total, has left traces that are incontestable and easily dis-
cernible (Plan 36, 1-9). Similar traces remain from the last metres as well (Plan
36, 14-15). The middle section has left no traces at all. Yet the evidence pro-
vided by the spring and the pits is such that the line as a whole can be cal-
culated precisely.
The boundaries and extent thus determined agree perfectly with the infor-
mation in the ancient sources. Indeed, the area is below the Acropolis of Clas-
sical times and it is not far from the Erechtheion. It is rocky and it includes
the cave of Pan, which is just above the fortification. The trireme carrying the
peplos followed the Panathenaic Way, reaching the level of the cave after cir-
cumventing the Eleusinion, went a little further along and stopped where the
gradient became very steep, beside the Pythion. As a unit the Pelargikon was
clearly distinct from the Asklepieion, the Areopagus, the tomb of Talos, the
Anakeion, the Eleusinion and the Pythion. Moreover it was large enough for
refugees to stay in it. We can well imagine it full of stones that had rolled
down from the Cyclopean wall of the Acropolis or had remained from its con-
struction, with all kinds of greens and weeds sprouting from amongst the
stones, in those days just as now. As for the sanctuaries and altars, Kavvadias'
excavation brought to light quite a few such constructions. The poorest and
most haphazard of these will, no doubt, have been erased by the passage of
time.
The Pelargikon as a space will evidently have been closed all around, com-
municating only with the top of the rock by way of the NW ascent. In this
chapter we have explained why an ascent or descent to the foot of the rock
is unlikely. In any case it was not needed. The area was clearly meant to be
used by the population as a refuge in times of danger or siege. Communica-
tion with the outside was quite unnecessary and a gate would probably have
been dangerous. Contact with the Acropolis was assured by the NW ascent
and the top of the wall was secure for the defenders.
It was also, and perhaps principally, constructed for purposes of security,
specifically to close off and make the area of the caves inaccessible and unas-
sailable by attackers. Because of the caves, moreover, the area could be inhab-
ited (just as it was later on during the Peloponnesian War). The chief merit,
however, of this addition is that it fortified a part of the rock that was lower
down, thus open to enemy attack and with its caves would have made an excel-
lent base for an attacker who could have occupied it. The very formation of
220
THE THIRD PHASE
the rock demanded the fortification of the Pelargikon, yet the difference in
level prevented its inclusion within the main fortification wall. The solution of
fortifying the lower level created the ideal Unterburg, increased the security of
the fortification system as a whole, and provided at the same time an area
suitable for refugees.
There is no precise evidence for its date. The way in which it was joined
to the main fortification wall shows that it must have been built at the same
time. The evidence from the well at 11 (Plan 36) leads to the same conclu-
sion. There is no reason at all to believe that the Pelargikon was later than
the Mycenaean fortification wall, while there is every reason to believe that it
was contemporary.
221
THE SLOPES
Apart from a section of the north slope, the Acropolis slopes have never
been systematically explored. Thus what Mycenaean material has surfaced, is
limited and has usually come to light in the course of exploration carried out
for other aims and interests. Yet every excavation reveals new evidence show-
ing that the area as a whole was inhabited during Mycenaean times, so that
we are justified in assuming that the lack of evidence is likely to be due sim-
ply to lack of exploration.
From the west slope there is no material. The first evidence begins to sur-
face in the Klepsydra area, beneath the NW corner of the rock. Three new
wells were added to the older ones. They were dug in Mycenaean times and
were in use down to the end of the period. There are also the two pits that
were found beneath the Classical paving of the Pythion. 386 The excavations of
Broneer to the east of the Pelargikon showed that the area between the rock
and the peripatos to the N of it had been systematically inhabited. Yet except
for the large amount of pottery retrieved, no durable evidence for its use sur-
vived other than a few uncertain traces of Mycenaean houses, which in the
meantime had disappeared. 387 Scattered ceramic finds in the sanctuary of Eros
and Aphrodite, and the existence of the sloping pathway leading to it, earlier
even than Mycenaean show that the place was in use, and that it was indeed
a cult place during the years we are examining. 388 The first concrete evidence,
however, is to be found in the area of the NE ascent.
386. Shear, Hesperia IX, 1940, p. 297; 539-540, 559 etc., Broneer, AJA 1940, p.
X, 1941, p. 7, fig. 7, Parsons, Hesperia XII, 255.
1943, p. 206. 388. Broneer, Hesperia I, 1932, p. 35;
387. Morgan, Hesperia IV, 1935, pp. IV, 1935, p. 124.
189-213, Hansen, Hesperia VI, 1937, pp.
222
THE THIRD PHASE
After this pathway was abandoned, a few small, poor houses were built, as
we have seen, over the pathway itself and the steps. Because of the steep slope
of the rock and the poor construction of the houses, only remains of a few
walls and parts of floors survived (Plan 37, 1, 2, 3, 4, 5, 6); so few and in such
poor condition that it was imp~ssible to retrieve their original plans or even
to determine their exact boundaries. 389
The houses were part of a more extensive settlement390 and they were not
in use for very long. Indeed they were abandoned not long after they were
built and they were left in some haste, as it appears from the position of the
finds that were lying on the floors. 391 Some imminent threat of danger will no
doubt have been the cause. Yet the settlement remains show no signs of
destruction by force, so it appears the danger ultimately passed them by. Even
so, the abandonment of the settlement was final and the houses were never
occupied again. 392
The sherds from beneath the house floors and found in situ abandoned on
the floors give the dates of construction and of abandonment of the houses.
They provide, moreover, important evidence for the chronology of the final
phase of the Mycenaean Acropolis.
The east slope with its enormous cave393 has not to date yielded any evi-
dence. From the south slope, a few sherds have been collected from the area
of the Odeion of Pericles. 394 A hearth of LH IIIB-C times is reported to have
been found in the temenos of Dionysos and a fill of the same period in which
the later peribolos wall of the sanctuary was founded. To this should be added
a considerable amount of pottery found scattered at various spots nearby. 395
Similar traces were discovered likewise on the SW slope in the form of pot-
389. Broneer, Hesperia II, 1933, pp. 352- 393. Broneer, Hesperia II, 1933, pp. 415-
355, 365-367, 372; IV, 1935, pp. 111-112, 417, Broneer-Pease, Hesperia V, 1936, pp.
pl. I. 247-272.
390. Broneer, Hesperia IV, 1935, p. 113. 394. J. Travlos, Praktika 1951, p. 44 and
391. Broneer, Hesperia II, 1933, p. 355; fig. 3.
IV, 1935, p. 111. 395. Newspaper Kathemerini, 27 Febr.
392. Broneer, Hesperia II, 1933, p. 355, 1962.
Antiquity 1956, p. 12.
223
THE THIRD PHASE
0 •
. • • •• III •
.~• ·
0 .
: ', ' . · · • p • A • • • ' •
0 5 10 20 30 40
Plan 37. The remains of the little Mycenaean houses built over the NE ascent.
tery fragments recovered from disturbed levels S of the Odeion of Herodes

Attikos,396 and from two wells, one a few metres S of the Odeion of Herodes
Attikos, the other just behind the rear wall of the Stoa of Eumenes. 397 Both
contained late Mycenaean sherds, evidence that the area was not without
inhabitants in that time.
396. Ergon 1956, p. 7. 1959 by the Director of the Acropolis, J.

397. I am indebted to the excavators for Meliades, the second by E. Fiandra of the
this information, which has not yet been Italian Archaeological School.
published. The first well was excavated in
224
THE THIRD PHASE AS A WHOLE
The Acropolis at the end of the Mycenaean period presented much the
same picture as that seen in other contemporary citadels. The fortification wall
surrounded the top of the rock, running along the brow. It had a double
entrance with two approaches (Plan 38, 1) protected by a tall bastion, placed,
as at Mycenae, to the right of the entrance as you approach it. Lower down,
on the NW side of the rock, the extension of the fortification known as the
Pelargikon (Plan 38, 2) enclosed the flat area with the caves, communicating
with the rock by way of the NW passageway (Plan 38, 3). Further east, at 4,
was the descent to the N spring and the cave of Aglauros (Plan 38, 5). The
NE approach was no longer in use. It was cut off at the top by the building
of the fortification wall. Lower down, small short-lived houses concealed it
(Plan 38, 6). Within the fortification, the visitor to the Acropolis came upon
the W fa<$ade of house 3 and east of that, complex 5, probably a guard-house.
To the N are terraces I, II, III, IV and V, on which stood the palace. W of ter-
race I there was the earlier Late Helladic building. Erected at the SE corner
were buildings 15, the nearby guard-house, and also graves 16. Houses 17 and
the group of graves at 18 (Plan 38) are against the S wall, further to the W.
On the NW slope there are wells around 22. Traces of habitation, few but
clear, were found on the S slope and, on the SW slope, not marked on the
plan, there are two more wells in the area of the Odeion of Herodes Attikos.
Thus the imposing fortress rose from the midst of a thickly inhabited area,
the seat of the dynast of Athens and symbol of his power.
225
~
0 100 200 300

~mm!l!!!!!!!!!!!!!!!!!!!!;;;!!!!;;;;;;;;i;;;;;;;;;;;;;:;;;;;;;;;;;:;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;:;;;;;;;;;;t-:=============================liiim:;;;;;;;;;;;;;;;~;;;;;;;;;~;;;;;;;;;o;;;;;;;;;o~;;;;;;;;;;;;;;;;s;;;o;;;;;;;e;;;;;;m;;;;;;;;;;;;;;;;;~M
Plan 38. The Acropolis during the final years of the Mycenaean period.
CHRONOLOGY
THE WALL
There are various kinds of evidence for dating the building of the wall,
direct and indirect, all of which agree with each other in determining the fol-
lowing chronology.
1. HOUSES OF THE NE ASCENT. The date of building of the little houses, which,
as we saw, stood on the pathway after the wall was constructed, provide an
indisputable terminus ante quern for the construction of the wall and, chiefly,
that of their abandonment.
The pottery found on the house floors provides the date of their aban-
donment, 398 and comprises both complete and fragmentary vases. They are
datable in the first years of LH IIIC times. Collected in addition were sherds
that evidently belong to the time when the settlement was functioning, going
back to the last years of the LH IIIB period. 399
It is clear, especially from the kylikes found, 400 that the settlement was
inhabited during the later years of LH IIIB, and abandoned at the beginning
of LH IIIC. Since the houses were not in use for long, as witness their poor
construction, we may conclude that the NE ascent was closed and the wall
built shortly before the end of LH IIIB.
2. SHERDS FROM THE s WALL. In 1938, Kolbe excavated inside the wall SE of
the Museum. He collected sherds that he termed Late Mycenaean and dated
them to around 1200 B.C. In a fill next to the SW corner of the Parthenon,
that had not been touched by Kavvadias, he found also fragments of a stirrup
398. Broneer, Hesperia II, 1933, pp. 365- mark, Chronology pp. 19-20, 77.
367, figs 39-45. 400. Broneer, Hesperia II, 1933, p. 372
399. Ibid. figs 37-38, see also ibid. pp. and fig. 44, Mountjoy pp. 45-46.
365-367, Antiquity 1956, p. 12, and Furu-
227
THE THIRD PHASE
jar, one of which was among the stones of the fortification wall. This vase too
is termed Late Mycenaean. 401 Broneer, who saw these finds, discovered that
they had been found together with many fragments of plain kylikes and cups
and he dated them to LH IIIB2. 402 These finds provide direct evidence that
the wall was built at that time.
3. SHERDS FROM THE BASTION OF THE w ENTRANCE. Sherds were recovered

from within the fill of the Mycenaean bastion, the latest of which is described
by Welter as Late Mycenaean - LH III and dated to ea 1200 B.C. 403 Accord-
ing to Broneer, they are contemporary with the sherds from the spring and
with those found by Kolbe, thus belonging to late LH IIIB times. 404
4. SHERDS FROM THE FOUNDATIONS OF THEN WALL, BESIDE THE MEDIAEVAL
BUTTRESS (see Plans 21, 6; 22, 4; 39, 5, and Appendix II, group 5). These sherds
were gathered during the clearing of the stones of the wall and come from
the yellowish earth joining the stones to the rock. Most of them, including the
latest, belong to LH IIIB2.
The conclusions to be drawn from a study of this material are twofold.
First, the wall was built in its entirety at one and the same time. This is evid-
ent from the fact that the finds collected from various different places (the N
wall next to the buttress, the NE ascent, the SE wall, the S wall, the W bas-
tion) all agree. Second, it was built at the end of LH IIIB times. On the basis
of what is known, this will have been towards the end of the 13th century B.C.
THE OTHER BUILDINGS
1. NORTH FOUNTAIN. It was in use for no more than twenty-five years. Much
pottery was recovered from the period of its construction and use, and also
from its subsequent abandonment. The pottery that dates its period of use
401. AA 1939, p. 235, FuF 1939, p. 428, Antiquity 1956, p. 13, Mountjoy p. 40.
Bericht VI Intern. Kongr. Archaologie 1940, 403. AA 1939, pp. 6 and 14.
p. 345, Research and Progress 1940, p. 257. 404. Antiquity 1956, p. 13.
402. Hesperia VIII, 1939, p. 423, n. 176,
228
THE THIRD PHASE
belongs to the last years of LH IIIB. 405 In addition, a study of the pottery from
the houses on the NE ascent provides a more exact dating of its construc-
tion in relation to the building of the wall. As Broneer observed, the later
shapes, numerous among the finds from the fountain, are missing in the pot-
tery from the houses built on the NE ascent. 406 This means that they were
abandoned while the fountain was still in use. This excludes the possibility that
the fountain precedes the houses, as it would have if it had been constructed
at the same time as the wall. For however short the duration of the settle-
ment, it cannot have been shorter than the use of the fountain, since it lasted
throughout the entire time of transition from LH IIIB to LH IIIC. The houses,
therefore, abandoned before the fountain went out of use, were evidently built
shortly before the fountain or, at most, at the same time. The wall, earlier
than the houses, likewise precedes the fountain, which thus appears to have
been a later addition. It may well be that the work of founding the wall led
to the discovery of the underground vein and thus to the construction of the
fountain, in this case soon after the wall was erected. In any case, not long
after the fountain was built, during the middle years of LH IIIC to judge by
the Granary style pottery which has just made its appearance, 407 it fell into
disuse and served instead as a dump until Early Protogeometric times.
2. HousEs ON THE NE ASCENT. As we have seen, the houses were built imme-
diately after the wall and were in use until the beginning of LH IIIC. It is
noteworthy that they were abandoned while the North Fountain was still being
used and before it fell into disrepair.
405. Broneer, Hesperia VIII, 1939, pp. 406. The latest of the vases from the fill
416-423, Daniel, AJA 1940, pp. 552-559, of the fountain are conspicuously absent
Furumark, OpArch 1944, pp. 196-197, Bro- among the pottery from the houses (Bro-
neer, Antiquity 1956, p. 13, W. Gauss (who neer, Hesperia VIII, 1939, p. 424).
follows J. B. Rutter) in S. Deger-Jalkotzy 407. Ibid. pp. 349, 366, Daniel, AJA
(ed.), The Beginnings of the Dark Ages in 1940, p. 558, Broneer, AJA 1948, p. 112,
Greece: LH IIIC Chronology and Synchro- Mountjoy p. 40.
nisms (Wien 2003) pp. 93-102.
229
THE THIRD PHASE
3. GRAVES NEXT TO THE SE CORNER (Plan 38, 16). The largest tomb alone was
equipped with plain pottery, placed beside the head of the dead. 408 The shape
of the vase dates it to advanced LH IIIC times.409
4. BRICK WALL SE OF THE PARTHENON (Plan 38, 17). Only two of the sherds
found together with the bronzes between this wall and the fortification wall
have been published and illustrated. 410 The latest of the two, the lower part
of which is illustrated, shows the lower part of a human figure. It is LH IIICl
in style and it dates the hoard of bronzes and the mud-brick wall.
5. GRAVES s OF THE p ARTHENON (Plan 38, 18). Apart from two pebbles, the
graves contained a krateriskos of reddish, not very fine clay, painted unevenly
brown inside and out except for the base. 411 Like the vase from the other
grave, it is advanced LH IIIC in date.
There is no chronological information for the other buildings. The habita-
tions of the SE comer must have been contemporary with the graves in the
area. The sherds found with the bronzes support this view. It is clear in any
case that all the houses were built after the wall and that the Acropolis was
inhabited down to Protogeometric times.
408. Kawadias-Kawerau pp. 29, 95, Kav- 1889, p. 58, figs 22 and 23, La Grece p. 156,
vadias, Deltion 1888, p. 170, Wolters in figs 500-501, Graef-Langlotz I n°8 202 and
Graef-Langlotz I p. XXXIV, figs 5 and 6. 222.
409. Furumark, MP p. 36, fig. 8, n° 58. 411. Kawadias-Kawerau p. 37, Kawa-
410. AM 1888, p. 108, BCH 1888, pp. dias, Deltion 1888, p. 83, AM 1888, p. 228,
244-245, Kawadias-Kawerau p. 37, Kawa- Graef-Langlotz I n° 176.
dias, Deltion 1888, p. 30, Montelius, VHAM
230
THE THIRD PHASE
Thus the general picture that emerges from the existing evidence is as follows:
LH IIIB2 The entire wall together

with the W bastion is erected.
Immediately afterwards, the
houses of the NE ascent
are built.
The North Fountain is
constructed.
The North Fountain is The Acropolis is turned

in use. The houses into a fortified citadel.
LH IIIB2-IIIC1 of the NE ascent It is inhabited by the
are abandoned. ruler and some of his
subjects
LH IIIC2 The North Fountain is

Beginning of the given up. Houses are
Granary Class built within
the Acropolis.
LH IIIC, down The Acropolis is inhabited

to the beginning by the ruler and some of
of Protogeo- his subjects.
metric times
231
III. CONSTRUCTION METHODS
THE TERRACES
The terrace walls were not built consistently everywhere. The factors deter-
mining their construction were their height and the formation of the rock
behind them, that is, the mass and weight of earth that they had to bank. This
determined the type of foundation, the size of stones used and the thickness
of the walls.
If required by the slope and anomalies of the rock, the surface where the
wall was to be set was first prepared. A characteristic example of this sort of
preparation is the trench in which the W terrace wall of terrace m was set.
The rock was only slightly hollowed out, enough to hold the lowest course of
the outer side in place. A shallow trench was thus cut that followed a straight
line and was distinct only along the outer side of the terrace. The working of
the hollow of the bed of the trench depended on the surface of the rock within
it, deeper where the surface was higher, shallower where it was already lower.
Thus the bed of the trench, while actually sloping from S to N, appears to be
more or less level. Its greatest depth, 0,35 m., is at about the middle length-
wise toward the W rim of the trench. A blunt stone tool was employed that
left no trace on the rock and produced rounded edges and corners. The work-
ing of these surfaces, the boundaries of which are not always clear, is limited
mainly to the trimming of projections that interfered with setting the big
stones and to creating a rim along the outer side of the trench to hold the
blocks in place. The bed of the cutting is in any case barely even and it can
easily be distinguished from the later smooth cuttings in the rock (Fig. 43).
Any depression in the rock, either in the bed of the trench or just outside
it, was filled with a mixture of mud and small stones, thus forming an under-
pinning for the big blocks of the terrace wall.
This preparation was not carried out everywhere, but only where it was
considered necessary. In some cases another less costly and troublesome
method was employed. A series of fairly small stones was set along the length
232
CONSTRUCTION METHODS
Fig. 43. Preparation of the rock for the foundations of the terrace walls: the bed
of the trench W of the supporting wall of terrace Ill
of the line of the terrace face. Their height and width was such as to level the
surface on which the large stones were set. These small stones were chosen
carefully, well wedged into place and held by the weight of the wall that cov-
ered them. They formed an exceedingly stable foundation (Fig. 44).
The wall itself was then constructed of large or small stones, according to
need. Gaps in the fa~ade were filled in with smaller stones. On the inner side,
in addition to small stones, a yellowish clay was used, not as filler between the
stones but to seal the joins of the Cyclopean construction so that the earth of
the terraces would not wash out.
The big stones were set on the outer side, which was the visible side of the
terrace wall (Fig. 16). They were, as a rule, smaller than the stones used for
the fortification walls and the surfaces were less smooth. The construction was
indifferent, the greatest attention, naturally enough, focussed on the outer side
only. The stones of the interior were smaller and unevenly laid, as a fill rather
than in straight courses.
233
Fig. 44. Foundation of the terrace wall directly on the uneven rock without cut-
ting: the NW comer of terrace W from the W.
The thickness of the supporting walls varied according to circumstances,

but in general ranged between 1 and 1,50 m. Only in the W wall of terrace
III, the largest and most heavily built of all, is this width surpassed. In places
where the terrace was lower, the stones are smaller and the walls are nar-
rower.
THE FORTIFICATION
The few sections of the wall preserved and visible today show an unusu-
ally good quality of workmanship and very careful construction. Indeed the
builders paid far more attention to its stability and durability than to aesthetic
results. Yet they were not entirely oblivious to this aspect. A number of refine-
ments show both a highly developed building skill and an attempt to give the
wall a form that suggests an aesthetic approach to its final appearance.
The wall is founded everywhere directly on the rock, at the very edge. The
rock is, to be sure, uneven and at most places the surface had therefore to be
worked before laying the foundations. The preparation is of two main types.
234
Where the rock slopes and a limited area needs to be worked, the surface has
been carefully cut to form parallel stepped levels with smooth surfaces,
rounded corners and ends more or less regular. Such cuttings are visible today
in front of the Pinakotheke, at the beginning of the Pelargikon and above the
NE ascent (see Figs 17, 23, 39, 40). Where the rock is approximately level,
all projections have been removed, producing thus a rough and uneven sur-
face but one suitable for the stability of the wall foundations. Preparations of
this sort are visible today at the E wall, between the Belvedere and the
Museum, and on the surface at the edge of the plateau just after the begin-
ning of the Pelargikon. Actual construction of the wall itself followed these
initial preparations. The large Cyclopean blocks were founded either directly
on the rock or on a series of smaller stones providing the necessary level sur-
face, as were the terrace walls. This method has been employed primarily
along the inner face where the lower courses are hidden by fill and do not
show.
From the standpoint of construction, the wall is made up of two thick par-
allel faces, the outer and the inner, carefully built of large, heavy stones laid
in irregular courses in the Cyclopean style. Small stones fill the spaces between
the massive stones, thus blocking the gaps and providing a more stable sup-
port for the course above (Fig. 24). Between the two faces is the core, care-
lessly made of smaller stones joined with the yellowish insulating clay we have
seen in the terrace walls, and of which very little remains in the visible sec-
tions today. The usual building material was Acropolis limestone, with occa-
sional pieces of schist or stone from the Pnyx. The blocks used for the faces
are generally of large proportions. 41 2 If not naturally smooth, the surfaces may
show traces of hammering to achieve a smooth and level face (Fig. 45) , as at
Tiryns. 413 The Acropolis rock itself is a good source for large stones with a
smooth surface because many small veins of a softer composition run through
it in all directions, so that large pieces with relatively flat surfaces can be bro-
412. The outer side of the wall com- 1,70x0,65 m., are in the leg of the wall S
prises stones usually with a maximum di- of the Propylaia. The average size of stones
mension of 1 m. The largest of those acces- used is 0,90-1 ,lOx0,55-0,75 m.
sible, measuring 1,90xl,35 m., l ,80xl,15 m., 413. Muller, Tiryns III p. 178, fig. 79.
235
Fig. 45. Block of the inner face of the SE fortification wall, showing traces of
hammering on the surface.
ken off. From the material collected, the builders obviously chose the largest
and best pieces for the faces of the wall.
The wall faces are generally vertical, to the extent allowed by the mater-
ial. Wherever there is a deviation from the vertical, it was clearly not by
design.
Where the wall runs along in a straight line or curves, the construction is
simple and presents no problems. If there is an abrupt change of course or of
size, the construction may differ from the usual and it is worth noting the
building technique employed. A characteristic example is provided by the
inside corner at the end of the S wall and beginning of the W, to the S of the
Propylaia (Plan 33, 2, 3). At that point, where the two legs of the wall meet
to form an acute angle, the foundation on the rock is common to both. It
comprises small, irregular stones placed so as to form a curving and irregular
corner (Fig. 46). The differentiation of the two legs begins from the present
surface upwards (coinciding with the height of the Mycenaean fill), at the
236
point of the first vertical joint. At this level, the interior line of the S leg
extends some 2,50 m. within the width of the W wall. This had been observed
by Kawerau, 414 who concluded that the S leg had been constructed after the
W and added to it. This, however, is not the case because, as we have seen,
the foundation is in one piece, and because the extension of the line of the S
leg is not more than half of the width of the W leg. It is due simply to the
method used to achieve a regular joint above the level at which it was visible.
First they built the outer face of the W leg and the shared interior founda-
tion. Then they built the inner face of the S leg, carrying it for some distance
within the width of the W leg, and finally the inner face of the W wall was
added, the end of which, built against the straight surface of the S leg, thus
made a fairly regular vertical joint (Fig. 46).
Another example of interest is the bastion of the W entrance. In this place
they had to construct a large bastion of considerable height, compact and
founded on the sloping spur of the rock. If it had been built like the wall, with
an outer casing of massive stones and an inner compact fill of small stones
and earth, the tremendous weight of the fill, increased by natural compres-
sion (settling), would have broken the W face of the bastion, which in any
case was subject to the greatest thrust because of the slope. This they avoided
by building within the bastion a strong cross-wall, that ran parallel to the W
wall some 4,50 m. within it. It was founded not directly on the rock but on
the embankment of the fill,415 and held the fill to the E, thus greatly reliev-
ing the load on the W fa~ade. Since it was also built into the side walls of the
bastion, it joined the whole construction together solidly.
The thickness of the fortification wall was not constant throughout. In
places where both faces are preserved, the greatest thickness, in the stretch
S of the Propylaia, reaches 6 m. East of the Museum it is at its narrowest,
3,60 m. 416 For the stretches that have not survived, we cannot be so sure, al-
though the various existing indications give a fair approximation. At a mini-
414. Kavvadias - Kawerau pp. 141-143, sider this and interpreted it as a terrace
pl. H. wall, unconnected with the bastion.
415. Balanos, Ephemeris 1937 f , p. 788. 416. See n. 275 .
Bundgaard (Mnesikles p. 69), did not con-
237
Pig. 46. The foundation of the SW comer of the fortification wall by the Propylaia.
mum it will have been at least 3 m. Quite apart from the question of stabil-
ity, this will have been necessary for reasons of security in case of siege, since
for purely constructional reasons, a thickness greater than around 1,50-2 m.
is not statically necessary for walls of that height and type of construction. The
top of the wall, however, had to be wide enough to accommodate the circu-
lation of the defenders, allowing them enough room to fight. This is the sole
reason why the width of the wall, as we have seen, is increased I
to 5 and 6
metres at the vulnerable and easily approached spots in the fortification, such
as the entrances, for at the points where the main thrust of the enemy force
would be directed, as many fighters as possible would have to be assembled
and they needed room to manoeuvre freely.
238
Nowhere is the full height of the wall preserved, nor have we any clear
indications. On the basis of what evidence we have, we estimate a height of
at least around 8-10 m. The W bastion may have been higher. 417
THE NORTH FOUNTAIN
The daring, imagination and practical outlook of the technicians of the

time are pre-eminently evident in the construction of the descent to the North
Fountain. The difficulties they had to face lay not only in the steepness of the
descent within a narrow, dark fissure deep in the rock, but also in the fact
that the walls of the fissure were uneven, with projections and angles con-
stantly changing the shape of the fissure itself. The fact that in the eight flights
of the descent, the method of construction was altered four times, shows an
unusual degree of adaptation to the technical problems.
The beginning of the descent is precisely at the edge of the main mass of
the Acropolis rock. The edge has been worked so as to form a regular step
with a rounded edge, smooth and straight, facing N. The step leads to a level,
within the fissure, forming a landing. The traces of the Mycenaean descent
begin precisely at 0,45 m. below the step. The traces, however, are those of
the support for the beginning of the stairway, so that the thickness of the top
step, clearly wooden, must be subtracted from the 0,45 m. The first step of
the descent will thus have had a height of around 0,30 m.
Below this landing, the descent begins. The first two flights were con-
structed as follows. In each wall of the fissure a series of two facing and cor-
responding rows of somewhat irregular hemispherical cuttings were made by
means of a stone hammer. The line follows the descent of the stairway. While
nearly all the cuttings are preserved on the S face of the fissure, on the oppos-
ite side later cuttings have all but obliterated the traces. The only one well
preserved and accessible today has a diameter of 0,14 m., while others reach
0,17 m. The space between them ranges from 0,35 to 0,60 m. and the differ-
ence in height from one to the next is around 0,18 m.
417. See supra, pp. 186-187. The wall today 8,25 m. high, clearly lower than its
around Grave Circle A at Mycenae stands original crown, which has not survived.
239
Fig. 47. Reconstruction drawing of the construction of the upper sections of the
descent to the North Fountain (Broneer, Hesperia VIII, 1939, fig. 6).
The hollows are sockets for wooden beams, round in section, that were
placed horizontally across the fissure to support the wooden steps. On the
basis of the above measurements, we have a series of beams averaging 0,15
m. in diameter, that supported steps with a tread 0,30-0,60 m. wide and a riser
with an average height of 0,18 m. This system of construction could not be
continued unchanged as far as the end of the first flight to the beginning of
the second, or the second to the third, because it took up the entire space of
the fissure. Indeed, at these turning points hollows have been cut in only one
side of the fissure, leaving an empty space corresponding to 8 steps at the end
of the first section and 10 steps at the end of the second. Here the steps were
inserted in only one side of the fissure. They ended at about the middle of
the width where they rested on a slanting beam, the upper end of which rested
on the last full step of the flight and the lower end on the landing where the
flight ended (Fig. 47). Since, as Broneer observed and as the visitor can see,
there was continuous circulation of air to the bottom of the fissure and there-
fore very little dampness, there was little danger of the wood rotting, and if
240
Fig. 48. Reconstruction of the third, fourth and fifth sections of the descent to the
North Fountain (Broneer, Hesperia VIII, 1939, fig. 18).
need arose the stairs could always be replaced piecemeal. The last part of the
second flight and the beginning of the third have left no traces, apparently
because these steps and the landing between them were founded on fill in the
fissure. From this point downwards, the construction changes again. Wooden
steps could not be kept in repair at this depth because of the dampness.
Another solution was therefore found for the third, fourth and fifth flights,
the details of which emerge from traces on the rock and from destroyed
remains of the stairway. At intervals less than 1 m. a series of wooden beams
were placed upright, at a distance from the wall of the rock equal to the width
of the stairway. The lower ends stood in hollows like those made for the pre-
vious ones, but cut vertically into the rock. The beams were joined at various
points with others, which ran horizontally into the wall of the fissure where
the ends were held in rock-cut hollows. The construction was reinforced by
241
Fig. 49. The two final sections of the descent to the North Fountain, built on the
rock (Broneer, Hesperia VIII, 1939, fig. 19).
other beams, horizontal and parallel to the wall of the rock. These were fast-
ened to the interior of the vertical beams, which were thus held in place. This
formed a wooden framework, a sort of loose lattice, with a series of roughly
square divisions along the face of the wall of the fissure. The framework was
filled with a mixture of mud plaster and relatively small stones. On top of all
this the steps were placed. These consisted of coarsely worked limestone slabs,
joined to the infrastructure with a layer of yellowish clay that held them firmly
in place (Fig. 48).
242
This kind of construction reinforced by timbering was not unknown in

Mycenaean times and was used in wall construction as well. 418 It is quite strong
as long as the wood is not exposed to dampness.
The few steps of the sixth flight of the descent were of stone, held on a
substructure of wood but wedged into both walls of the narrowing fissure.
They were thus held in place even after the wood disintegrated.
The seventh and eighth flights did not present the same difficulties of con-
struction. At this depth, the limestone comes to an end and schist (kimilia),
which is much softer, begins. This they managed to dig into and they prepared
a strong stone substructure on which the steps were laid. The final step of the
eighth flight, 1,55 m. wide and much larger than the others, was inserted
beneath the other steps, which it thus supported (Fig. 49).
Below this final flight of the descent, was a well, 8 m. deep. It was dug out
of the schist and its walls were therefore in danger of collapsing. Traces of
wood show that the walls had been shored up with boards and small beams.
THE HOUSES
Within the Acropolis, remams of houses were preserved in the Pina-

kotheke, in the SE corner of the fortification wall and S of the Parthenon.
There is in addition the earlier LH I house N of the Erechtheion, the con-
struction of which, with a stone foundation and a fine floor on the interior,
has already been described.
The later LH III houses are all built in the following way: the lower part
of the walls is constructed of small unworked stones, set with a small amount
of mud-plaster. In wall 3 on Plan 32, this substructure reached a height of
around 1 m.419 In other cases the wall substructure was founded directly on
the rock,420 or on the fill, which had collected on the surface. 421 Above this
foundation, rose a wall of unbaked brick. Only in one case has a floor sur-
418. See also Wace, BSA 25, 1921-1923, 420. Ibid. p. 51, Kawadias-Kawerau p.
pp. 42, 88 f. and fig. 20. 59.
419. Montelius, VHAM 1889, p. 51. 421. Kawadias-Kawerau p. 103.
243
vived. It was made of a layer of small irregular stones covered over by a layer
of mud-plaster. 422
The houses of the NE ascent have floors of tamped earth and walls roughly
built of small stones and stone slabs that served as bases for wooden roof
supports.423 Although these are simple, crudely made constructions by com-
parison with the houses erected within the Acropolis, they were evidently built
according to the following general rules: stone foundations, superstructure of
unbaked brick and wooden columns that rested on stone bases.
The preservation of these remains is in general so poor that it is imposs-
ible to have any idea of the arrangement of space. As far as can be observed,
the rooms were rectangular and ample.
THE PALACE
Of the palace, only a single column base and the sandstone steps remain,
and these are not in their original places. As evidence they are admittedly
insufficient for a restoration of the palace. They are not even enough for pos-
itive conclusions as to the appearance of the palace. To be sure, the building
had wooden columns, and evidently a second floor if the stairway from which
the steps come belongs to the palace and not to the terraces. For the rest, we
can only surmise what it may have looked like, based on comparisons with the
palaces of Mycenae, Tiryns, Pylos and, perhaps, Gia.
THE STAIRWAYS
The stairways, with wooden or stone steps, and sometimes cut out of the
rock itself, are constructed without any symmetry whatsoever. This is clear
from the measurements. Such a conception was first applied in Classical times.
It was inherited by the Roman architects and in turn by the Renaissance,
which passed it on to modern times. Popular architecture ignores it altogether.
422. Kavvadias - Kawerau p. 103. 423. Broneer, Hesperia II, 1933, p. 355.
244
It is rare to find two steps in a Mycenaean stairway that have exactly the same
measurements. Many are far from ample, with narrow treads .and high risers.
In the descent to the North Fountain, which we have already discussed, the
steps of the last built sections have treads ranging in width from 0,26 to 0,40
m. and risers varying between 0,12 and 0,25 m. In the NE ascent, the steps
discovered by Broneer are few and divided into many units, so that there is
no group suitable for study. The others, excavated by Kavvadias, were con-
solidated by him with mortar and by the addition of marble in places, so that
it is uncertain whether they still have their initial proportions. Furthermore,
neither their number nor their arrangement agree precisely with Kawerau's
plan. Even the sandstone steps of the palace, which belonged to a large and
imposing stairway, are not precisely similar in all their measurements.
Certainly, in the stairways preserved, the steps, when the incline allows,
alternate with short ramps, the surface of which consists of tamped earth and
pebbles. The steps themselves are not always horizontal, but are sometimes
on a slant. They are made of coarsely hammered stone slabs, occasionally set
on edge, the rest of the tread being formed by the same mixture of earth and
pebbles as used for the sloping surfaces.
THE GRAVES
All the LH graves on the Acropolis are cist graves, the sides lined with
small irregular slabs set without any binding material, and covered with other,
similar slabs. 424 They are dug into the fill and rest directly on the rock (see
Fig. 28, left).
424. Kavvadias - Kawerau pp. 37, 39, p. XXXIV, fig. 5).

95, 99 and Wolters (Graef - Langlotz I
245
EPILOGUE
THE ACROPOLIS OF ATHENS

IN RELATION TO THE OTHER MYCENAEAN CITADELS
EPILOGUE
The remains of the Acropolis of Athens are few and in poor condition. Yet
the picture they present is clear and comprehensible. It is completely remin-
iscent of the citadels of Mycenae and of Tiryns both in general and in detail.
As were they, it too was inhabited without a break from very early times, well
before it became a fortress. After that, during LH times, it saw more system-
atic and intensive use. Apart from the remains of the room N of the
Erechtheion, 425 nothing is left of the first periods of this era. Until the con-
struction of the terraces, there is an intervening gap of centuries. Yet there are
ceramic finds from the first great excavation of the rock, as well as evidence
provided by the graves found and excavated by the American School of Clas-
sical Studies in the Ancient Agora. 426 These, dating from Early LH II to the
final years of LH III, show that the area immediately around the Acropolis and,
to be sure, the Acropolis itself continued to be inhabited throughout this time.
Building activity at Mycenae falls into three phases, including both the for-
tification wall and the palace. The first phase belongs to the LH IIIA period.
Dating to that time are the N Cyclopean wall and the first palace construc-
tion, including the Pillar Room. During the second phase, in LH IIIB, the SW
wing of the palace was renovated and the N Postern Gate, the Lion Gate and
the section of the fortification that included Grave Circle A were added to
the fortification wall. Later on, during LH IIIC times, the grand staircase of
the palace was built and the small extension taking in the Perseian fountain
was added to the fortification. 427 The acropolis of Tiryns likewise has three
building phases, all three Late Helladic428 as at Mycenae. 429
425. Supra, pp. 73-75. Mycenae and their chronology see Mylo-
426. Furtwangler-LOschke, M. V p. 35, nas, EpistEpetAth 1955/1956, pp. 167-177,
fig. XVI, Graef-Langlotz I pp. 6-22 and pls Ancient Mycenae pp. 37-38, 64-66, and
2-7, Travlos, Iloleoo. fig. 7, Shear, Hespe- Ephemeris 1958, pp. 153-207, Ergon 1959,
ria IX, 1940, pp. 274-291, E. Townsend pp. 93, 96-97; 1961, pp. 149-150, 154. See
Vermeule, Hesperia XXIV, 1955, p. 188, also Wace, Mycenae pp. 86-90.
and n. 4, with relevant information and bib- 428. Miiller, Tiryns III pp. 205-209.
liography. Also Mountjoy p. 16. 429. Mackeprang, AJA 1938, pp. 556-
427. For the various building phases of 559.
249
EPILOGUE
In terms of building development, the Acropolis of Athens has much in

common with these two citadels, especially during the last two phases. Indeed
the second period of the Athenian Acropolis, the time of the terraces, coin-
cides, with some delay, with the second phase of Mycenae and Tiryns. The
third period of the Athenian Acropolis, when the fortification was erected,
coincides with the third phase at the other two citadels. After that, only the
houses and graves were added. In other words, some time after Mycenae and
Tiryns had constructed fortification walls for the first time, the Acropolis of
Athens acquired a series of terraces, constructions which preceded the forti-
fications and the purpose of which was not defensive. The Athenians of that
time clearly felt no need of protection. Perhaps they were not important
enough to be in any serious danger. Be that as it may, by the time they arrived
at their first building phase, the other two sites had gone beyond this phase
and had become citadels. Soon, however, either in anticipation of some spe-
cific danger, or because power politics was on the rise, Athens too built a for-
tification system complete with bastion, thus copying the developed phase of
the fortifications of her two models. This was the time when Mycenae and
Tiryns extended their fortifications and built the first bastions, and Gia in Boe-
otia was fortified. 430 The third phase was not followed by Athens, evidently
because there was no reason, since at Mycenae the extension was made specif-
ically to include Perseia, 431 whereas at Athens the corresponding North Foun-
tain, constructed after the wall, lay within the fortification. The purpose of the
430. See I'J..6.~ I (Athens 1989) 256-258. of the citadel hill of Mycenae, however,
431. In his introduction to the Greek preclude this apparently simple solution.
edition, Professor Marinatos rejected this Indeed, the pipe runs underground as close
view with the argument that since the pipe as possible to the base of the rock, but it
through which the water of the spring flows cannot be taken any further because the
ended so close to the fortification wall, limestone was too hard for the tools and
rather than adding such a length of wall, it technical means of the time, which were
would have been simpler just to lengthen inadequate for opening the channel re-
the channel a few more metres, thus bring- quired. Thus, because of the gradient of the
ing the water within the already existing rock, the underground cistern where the
wall. The morphology and composition channel ended and the water was collected,
250
EPILOGUE
extension at Tiryns, which by now is being systematically explored, appears to

have been to increase the fortified area, whereas at Athens the wall already
included all the available space.
Thus the Acropolis of Athens, with initial delay, reached the same state of
fortified strength as that achieved by the other two great citadels at about the
same time, around the end of LH IIIB. In the course of this period, the char-
acteristically Mycenaean principles governing the construction of fortifications
were finally worked out and applied, according to which the fortifications, that
is the Cyclopean wall, were built adapted to the rock formation, thus creat-
ing a protected but accessible area. The invention and building of bastions for
protecting the entrances form part of this same concept. In Athens, all this
was carried out as part of a fully-fledged program without preliminary stages
of development and without the possibility of subsequent changes.
In the other citadels, the fortification systems show a long, inherent devel-
opment, crystalised after changes and experiments which we can see applied
from stage to stage. Despite a slow beginning, the Acropolis of Athens shows
a sudden application of these principles in fully developed form, at almost the
same time as the others. There is thus no doubt that they found these prin-
ciples ready, and they imitated them. This conclusion is supported also by
chronological facts; for the fortification of Athens belongs chronologically to
the final years of LH IIIB, whereas at the other sites, the second phase goes
back to the advanced years of the same period. Thus it is somewhat - but only
slightly - earlier. The imitation is indeed worthy of the prototype. In some
details it may be even better.
had to be built a few metres beyond and NE extension of the wall. The further
lower than the brow of the rock (on which course of this extension was suggested by its
the fortification wall was founded) , and adaptation to the somewhat lower brow of
therefore outside the fortified area. So, the little hollow plateau east of the old for-
using a natural crevice in the rock at the tification, so that by adding almost 140
foot of the NW corner of the original for- metres of wall, a small area was adjoined to
tification wall, the Mycenaean technicians the citadel, which obviously cannot be char-
made a stepped descent to the spring, the acterised as a "lower citadel".
beginning of which was protected by the
251
EPILOGUE
The material remains on the Acropolis are, admittedly, very few and it is
therefore not possible to draw completely firm conclusions from comparisons
with the details of the other citadels. Even so, on the basis of what is pre-
served, it can be seen first of all that the Acropolis has more in common with
Tiryns than with Mycenae, from which it borrowed the form of its entrance.
For the rest, the diameter of the column base agrees with the diameters at
Tiryns and the stones of the fortification wall show the same careful choice
and workmanship. At Mycenae, except for the Lion Gate and the N Postem
Gate, where the conglomerate stones of the fa<;ade are regular and the sur-
face is the result of the type of stone used,432 the limestone blocks of the
Cyclopean wall of the second, and even the third, phase, are practically
unworked. At Tiryns, however, both on the preserved sections of the first wall
and on the remnants of the second, we may observe the same quality of work
as at Athens. 433 The fortification wall at Athens, made up of large but not
enormous stones, placed lengthwise like stretchers, with a few spaces between
them, shows similar dimensions in its construction and the same arrangement
of the stones as that of the first wall at Tiryns.
As for the architect, the man who undertook the entire work and made
the decisions about its details, it is impossible to pick out and determine pre-
cisely the origin of the influences to which he will have been exposed. We
would say, however, that the technicians who carried out the work in all like-
lihood learned their trade, directly or indirectly, at Tiryns.
The massive and overwhelming sight presented by the fortifications of the
Acropolis of Athens fully justifies the veneration shown by the ancients, who
recorded that they were built by beings of supernatural strength, to whom they
referred with the suitably invented names Hyperbios (the all-forceful) and
Agrolas (huge boulder). 434 It explains too the outlook of the vase painter who
represented Athena leading a giant who carries on his shoulder a huge stone
destined for the wall. 435 Its construction bears witness to the capability and
432. Marinatos-Hirmer, Kei/rrJ xai Mv- 434. Paus. I 28, 3.

xnvaixq 'EA.A. a~ p. 58. 435. See Hauser, Strena Helbigiana pp.
433. See Muller, Tiryns III pp. 55-57. 115-121, fig. p. 116.
252
EPILOGUE
high technical level of its builders, whose work is preserved in few and barely
accessible remains so that it is difficult to appreciate it fairly. To conclude with
the words of Blegen436 : "The regal magnificence of the walls themselves, which
places them on a qualitative level with those at Mycenae, has seldom been
adequately recognized."
436. HSCP 1940, suppl., p. 1, n. 1.
253
APPENDIX I
ANCIENT TESTIMONIA
(By topic and chronological order)
I. THE FORTIFICATION WALL BUILT BY THE PELASGIANS/

TYRRHENIANS
1. Herod. VI 137 (Hekataeus, 545-475 B.C.)
nEAacryoi ElTElTE EK TflS :ATTlKflS VlTO When the Pelasgians were cast out of
:A6T)vaiwv E~El3Ai)6T)crav, EhE wv 8Ti Attica by the Athenians, whether justly
81Kaiws ElTE a8iKws· TOVTO yap OUK or unjustly, - as to that I can say nothing,
EXW cppacra1, TIAT)v Ta AEyoµEva , 0T1 beyond what is recorded, namely, that
'EKaTaios µ€v 6 'HyT)crav8pov E<J>TJ<YE Hecataeus the son of Hegesandrus de-
EV TOlO"l A6yo1cr1 AEywv a8iKws · ElTEi- clares in his history that the act was
TE yap 18e1v Tous :A6T)vaiovs Tiiv unjust; for (says Hecataeus) when the
I \ I l - < \ \
XWPllV, TllV O"CJ>lO"l aVTOlO"l \JlTO TOV Athenians saw the land under Hymettus
•
'YµT)CYO"OV Eovcrav E8ocrav neAacryo1- which, being their own, they had given to
0"l o1KT1crai µ1cr66v TOV TElXEOS TOV the Pelasgians as a dwelling-place in re-
lTEpi Tiiv CxKp6TioA1v KOTE EAT)AaµE - ward for the wall that had once been
vov, TaVTTJV ws 18E1v Tous :A6T)vaiovs built round the acropolis, - when the
E~EpyacrµEVTJV Ev, Tiiv lTpOTEpov Elvai Athenians saw how well this place was
KaKi}v TE Kai TOV µ118Ev6s a~i11v . Aa- tilled which erewhile had been bad and
13e1v cp66vov TE Kai 'iµEpov TflS yfls, worthless, they grudged and coveted the
Kai OUTW E~EAaUVElV aUTOUS ou8Eµiav land, and so drove the Pelasgians out on
&M11v 1Tp6cpacr1v 1Tpo"icrxoµ€vovs Tovs this and no other pretext. But the Athe-
:A611vaiovs. ws 8€ auToi 1'611vaio1 AE- nians themselves say that their reason for
yovcr1, 8tKaiws E~EAacrai. KaT01K11µE- expelling the Pelasgians was just. The
vovs yap TOUS n EAacryous UlTO T~ Pelasgians, they say, issued out from
'YµT)crcr~ EV6EVTEV 6pµwµEvovs a81- their settlement at the foot of Hymettus
KEE1v Ta8E · cponav yap aiei Tas crcpe- and dealt wrongfully with the Athenians
TEpas 6vyaTEpas TE Kai Tous Tia18as in this wise: neither the Athenians nor
1
E1T u8wp ElTl Tiiv 'EvvECxKpovvov · OU any other dwellers in Hellas had as yet
yap Elvai TOVTOV TOV xpovov crcpicr1 servants at that time, and their sons and
257
APPENDIX I
KW ou5e Toten &Moten "EM11crt OlKE - daughters resorted to the Nine Wells for
Tas· oKws 5e eA6otEv avTai, Tovs water; and whenever they came, the Pe-
nEAacryovs V1TO u~pt6s TE Kai 6At- lasgians maltreated them out of mere
ywpi11s ~tacreai cr<peas . Kai TavTa arrogance and pride. Nor yet were they
µevTot cr<pt ouK aTioxpav TiotEEtv, content with so doing, but at last were
&Ma TEAos Kai e Tit~ovAEvovTas eTit- caught in the act of planning to attack
XElPTlenV <pav11vat E1T avTo<pwp~.
, "V ' ' , ,
Athens. The Athenians, by their own

ewvTovs 5e yEvecreai TocrovT~ eKEi- showing, dealt so much more rightly than
vwv &v5pas aµEivovas , Ocr~ TiapEOV the Pelasgians, that when they might
EWVTOlen cl1TOKTElVOl TOVS nEAacryovs, have killed them, caught plotting as they
ETIEi mpEas EAa~ov ETit~ovAEvovTas , were, they would not so do but bade
OUK €6EAT;crat, aMa cr<pt 1TpOEt1TElV them depart out of the country. There-
EK TT;s yT;s e~tevai. Tovs 5e ouTw eK- upon the Pelasgians departed, and took
xwpf)cravTas aMa TE <JXElV XWpia Lemnos in possession, besides other
Kai 5Ti Kai /\T;µvov. EKEiva µev 5Ti places. This is the Athenian story; the
'EKaTaios EAE~E , TavTa 5e :A611vaiot other is told by Hecataeus. (Loeb, A. D.
Aeyovcri. Godley)
2. Aristoph. Av. 1139 (414 B.C.)

(the wall of Cloudcuckooland is being built)
.... "ETEpot 5' E1TAtv6o<p6povv TIEAap- .... another ten thousand storks (rccA.aQ-
yoi µvptoi. yol) made bricks. (Loeb, J. Henderson)
3. Dion. Hal. Antiq. Rom. A, XXVIII, 4 (lst cent. B.C.)
MvpcriAos 5e Ta EµTiaAtv aTio<paiv6- But the account Myrsilus gives is the

µEvos 'EMaviK~ Tovs Tvpp11vovs <p11- reverse of that given by Hellanicus. The
crtv, E1TEt5Ti TTiv eavTwv E~EAtTiov, ev Tyrrhenians, he says, after they had left
TTJ 1TAOVT) µETOVOµacr6T;vat nEAap- their own country, were in the course
yous ... Kai Tois :A611vaims To TEixos of their wanderings called Pelargoi or
TO 1TEpi TTiv :AKp01TOAtV, TO nEAap- "Storks" ... and they built the wall round
ytKOV KaAovµEvov , TovTovs TIEpt~a the citadel of Athens which is called the
AEiv. Pelargic wall. (Loeb, E. Cary)
4. Paus. I 28, 3 (2nd cent. A.D.)
TT) 5e clKp01TOAEt, TIATiv ocrov Kiµwv The Acropolis is surrounded by a wall; a
258
APPENDIX I
c_;,Ko86µ11crEv avTT)S 6 M1AT1a8ou, TIE- part was constructed by Cimon, son of

p1j3aAEiv To Ao1Tiov AEYETa1 Tov TEi- Miltiades, but all the rest is said to have
xovs n EAacryovs OtKi}craVTCxS TIOTE been built round it by the Pelasgians,
vTio TT,v aKp6TioA1v· cpacri yap 'A- who once lived under the Acropolis. The
yp6:Aav Kai 'Y.rrep131ov Tovs o1Ko5o- builders, they say, were Agrolas and
µf}cravTas Elva1, Tivveav6µEvos 8€ oY- Hyperbius. On inquiring who they were
TlVES ilcrav, ov8h &Mo E8vvaµ11v µa- I could discover nothing except that they
6Eiv r; LlKEAovs To E~ apxl1s ovTas were Sicilians originally who emigrated
ES 'AKapvaviav µETOlKT)crai. to Acamania. (Loeb, W. H. S. Jones)
5. Schol. Aristoph. Av. 1139
TIEAapyoi: 816: TO nEAapylKOV TElXOS TIEAapyoi: in allusion to the building of

TOVS CxTIO T upp11vias TJKOVTOS ava- the Pelargic wall by men who came from
O"TllO"Ol . Tyrrhenia. (William G. Rutherford, Mac-
Millan ed.)
6. Schol. Aristoph. Av. 832
To nEAapyiK6v: <oTI) 'A6i}v11m TO nE- T 0 nEAapy1K6v: observe that the wall

AapylKOV TEixos ou µeµv11T01 KaMi- of this name was at Athens. It is men-
µaxos· "Tupcr11vwv TEix1crµa nEAap- tioned by Callimachus: -"The Pelargic
y1K6v". wall of Tyrrhenian building". (William
G. Rutherford, MacMillan ed.)
7. I. Bekker, Anecd. Gr. I (Berlin 1814) p. 299, ll. 16-18
nEAapy1K6v: To VTIO Tuppf}vwv Ka- Pelargicon: The wall built by the Tyrrhe-
TOO"KEVO<r6Ev TElXOS, oOs 6EacraµEV01 nians, whom they named storks (TIEAap-
810 TOS criv86vas, as E<p6pouv, TIE- yous) for the garments of fine cloth
Aapyovs wv6µacrav. which they wore.
8. Hesych. s.v. nEAOO"TlKOV (5th cent. A.D.)
TIEAacrT1K6v· TE1xiov ouTw Ev 'A6f}va1s Pelasticon. A wall of that name built by

KOAovµEVOV T upp11vwv KTlO"CxVTWV. the Tyrrhenians.
259
APPENDIX I
9. Phot. Lexicon, s.v. nEAapy1K6v (820-891 A.O.)
nEAapy1K6v· TO UTIO TWV Tvpp11vwv Pelargicon. The wall of the Acropolis,

KaTaCTKEvacr6Ev TTlS CxKpOTIOAEC.US TEl- built by the Tyrrhenians. These were
xos· TOVTOVS yap KA116T)v01 nEi\ap- named Pelargoi, as it were Pelasgoi,
yous, ofov nEAacryous, ws ni\av11Tas being migrants. Or, because the Atheni-
Tivas · Ti 8T1 i86vTES avTous npwTov ans, when they saw them, likened them
oi f\611vaio1 cr1v86vas Aaµnpas nEp1- to storks (nEAapyoi) because they wore
PEPA11µevovs nEAapyoi's EiKacrav. bright garments.
10. Etymologicum Magnum , s.v. nEi\apy1K6v (lOth cent. A.O.)
nEAapy1K6v, TO UTIO Tvpp11vwv Ka- Pelargicon, the wall which was razed to
TacrKa<pEv TEixos, oOs Kai 6EacraµE - the ground (KaTacrKa<pev, obviously in-
vo1 TlVES, nEi\apyous wv6µacrav 81a stead of KaTacrKEvacr6ev, built, supra
Tas mv86vas, &s E<p6povv. nos 7, 9) by the Thyrrhenians, whom they
named Pelargoi for the garments which
they wore.
II. THE FORTIFICATION IN GENERAL
11. Herod. V 64-65 (484-410 B.C.)
Ki\Eoµev11s 8E an1K6µEvos ES To &crTV Then Cleomenes, when he and the Athe-

&µa f\611vaiwv Toi'm povAoµevo1cr1 E1vai nians that desired freedom came before
e:Aev6epo1cr1 eno:A16pKEE Tovs Tvpav- the city, drove the despots' family within
vovs CxTIEpyµevovs EV T~ nEi\acry1K~ the Pelasgic wall and there beleaguered
TEiXEi Kai ou8ev Tl TICxVTC.US &v E~Ei them. And assuredly the Lacedaemonians
i\ov n E10'10'TpaTi8as Ol /\aKE 801µ6v101 · would never have taken the Pisistratid
ovTE yap Ene8p11v EnEv6Eov no1i)cra- stronghold; for they had no mind to
cr601, Ol TE nE10'10'TpaTlbOl crho1cr1 Kai blockade it, and the Pisistratids were well
n0Toi'cr1 EU napEcrKEva8aTo· noA1op- furnished with food and drink; and the
Ki)cravTES TE &v ilµepas oAiyas anaA- Lacedaemonians would but have besieged
ACxO'O'OVTO ES TT,v :L napT11v. the place for a few days and then returned
back to Sparta. (Loeb, A. 0. Godley)
12. Herod. VIII 51-53 (484-410 B.C.)
(Oi pappapo1) ... EyevovTo Ev T~ (the barbarians) ... arrived in Attica,

f\TT1KTJ, Kai\i\1a8Ec.u &pxovTos f\611- Calliades being then archon at Athens.
260
APPENDIX I
I \ ( t >I \ >I
va101cn. Kal atpEOUO"l Ep11µov TO a- There they took the city, then left
CYTU, Kai TlVaS oi\iyous EVpicrKOUO"l desolate; but they found in the temple
TWV f\611vaiwv EV T~ ip~ EOVTas, some few Athenians, temple-stewards
I ,._, C - \ I )
Taµ1as TE TOU 1pou Kal 1TEV11Tas av- and needy men, who defended themsel-
6pc.0TIOVS, o'i <ppa~aµEVOl TT)v CxKp6- ves against the assault by fencing the
1TOA1V 6vpl)cri TE Kai ~UAOlO"l f)µuvov- acropolis with doors and logs; these had
1
TO TOVS ElTlOVTas, &µa µEv V1T acr6E- not withdrawn to Salamis, partly by rea-
VEi11s 13iou ouK EKXwpi)cravTES ES :La- son of poverty, and also because they
i\aµ iva, Tipos 8E auToi 8oKeovTES E- supposed themselves to have found out
~Evp11Keva1 TO µavTi)1ov TO Ti nuei11 the meaning of the Delphic oracle that
crqn EXPll<YE, TO ~vi\1vov TElXOS av6:- the wooden wall should be impregnable,
i\c.vTOV EcrEcr6a1 · auTo ofi TOVTO elvai and believed that this, and not the ships,
To KpTJcrq>vyETov KaTa To µavT'fl1ov was the refuge signified by the prophecy.
Kai OU TCxS VEas. oi OE nEpcra1 i~6- The Persians sat down on the hill over
) \ ' , - >
µEVOl ElTl TOV KOTOVTlOV TTJS aKpo- against the acropolis, which is called by
TI6i\1os oxeov, TOV f\611va1o1 Kai\E- the Athenians the Areopagus, and be-
OUO"l f\pi)1ov TI6:yov, ETioi\16pKEov Tp6- sieged them by shooting arrows wrap-
1TOV TOt6v8E. OKWS CYTUlTlTEloV lTEpi ped in lighted tow at the barricade.
TOVS OlCYTOVS 1TEp16EVTES O\l)ElaV, ETO- There the Athenians defended themsel-
~EUOV ES To q>p6:yµa . Ev6avTa f\611- ves against their besiegers, albeit they
vaiwv Ol Tioi\1opKEOµEV01 oµc.vs f)µu- were in extremity and their barricade
,
VOVTO, KallTEp ,
ES TO' ECYXaTOV
,,
KaKOU- had failed them; nor would they listen to
, ,
aTI1yµEVOl Kal' TOU q>payµaTOS
, - 1Tp0- the terms of surrender proposed to them
0E8WKOTOS. OUOE i\6yous TWV nE1<Jl- by the Pisistratids, but defended themsel-
CYTpaT18ewv lTpO<Yq>EpOVTWV lTEpi 6µ0- ves by counterdevices, chiefly by rolling
i\oyi11s EVEbEKOVTO, aµuv6µEV01 OE &i\- great stones down on the foreigners
Aa TE avTEµ11xavwvTo Kai 5Ti Kai when they assaulted the gates; insomuch
Tipocr16vTwv Twv 13ap136:pwv Tipos that for a long while Xerxes could not
TCxS lTUi\as oi\otTpOXOVS CxlTiEcrav, take the place, and knew not what to do.
WCYTE ZEP~llV ElTl XPOVOV cruxvov CxlTO- But at the last in their quandary the
pil)<Jl EVEXEcr601 ou 8uv6:µEvov crq>eas foreigners found an entrance; for the
€i\Eiv. xp6v~ o' EK TWV aTI6pwv Eq>6:v11 oracle must needs be fulfilled, and all the
oi) TtS E~ooos Toicr1 !3ap136:po1m · EOEE mainland of Attica be made subject to
yap KaTa To BE01Tpo1T1ov Tiacrav the Persians. In front of the acropolis,
Tiiv f\TTtKTiv Tiiv Ev T~ i)TIEip~ yE- and behind the gates and the ascent
vecr6a1 UlTO nep<JT)<Jl. Eµ1Tpocr6E WV thereto, there was a place where none
Tipo TT)s aKpoTI6i\1os. 0TI1cr6E OE Twv was on guard and none would have
lTVi\EWV Kai TT)S av6oou, T~ ofi OVTE thought that any man would ascend that
TlS Eq>vi\acrcrE oth' &v T)i\TI1crE µi) KOTE way; here certain men mounted near the
261
APPENDIX I
TlS KaTCx TaUTa ava~airi av6pWTIWV, shrine of Cecrops' daughter Aglauros,

TaUTT) ave~ricrav TlVES KaTCx TO ipov though the way led up a sheer cliff.
TflS KeKpOTIOS evyaTpos 'l\yi\avpov, When the Athenians saw that they had
I > t >
I -
KalTOl TIEP aTIOKpriµvov EOVTOS TOV ascended to the acropolis, some of them
xwpov · ws 8€ El8ov aUTOVS ava~E cast themselves down from the wall and
~flKOTas oi 'l\6rivai01 ETii TT)v aKpo- so perished, and others fled into the
TIOAlV, o'i µEv EppiTITEOV EWVTOVS inner chamber. Those Persians who had
KaTCx TOV TEiXEOS KCxTW Kai 8tE<p6Ei- come up first betook themselves to the
pOVTO, o'i 8€ ES To µeyapov KaTE- gates, which they opened, and slew the
<pEvyov . TWV 8€ nEp<rEWV oi ava~E suppliants; and when they had laid all the
~flKOTES npwTov µh ETpanovTo npos Athenians low, they plundered the temple
TCxS nvi\as, TaUTas 8€ avoi~aVTES and burnt the whole of the acropolis.
TOVS iKETas E<povEvov · ETIEi 8€ crept (Loeb, A. D. Godley)
navTES KaTE<rTpwvTo, To \pov crvi\ii-
cravTES Evenpricrav nacrav TT)v aKp6-
noi\1v .
13. Herod. IX 13 (484-410 B.C.)
('O Map86v1os) ETIEi 8€ ouK (Mardonius) when he could not move

ETIEt6E, nv66µEvos TICxVTa i\6yov, npiv them, and learnt all the truth of the mat-
f) TOVS µETCx navcraviEW ES TOV 'lcr6µov ter, he drew off from before Pausanias'
Ecr~ai\Eiv, VTIE~EXWPEE Eµnpi)cras TE army ere it entered the Isthmus; but first
TCxS f\6i)vas , Kai El KOU Tl 6p6ov fiv he burnt Athens, and utterly overthrew
Twv TE1xewv ii Twv oiKriµcnwv ii and demolished whatever wall or house
Twv ipwv, navTa KaTa~ai\wv Kai or temple was left standing. (Loeb, A. D.
O"VYXW<ras. Godley)
14. Aristoph. Lys. 1150 seq. (411 B.C.)
J\ YLILTPATH: .. OuK 1cr6' off uµas LYSISTRATA: ... Do ye not mind, when ye
o\ /\aKWVES au61s au I KOTWVCxKOS I Wore skirts of hide, how these Laco-
cpopovvTas EA6ovTES 8opi I noi\Aovs nians came I And stood beside you in the
µEV av8pas 8ETTai\wv clTIWAE<rav I fight alone, And slew full many a stout
noi\Aovs 8' ETa{povs 'lnniov Kai ~vµ Thessalian trooper, I Full many of Hip-
µaxovs ... pias' s friends and helpers ... (Loeb, Ben-
jamin Bickley Rogers)
262
APPENDIX I
15. Aristot. Ath. Const. XIX 5-6 (384-322 B.C.)
(Qi /\aKc801µ6v101) .. . Ki\coµEv11v E~E ... they dispatched their king Cleomenes
TIEµ\J)av TOV l3acr1i\Ea crT6i\ov EXOVTa by land with a larger army; he won a
µEi(w KaTo yliv, 8s eirci Tous Twv victory over the Thessalian cavalry who
0ETTai\wv l1T1TEIS EViKTJO"EV KWAVOV- tried to prevent his reaching Attica, and
Tas auTov Eis Tfiv :ATT1K'fiv irap1E- so shut up Hippias in the fortress called
va1, KaTaKi\cicras TOV 'liririav Eis TO the Pelargicon and began to lay siege to
Kai\ovµEVOV nci\apy1KOV TElXOS E1TO- it with the aid of the Athenians. While
i\16pKEl µETO TWV :A611vaiwv. irpocr- he was sitting down against it, it occurred
Ka611µEvov 8' auTov crvVE1TEO"EV V1TE- that the sons of the Peisistratidae were
~16vTas ai\wvai TOVS TWV nElO"lO"Tpa- caught when trying secretly to get away;
Tl8Wv vic'is · wv AT)q>6EvTwv 6µoi\o - and these being taken they came to
yiav eiri Tfj TWV irai8wv O"WTTJpiq terms on the condition of the boys'
iro111craµEvo1 Ka\ TO EavTwv ev 1TEv6' safety, and conveyed away their belong-
fiµEpms EKK0µ1craµcv01 irapE8wKav T'fiv ings in five days, surrendering the Acro-
aKp6iroi\1v Tots :A611vai01s eiri :t\pira- polis to the Athenians; this was in the
KTi8ov OPXOVTOS ... archonship of Harpactides ... (Loeb, H.
Rackham)
16. I. Bekker, Anecd. Gr. I p. 419, 1. 27 (Cleidemos, middle of 4th cent. B.C.)
"Aire8ov · TO icr6irc8ov Kai TO 6µa- "Airc8ov: that which is even and level.
i\6v. 0ovKv8i811s. TO icr6irc8a . Ki\ci- Thucydides: The level grounds (icr6irc-
811µos · Ka\ fiirE81(ov T'fiv aKp6iroi\1v, 8a). Cleidemus: they levelled the Acrop-
1TEplEl3ai\i\ov 8E EVVECx1TVAOV TO nc- olis and surrounded it with the Pelar-
i\apy1K6v . gicon of the Nine Gates.
17. Parian Chronicle 1. 60 (264/3 B.C.) (ed. Jacoby, Berlin 1904)
aq>' IOU :t\pµ681os Ka\ [:AplO"TOYE Ji- Since Harmodius and Aristogeiton killed
TWV 01TEKTE[ ivav I "lirira ]pxov nH- Hipparchus, successor (?) of Peisistratus,
O"lO"TpCxTOV 8[ ia ]8[ olx]ov (;) Ka\ :A- and the Athenians expelled the descend-
611vaio1 [ E~aVEO"T ]11crav I TOVS nE10"1- ants of Peisistratus from the Pelargicon
0"TpaTi8as EK T[ OU n]ci\alcry1KOV TEi- enceinte, 248 years, in the archonship of
xovs. ETll HHLiLiLiLinlll, I apxovTOS Harpaktides.
:A6i)v11mv :t\[p Jir[ aKTi8ov].
263
APPENDIX I
18. Schol. Aristoph. Lys. 1150
.. .6py1cr6evTes oi /\aKc.vves Ki\eoµevri .. . the Lacedaemonians got angry and

Tov j3acr1i\ea cruv µei~ov1 E~ETIEµ\¥av sent their king Cleomenes with a larger
<rTOA(}> ' Kai v1Ki)cras Tovs 8eTTai\ous force. He defeated the Thessalians and
e1crf1i\6ev eis Tiiv /\TTlKTiv Kai Tov entered into Attica and shut Hippias
'11T1Tiav O'UVEKAflO'EV ElS TO nei\ap- up in the Pelargicon, till the children of
YlKOV Teixos EWS oi Tiai8es Twv Tv- the tyrants were captured. (William G.
pavvc.vv Eai\c.vcrav . Rutherford, MacMillan ed.)
19. Suda s.v. aTie8a and ftTie81~ov (end of lOth cent. A.D.)
"ATie8a. Ta 1cr6Tie8a. Ki\ei8riµos · Kai ,f\1Te8a: Level ground. Cleidemos: And

TtTIEbl~OV Tiiv CxKp01TOAlV, 1Tep1ej3aMov they levelled the Akropolis and built
8e EVVECx1TUAOV TO nei\acry1K6v . around it the Pelargicon of the Nine Gates.
)HTIEbl~OV . 01Te8a yap Ta l0'01Te8a. ,HTie81~ov : aTie8a i.e. level ground. Clei-
Ki\ei8riµos · Kai TtTIE8i~ov Tiiv aKp6Tio- demus: And they levelled the Akropolis
/\1v, Tiepu~ j3aMov 8f: EvveaTiv/\ov To and built around it the Pelargicon of the
n e/\acrytKOV. Nine Gates.
III. THE PELARGICON AFTER THE PERSIAN WARS
20. Cratinos fr. 321 (528-423 B.C.)
Xai'p ' w xpva6Kepc.v j3aj3aKTa Ki)- Hail, revelling he-beast, golden-homed

i\wv, I nav, nei\acry1KOV apyov Eµ- Pan, Whose haunt is in the vacant Pe-
j3aTEVC.VV. lasgicon. (John Maxwell Edmonds,
Brill ed.)
21. Thuc. II 15-17 (470-394 B.C.)
To 8e 1Tpo Tov f1 aKp61Toi\1s il vuv Before this what is now the Acropolis
oucra 1TOAlS fiv Kai TO U1T' atnTiv
J was the city, together with the region at
Tipos v6Tov µai\1crTa TeTpaµµevov. the foot of the Acropolis toward the
TeKµi)p1ov 8E- Ta yap iepa <Ta> Ev south. And the proof of this is as follows:
aUTTJ TTJ CxKp01TOAfl Kai &Mwv eewv On the Acropolis itself are the sanctua-
EcrT1, Kai Ta E~c.v 1Tpos TovTo To µe- ries of the other gods as well as of
pos TTlS TI6i\ec.vs µO:Mov 18puTm, TO Athena, and the sanctuaries which are
264
APPENDIX I
TE TOV ~lOS TOV ,O:Auµniou Kai TO outside the Acropolis are situated more
nve1ov Kai TO TflS rT]s Kai TO <Tov> in that quarter of the city, namely those
ev /\iµvais ~lovvcrou, ~ To apxaio- of Olympian Zeus, of Pythian Apollo, of
TaTa ~10VV0"1a TTJ 8c.v8EKCxTT) TI01- Earth, and of Dionysus in Limnae, in
ElT01 ev µ11vi :Av6EcrT11p1wv1, wcrnEp whose honour are celebrated the more
Kai oi an' l\6nvaiwv "lcuVES ET1 Kai ancient Dionysia the twelfth of the
vvv voµi(oucr1v. 18puTa1 8€ Kai &Ma month Anthesterion, just as the Ionian
( \ I ) - \ - I
1Epa TaUTl) apxa1a. Kal TT) KPTJVT) descendants of the Athenians also are
TTJ vvv µev Twv Tvp6:vvc.vv ovTc.vs wont even now to celebrate it. In that
crKEuacravTc.vv ,EvvEaKpovv~ Ka:Aou- quarter are also situated still other an-
µEVT), To 8€ n6::Aai cpavEpwv Twv nri- cient sanctuaries. And the fountain now
ywv ouawv KaM1p6l) wvoµacrµEVT) called Enneacrunos, from the fashion
EKE'ivoi TE eyyus OU<YT) TO ni\eicrTOV given it by the tyrants, but which an-
&~1a EXPWVTO, Kai vvv ETl ano TOV ciently, when the springs were unco-
apxaiov np6 TE yaµ1KWV Kai ES &i\- vered, was named Callirrhoe, was used
Aa TWV lEpwv voµi(ETal T~ u8aT1 by people of those days, because it was
XPfl0"6al. Kai\Efrat 8€ 810 T'fiV TIO- close by, for the most important cere-
AalOV TaVTT) KaToiK11cr1v Kai Ti aKp6- monials; and even now, in accordance
no:A1s µexp1 TOU8E ET1 vn' ~erivaiwv with the ancient practice, it is still custo-
n6i\15. TTJ 8' ovv eni no:Au KaTo Tiiv mary to use its waters in the rites preli-
xwpav auTov6µ~ OtKflcrci [µETElXOV J minary to marriages and other sacred
Ol ~611vafo1, Kai ETIC18'fi ~UV~Kicr611- ceremonies. And, finally, the Acropolis,
crav, 810 TO E6os EV TOlS aypo1s because the Athenians had there in early
oµc.vs oi TIAEious TWV TE apxaic.vv times a place of habitation, is still to this
Kai Twv vcrTEpov µexp1 Tov8E Tov day called by them Polis (or city). Because,
no:Aeµou yEv6µEvoi TE Kai o1Kflcrav- then, of their long-continued life of in-
TES, OU pq8ic.vs naVOlKTJO"iq TOS ava- dependence in the country districts, most
O"TCxO"ElS eno1ovvTo, &Mc.vs TE Kai of the Athenians of early times and of
apTl CxVEli\TJq>OTES TOS KaTaO"KEUOS their descendants down to the time of
µETO TO M1181K6:- ej3apvVOVTO 8€ Kai this war, from force of habit, even after
xa:AEnws Eq>Epov o1Kias TE KaTaAEi- their political union with the city, con-
novTES Kai iEpo & 810 navTos Tjv
) _, - \ -\)
aUTOlS EK TTJS KaTa TO apxa1ov TIO-

tinued to reside, with their households
in the country where they had been
AlTEias n6:Tp1a, 8ian6:v TE µeMov- born; and so they did not find it easy to
TES µETaj36:Mciv Kai ou8€v &Mo ft move away, especially since they had
n6:A1v Tfiv auTov ano:AEinc.vv EKaCYTos. only recently finished restoring their es-
enci8Ti 8€ acpiKOVTO ES TO OO"TU, 6:Ai- tablishments after the Persian war. They
y01s µEV TlO"lV VTiflPXOV OtKflO"ElS Kai were dejected and aggrieved at having to
265
APPENDIX I
lTapo <pii\wv TlVOS Ti OlKEiwv KaTa- leave their homes and the temples which
cpuyi;, oi 8e TioMoi Ta TE Epf1µa TflS had always been theirs, - relics, inherited
TIOAEWS 4)KT)crav Kai TO iEpo Kai TO from their fathers, of their original form
ilp~a lTCxVTa lTAiiv TflS CxKpOlTOAEWS of government - and at the prospect
Kai TOV 'Ei\Eucr1viou Kai Ei Tl &Mo of changing their mode of life, and facing
~E~aiws KAT)CYTOV ftv· TO TE nEi\ap- what was nothing less for each of them
y1KOV Kai\ovµEvov TO vTio Tiiv aKp6- than forsaking his own town. And when
Tioi\1v, 8 Kai ElTCxpaT6v TE fiv µii they came to the city, only a few of them
OlKElV Kai Tl Kai nu6tKOV µaVTEiou were provided with dwellings or places
CxKpOTEAEVTlOV T010V8E 81EKWAUE, i\f. - of refuge with friends or relatives, and
11
yov WS TO nEi\apytKOV apyov aµEl- most of them took up their abode in the
vov ", oµws VlTO TflS 1Tapaxpf1µa vacant places of the city and the sanc-
avayKT)S E~~Kii6T). tuaries and the shrines of heroes, all
except the Acropolis and the Eleusinium
and any other precinct that could be
securely closed. And the Pelargicum, as
it was called, at the foot of the Acropolis;
although it was under a curse that for-
bade its use for residence, and this was
also prohibited by a verse-end of a
Pythian oracle saying: "The Pelargicum
should better be left vacant" nevertheless
under stress of the emergency was oc-
cupied. (Loeb, C. F. Smith)
22. W. Dittenberger, SJG 3 83

(Decretum de primitiis Eleusina consecrandis, ea. 423/2 B.C.)
[A]aµTiov EllTE ' To µev &Ma Ka6a- Lambon said: The rest according to the
TIEP a\ xcruyypacpai TES CxlTapxEs TO decree regarding the first fruit offerings
KapTio To'iv 6Eo'iv ... Tov 8e ~acr[ 1]- to the gods ... and let the basileus delimit
i\ea hopicra1 To hiEpo To Ev T[o]1 the sanctuaries in the Pelargicon. And let
nEi\apy1KOl, Kai TO AOllTOV µE Evh1- no one in the future found altars in the
8pvEcr6a1 ~oµos EV TOl nEi\apytKOl Pelargicon without the consent of the
avEu TES ~oi\Es Kai TO 8f.µo . µE8€ Council and the People, nor quarry
TOS i\i6os TEµVEV EK TO nEi\apytKO stones from the Pelargicon or carry out
µE8€ y'Ev €xcrayEv µE8€ i\ieos. €ov 8€ soil or stones. And if someone trans-
266
APPENDIX I
TlS napa'3aivE1 T:OVTOV Tl CxTIOTl- gresses the above let him pay 500
VETO ncvTaKoaias opaxµas, Eaayyc- drachmas and let the basileus impeach
AETo OE h[ o J '3acr1AcVS ES TEV '3oAEV. him to the Council.
23. Aristoph. Av. 832-836 (414 B.C.)
EYE/\nl~HL : Tis oai Ka6e~El TT)s n6- EVELPIDES: And who shall hold the
Acws TO ncAapy1K6v; citadel's Storkade? (Pelargicon)
EnO't': "Opv1s acp' fiµwv TOV yevous HooPoE: A bird of ours, one of the
TOV ncpcrlKOV, OO"Ticp AEycTal OElVO- Persian breed, Everywhere noted as the
TaTOS clva1 navTaxov 'J\pcws vcoT- War-god's own Armipotent cockerel.
TOS. EVELPIDES: 0, Prince Cockerel! Yes,
EYE/\n.: ... n vcoTTE 8eanoTa· ws 8' He's just the God to perch upon the
6 6c6s ETI1Tfi8c1os oiKciv Eni ncTpwv. rocks. (Loeb, Benjamin Bickley Rogers)
24. Luc. Pesc. 42 (120-200 A.D.)
<l>l/\OLO<l>IA: Ba~ai, ws TIAflPTlS µEv PHILOSOPHY: Aha! What a lot of them!

Ti &vo8os c.061,oµevwv, ETici Tas 8uo The road up to the gate is full of men
µvas flKOucrav µ6vov. napa OE TO nc- hustling after the two minas, as soon as
Aacry1KOV &M01 Kai KaTa TO 1\crKAT}- they heard of them; others are coming
Tilciov ETcpo1 Kai napa Tov "ApEiov up beside the Pelasgicon; others by the
nayov ETl TIAcious, EV101 OE Kai precinct of Asclepius; even more of them
KaTCx TOV TOV TaAW TCx<pOV, oi OE along the Areopagus; some, too, by the
Kai npos To 1\vaKciov npoa6eµcvo1 tomb of Talus; and some have set lad-
KA{µaKas avepnoucr1 '30µ'31186v vi) ders against the temple of the Twin
~ia Kai '3oTpu8ov Ecrµov 8iKT}V, 1va Brethren and are climbing up with a
Kai Ka6' "QµT}pOV clTIW, aMa Ka- hum, by Heaven, and "in clusters" like
Kciecv cv µaAa noMoi KCxVTcv6cv ... swarming bees, to use the words of Ho-
mer; from that side right many, as also
from the other ... (Loeb, A. M. Harmon)
25. Luc. Pesc. 47-48 (120-200 A.D.)
<l>l/\OLO<l>IA: Ti npaTTElV avi)p 81a- PHILOSOPHY: What does the man intend
VOE1Ta1; to do?
IEPEIA: ~cAcaaas To &YK1aTpov iaxa- PRIESTESS: Baiting the hook with the fig
81 Ka\ T~ xpucri~ Ka0c,6µcvos ETii and the gold, and taking his seat on the
267
APPENDIX I
To aKpov Tov Teixiov Ka6T)KEV ES crest of the wall, he has made a cast into
TT)v n6A1v. the town!
<Dl/\OLO<DIA: Ti TOVTO, w nappTJ- PHILOSOPHY: Why are you doing that,
cna8ri, lTOlEis ; T) lTOV TOUS /dBovs a- Parrhesiades? Have you made up your
AlEVO'ElV 8u~yvwKas EK TOV nEAO- mind to fish up the stones out of the
-
cry1Kov ;... Pelasgicon? ...
('O nappTJO'lObTJS, OAlEVO'OS EVO EK (Parrhesiades pulls up a philosopher and
TWV q>lAOO'Oq>WV, AEyE1:) says:)
Ti TOVTO, w YEVVOlOTOTE ; EiATJYJOl How about it, my fine fellow? Caught,
AlXVEvwv TIEpl TOS TIETpas, EvBa Ai)- were you, gormandizing about the rocks,
crE1v TiATilO'OS UlTObEbVKws: where you hoped to slip under cover and
keep out of sight? (Loeb, A. M. Harmon)
26. Luc. Bis Acc. 9 (120-200 A.D.)
EPMHL :.. .. 'AMa µETO~U Aoywv fi8ri HERMES: ... But in the course of our talk
TIATJ<rla~oµEv T~ 'ATTlK~ · wcrTE TO we are already drawing near to Attica, so
µev Lovv1ov €v 8e~1q KaTaAhrwµev , let us leave Sunium on our right, and
ES 8e TT)v aKp6no::\1v anovEvwµEv now let us glide down to the Acropolis.
Ti8ri. Kal ElTEllTEp KOTal)El)i)KaµEv , au- Now that we have alighted, you sit down
TT) µev EVTavea lTOV ElTt TOV nayov here on the Areopagus somewhere,
KaBricro ES TT)v nvvKa 6pwcra Kal facing the Pnyx, and wait until I cry out
nEp1µevovcra EcrT' &v KTJpu~w Ta the proclamation from Zeus. If I climb
napa TOV Ll16s, EYW 8e ES TT)v Cx- the Acropolis it will be easier for me to
KpOlTOAlV ava~as paov OVTWS &irav- summon everybody from that point of
TOS EK TOV ETITJK6ov npocrKaAEcroµai. vantage for the voice.
LllKH : MT) npoTEpov aneA6T)s, w 'Ep- JUSTICE: Don't go Hermes, until you
µfl , nplv ElTIEiv OO'TlS OVTOS 6 npo- have told me who comes here, the
I > < I ( \
O'lWV EO'TlV, 0 KEpacrq>opos, 0 TTJV person with the horns and the shepherd's
crup1yya, 6 AOO'lOS EK TOlV O'KEAOlV . pipe and the hairy legs.
EPMHL: Ti q>i)s : ayvoEiS TOV nova, HERMES: What! Don't you know Pan, the
TWV Ll1ovvcrov 6Epan6vTwv Tov j3aK- most bacchanalian of the servants of
I 7 >I \ \ I
XlKWTaTov ; ovTOS ~KEl µEv TO npo- Dionysus? He formerly lived on Parthe-
cr6Ev ava TO nap6ev1ov, UlTO 8E TOV nion, but when Datis approached by sea
LlaT18os EninAovv Kal TT)v Mapa- and the barbarians landed at Marathon,
Bwva8E TWV j3apj3apwv an61)amv he came unasked to fight on the side of
i)KEV aKATJTOS Tois 'ABrivaio1s ~uµµa the Athenians; and since then, accepting
xos, Kal TO an' EKEivov TO UlTO T~ this cavern under the Acropolis, a little
268
APPENDIX I
CxKpOTIOAEl crrri}i\aiov TOVTO arroi\a- above the Pelasgicon, he lives in it, pay-
~6µEVOS oiKEi µ1Kpov vrrEp Tov nE- ing the usual tax as a resident alien.
i\acry1Kov ES To µEToiK1ov crvvTE - (Loeb, A. M. Harmon)
i\wv .. .
27. Philostr. Vit. Soph. II, a, V (ea. 200 A.D.)
.. .KaKEiva TIE pi TWV nava611vaiwv Moreover, I have been told the following
TOVTWV T;Kovov· TIETIAOV µEv avfl- facts concerning this Panathenaic fest-
cp6ai TllS VEWS i)8iw ypacpfls ~uv ival. The robe of Athene that was hung
ovpi~ T~ KOATI~ , 8paµEiv bE Tiiv on the ship was more charming than any
va\Jv OVX VTIOsuyiwv ayoVTWV, ai\i\' painting, swelling before the breeze, and
vrroyEio1s µ 11xavais ETIOA1cr6avovcrav' the ship, as it took its course, was not
EK KepaµElKOV 8e apacrav x1i\iq KW- hauled by animals, but slid forward by
TIT) acpEiva1 ETii TO 'Ei\Evcriviov Kai means of underground machinery. Setting
TIEp1~ai\ovcrav avTo rrapaµEi~a1 To sail at the Cerameicus as if with a thou-
nEi\acrylKOV Koµ1soµsv11v TE rrapa TO sand rowers, it arrived at the Eleusinium,
nueiov EA6Eiv , Ol vvv wpµl<rTal. and after circling it, passed by the Pe-
lasgicum: and thus escorted came by the
Pythium, where it is now moored. (Loeb,
Wilmer Cave Wright)
28. Schol. Aristoph. Av. 836 (Didymus: 63 B.C.-10 A.D.)
'Qs 8' 6 6Eos KTA: lii8vµos cp11cri To 'Qs 8' 6 6eos KTA: Didymus says that
nEi\apylKOV TElXOS ETii TIETpwv KEl- the Pelargic wall rests upon a rocky foun-
cr6a1. dation. (William G. Rutherford, Mac-
Millan ed.)
29. Pollux Onom. VIII 101 (2nd cent. A.D.)
... OVTOl rrapEcpvi\aTTOV µi} TlS EVTOS they watched closely that nobody
TOV nEi\acry1KOV KEipEl ii KaTCx TIAEOV would crop the plants nor dig in the
E~OpVTTEl , Kai T~ apxovTl rrapE8i- Pelasgicon and turned the offenders over
8ocrav. To 8E Tiµ11µa f}v TpEis 8pax- to the magistrate. The fine was three
µai Kai arri\ovv To ~i\a~os . drachmas and the damage considered to
be limited.
269
APPENDIX I
30. Schol. Luc. Bis Ac. 9
... TOTIOS 'A6iiv11cnv , clTIO nei\acrywv ... a place in Athens, after the Pelasgians
ev avTi.p oiK11cravTc.vv . ypacpETat Kai who dwelled there. Spelled also with
81a Tou p. an r.
31. Hesych. s.v. nei\apytK6v (5th cent. A.O.)
n EAapytKOV . clVTi TOU n EAacrytKOV . Pelargicon: instead of Pelasgicon; They

nei\apyous yap cpacrt TT)V 1XTTtKT)V say that the Pelargoi (storks) dwelled in
OtKilcrat · clTIO TWV nei\aaywv µETa - Attica; transferring it (the name) from
q>EpOVTES ETii TCx TITTJVCx. the Pelasgoi to the birds.
32. Eust. Thes. Schol. Dion. Per. 347 (end of 12th cent. A.O.)
. . . EcrTt 8E Kai ev 1\6T,va1s Tei'xos ... there is a wall in Athens, Pelasgicon

nei\acrytKOV 'fiTOl nei\apytKOV, WS 6 i.e. Pelargicon as the comic poet states in
Kc.vµ1KOS bT\Aoi ev Tois "Opvicnv, oTa his Birds, alluding, he says, to the Pe-
TWV nei\aaywv, cp11cri, Kai\ouµevc.vv lasgians, called also by the inhabitants
Trap ' 1XTTtKOlS Kai nei\apywv, 8ta TO of Attica Pelargoi, (storks) for being
TIi\aVTJTlKOV . migrants.
IV. THE NINE GATES
33. K. Miiller, FHG III p. 131: Polemon fr. 49 (210-170 B.C.)
Tils 8E noµnils TaVTTJS 'Hcruxi801, 8 This procession [is led] by the Hesy-
8T) yevos ecrTi nEpi Tas LEµvas 6eas, chidae who are the clan in charge of the
Kai TT)v i)yEµoviav EXEL Kai npo6v- worship of the Eumenides. And before
ovTa1 npo TllS 6ucrias Kp1ov 'Hcrvxl}> the sacrifice they offer a ram to the hero
iepov fip(.}> , TovTov ovTc.v Kai\ouvTES Hesychos whom they call thus as a
8ia TT)v Evcp11µiav· ov To iep6v EcrTt compliment; whose sanctuary is next to
napa To Kui\wvetov, EKTOS Twv 'Ev- the Cyloneion, oustside the Nine Gates.
vea nui\wv.
270
APPENDIX I
34. Schol. Soph. Oed. Col. 489
("Hcruxos 'fipc.us) ... ov To 1Ep6v EcrT1 (the heros Hesychos) ... whose sanctuary
napa To Kui\wvE1ov EKTOS Twv €v- is next to the Cyloneion outside the Nine
vEa nui\wv . Gates.
V. VARIA ON THE ACROPOLIS
35. Hom. Od. 7, 80-81
('H J\611va) 1KETo 8' ES Mapa6wva She (i.e. Athena) came to Marathon and
Kai Evpuayuiav J\6i}v11v. I 8vvE 8' broad-wayed Athens, I and entered the
,EpEx6flos TIUKlVOV 86µov . well-built house of Erechtheus. (Loeb,
A. T. Murray)
36. Herod. VIII 55 (484-410 B.C.)
"EcrT1 EV Tfj aKpon6i\1 TaVTTJ ,EpE- On that acropolis there is a shrine of

X6Eos TOV Y11YEVEOS AEyoµEvou Elva1 Erechtheus the Earthbom (as he is
v116s, Ev Tq, €i\ai11 TE Kai 6ai\acrcra called), wherein is an olive tree, and a
EVl, TO i\6yos napa J\611vaic.uv no- salt-pool, which (as the Athenians say)
O"E18Ec.uva TE Kai J\611vai11v Epicrav- were set there by Poseidon and Athene
Tas nEpi TllS xwp11s µapTvp1a 6E- as tokens of their contention for the land.
cr601 . (Loeb, A. D. Godley)
37. Apollod. Bib/. III 178 (180-109 B.C.)

7
HKEV ouv TIPWTOS nocrE18wv ETii So Poseidon was the first that came to
Ti)V J\TTlKT)v Kai ni\i}~as Tfj Tp10iv11 Attica, and with a blow of his trident on
KaTa µEcr11v TT)v aKp6noi\1v anEq>TlVE the middle of the acropolis, he produced
6ai\acrcrav, ftv vvv ,EPEX611i8a Ka- a sea which they now call Erechtheis. After
i\ovcrl. µETO 8E TOVTOV fiKEV J\611va, him came Athena, and, having called on
Kai no111craµEV1l TllS KaTaAii\J'EC.US KE- Cecrops to witness her act of taking pos-
Kpona µapTvpa Eq>vTEvcrE EAaiav, ft session, she planted an olive tree, which is
vvv EV Tq, nav8pocrEi~ 8EiKVUTal. .. still shown in the Pandrosium. (Loeb, Sir
James George Frazer)
271
APPENDIX I
38. Paus. I 26, 5 (2nd cent. A.D.)
"EcrT1 8E Kai oiKriµa 'EpEX6Hov Ka- There is also a building called the
i\ovµEvov . . . u8c.vp EO"Tiv EVbOV 6a- Erechtheum ... sea-water in a cistern ...
i\acrcr1ov EV q>pEaT1 . . . Kai Tp1aivris On the rock is the outline of a trident.
EO"Tiv EV Tfl lTETpq crxilµa · TaUTa 8E Legend says that these appeared as
AEYETOl nocrE18wv1 µapTvp1a ES Tiiv evidence in support of Poseidon's claim
aµq>1cr~iJTTlO"lV TllS xwpas q>avilval. to the land. (Loeb, W. H. S. Jones)
39. Paus. I 27, 2 (2nd cent. A.D.)
nepi 8E TllS EAaias OUbEV EXOVO"lV About the olive they have nothing to say
&Mo EllTElV Ti Tfl 6Ec'.;) µapTvplOV except that it was a token the goddess
YEVE0-601 TOUTO ES TOV aywva TOV produced when she contended for their
) \ ,..,, I
ElTl TT) xc.vpq . land. (Loeb, W. H. S. Jones)
40. Clemens, Protr. III 45 (160-220 A.D.)*
... f\6iwrimv 8E Ev aKpoiroAEl KEKpo- ... and at Athens, on the Acropolis, is

iros (£vv. tacpoc;). ws cpricr1v :AvTioxos that of Cecrops, as Antiochos says in the
EV Ti;) EVCxT~ TWV 'lcrTop1wv. ninth book of his Histories.
41. Arnobius, Adv. nat. VI 6 (beginning of the 4th cent. A.D.)*
In historiarum Antiochus nano Athenis in Antiochos, in the ninth book of his His-
Minervio memorat Cecropem esse man- tories, relates that Cecrops was buried in
datum terrae. the temple of Minerva, at Athens.
42. Theodoretos, 'E}..).. rJvtxwv Ocea1rcvnxiJ naOrJµa:rwv, JrcQi rij~ rwv µaervewv nµij~
H 30 (5th cent. A.D.)*
Kai yap f\6i)vrimv, ws f\vTioxos Ev TT) And in Athens, as Antiochos writes in his
EVCxTT) yeypaq>Ev iO"Topiq, &vw ye Ev ninth book of Histories, there is on the
TT) aKpoir6i\e1 KEKpoir6s €crT1 TCxq>os Acropolis the tomb of Cecrops next to
lTapa TfiV noAlOVXOV auTiJV. the patron goddess herself.
* The source of information common to historian Antiochos the Syracusan who

these three Fathers of the Church is the lived in the Sth cent. B.C.
272
APPENDIX II
THE POTTERY SHERDS
(that date the terraces and the fortification wall)
GROUPl
The sherds were found in a hollow of the rock at the NW corner of the W
terrace wall of terrace III (see Plans 9, 7 and 39, 1) in a fill composed of light
coloured soil and 4-5 stones of medium size, placed in order to even out the
hollow and to prepare it for setting the corner of the terrace wall. The hol-
low was not excavated by Kavvadias, and the contents had thus remained
0 100 200 300
Plan 39. The find-spots of the material dating the terraces and the wall.
275
APPENDIX II
e f
- - - - -- -- -- 10
cm
Fig. 50. Sherds of Group 1 (phot. by N. Tombazis).
undisturbed. The fill contained a mixture of sherds, mostly MH, but with some
earlier and later as well. Published here are the most characteristic, repre-
senting the various categories found (Fig. 50).
a. Rim fragment of an EH shallow bowl. Clay reddish, fine, slip of lighter
colour unevenly applied with a brush. Max. pres. dim. 49 mm., th. of wall
3 mm.
b. Wall fragment of an open vase, probably phiale, similar to the previous.
Clay reddish, unrefined, very micaceous. Exterior painted red, interior
brownish red. Max. pres. dim. 60 mm., th. of wall 4 mm.
276
APPENDIX II
c. Part of the rim of a wide-mouthed pithoid vessel, decorated with horizontal

relief bands, datable in latest EH times. Clay reddish, coarse and unre-
fined, with traces of a thin yellowish-white slip. Max. pres. dim. 95 mm.,
th. of wall 5-6 mm.
d. Fragment of the lower part of a large goblet (kylix) of late MH times. Clay
brownish red, unrefined. Matt grey paint. Max. pres. dim. 65 mm., th. of
wall 4-5 mm.
e. Approximately half of the low foot of a goblet (kylix) of late MH or early
LH times. Preserved at each end is the beginning of the bottom of the bowl
and the beginning of the base. Clay yellowish with greyish wash covered
with a yellowish shiny paint, unevenly applied. Pres. h. 35 mm., pres. diam.
36 mm. Imitation of a yellow Minyan goblet.
f. Low-footed goblet (kylix) with hollow base. Preserved is the beginning of
the bottom. Similar to the previous. Clay reddish, unrefined, paint brown
unevenly applied over the entire surface excepting the hollow of the base.
Pres. h. 35 mm., pres. diam. 34 mm.
g. Part of the discoid base of a LH II vase with a narrow bottom. Clay light
brown, wash reddish. Max. dim. 32 mm.
h. Wall fragment of an open vase, probably a kylix, of early LH III date. Clay
red, refined. Interior thickly painted brown and well preserved, exterior
thin brown unevenly applied and in places worn. Max. dim. 54 mm., th. of
wall 4 mm.
1. Bottom of a kylix of late LH IIIA or early LH IIIB times, coarse fabric.
Clay reddish unpainted, wash likewise, unevenly applied. Walls slightly

curving, almost straight. Pres. h. 41 mm., pres. diam. 65 mm., diam. of stem
20 mm., th. of wall 5-6 mm.
GROUP2
The group comprises a few MH and LH sherds, which were found within
the clean soil on which the stones of the northern supporting wall of terrace
III rested (see Plans 9, 9 and 39, 2). Published are the following diagnostic
samples (Fig. 51):
a. Minute sherd with clay of reddish orange colour, traces of bright red paint
on one side. Datable to early LH times. Max. dim. 24 mm., th. of wall 4 mm.
277
APPENDIX II
b
a
d e
0 10
cm
b. Minute sherd similar to the previous, but from another vase. Max. dim. 14
mm., th. of wall 5 mm.
c. Wall fragment from a large, probably open LH I vase. Clay grey. Interior
has a worn wash of bright yellow. Imitation of Minyan. Max. dim. 32 mm.,
th. of walls 7 mm.
d. Fragment of a strap handle mended from two joining fragments. Carelessly
made. LH I or II. Clay reddish, slip similar, lighter in shade. Pres. 1. 41
mm., cross-section of handle 8-13 mm.
e. Sherd from an early LH vase, clay bright red. Exterior has brownish grey
slip, unevenly applied. Max. dim. 27 mm., th. of wall 5 mm.
GROUP3
These sherds are numerous and were found in the undisturbed dull yellow
clay that sealed the joins between the stones of the N supporting wall of ter-
race V, at the point where it meets the E supporting wall of terrace IV (Plans
278
APPENDIX II
f
e
---------- 10

cm
13, 12P and 39, 3). They were collected from among the lowest courses of the
supporting wall, from places untouched by Kavvadias' excavation. The sherds
published below represent all the categories to which the finds belong (Fig. 52).
a. Sherd from a NL vase, clay unrefined and spongy, colour brownish grey.
The surface shows traces of uneven smoothing by burnishing. Max. dim.
39 mm., th. of wall 7 mm.
b. Rim fragment of an EH shallow bowl. Clay yellowish, slip similar and not
very shiny. Max. pres. dim. 30 mm., th. of wall 3,5 mm.
c. MH sherd, interior surface uneven. Clay reddish. Exterior painted a matt
brownish grey. On this fine horizonal lines in a dilute white. Between the
lines a band of broken zig-zag. Max. pres. dim. 34 mm., th. of wall 4 mm.
279
APPENDIX II
d. Early LH I sherd, markedly curved. Clay grey. Exterior has a dilute yellow
wash, in places worn off. Imitation of Minyan. Max. dim. 38 mm., th. of
wall 4 mm.
e. Minute sherd of reddish clay with traces of bright red paint on both inte-
rior and exterior surrfaces. LH I. Max dim. 16 mm., th. of wall 4 mm.
f. Rim fragment from a small LH I open vase. Clay reddish, painted bright
red inside and out. Max. dim. 32 mm., th. of wall 3,5 mm.
g. Sherd of reddish clay, with relatively shiny slip of the same colour. LH I.
Max. dim. 40 mm., th. of wall 4 mm.
h. Sherd of reddish clay. Exterior painted a rather dull brownish grey. LH I-
II. Max. dim. 23 mm., th. of wall 4 mm.
i. Fragment of the flat base of a fairly large platter of the beginning of LH
IIIA. The side walls are thicker than the base. Clay reddish, slip similar
and fairly shiny. Pres. diam. 55 mm., pres. h. 27 mm.
GROUP4
Group 4 comprises a few sherds, collected from the same place as those
of the previous group, but at a lower level, between stones of the terrace wall
foundations and beneath them (Plan 39, 4). The most representative are the
following (Fig. 53):
a. Wall fragment from a fairly large EH vase, of uncertain shape, slightly

curving. Clay reddish. Paint on exterior red. Interior unpainted. Max. dim.
73 mm., th. of wall 5 mm.
b. Part of the base of an unpainted LH II-III goblet. Clay pale reddish yel-
low. Pres. semi-diameter 29 mm., th. 3,5-5 mm.
c. Fragment of the rim of a small vase, everted obliquely, probably from an
amphoriskos or alabastron, early LH III. Interior brown paint, exterior dull
grey, fading. Max. dim. 20 mm., th. 4-5 mm.
d. Rim fragment of a skyphos with relatively thick walls. Clay pale red. On
the interior the surface is covered by brown paint, thickly and evenly
applied. The exterior is decorated with a panel of clumsily drawn pen-
dent concentric semi-circles bordered by vertical lines framed by a
280
APPENDIX II
0
~ - -- - -- - 10
cm
cross-hatched semi-circle. LH IIIB2. Max. dim. 45 mm., th. of wall

5 mm.
GROUPS
This group comes from the North fortification wall, where it runs beside
the Mediaeval buttress, NE of the Pinakotheke (Plans 22 and 39, 5). A great
many sherds, mainly LH but also some earlier, were recovered from the
refined yellow clay between the stones of the fill and the rock. The sherds
published here represent the various categories found. The latest date the con-
struction of the wall (Fig. 54).
a. Small sherd of a MH vase with slight curvature. Clay greyish brown, with
surface imperfectly burnished. The exterior is decorated directly on the clay
in brownish red matt paint. Max. dim. 34 mm., th. of wall 4 mm.
281
APPENDIX II
-· --, ,.. '
... ~
0 10
b. Handle fragment. Vertical stripe in matt grey along spine. MH. Pres. 1. 51
mm., diam. 12 mm.
c. Low foot of a goblet preserving part of the bottom and the beginning of
the slightly hollow base. Around the stem a narrow relief band. Clay red-
dish, not greatly refined. Exterior covered by bright red paint. LH IIB-
IIIAl : Pres. h. 30 mm., pres. diam. 66 mm.
d. Minute sherd covered inside and out with a thick, shiny, red paint. LH I.
Max. dim. 30 mm., th. of wall 5 mm.
e. Small sherd from a thick-walled vessel. Surface rough and uneven, without
slip. Clay yellowish. Decorated with two parallel lines in brownish black
paint, peeled off in places. LH IIIA-B. Max. dim. 32 mm., th. of wall 5 mm.
f. Fragment of the beginning of the shoulder of a small wide-necked vessel.
Clay reddish. Decoration in thick red paint. Preserved between neck and
shoulder is a trace of a horizontal band and descending from this a thin
282
APPENDIX II
vertical zig-zag column dividing the shoulder into panels decorated with
pendent semi-circles. LH IIIB2. Max. dim. 58 mm., th. of wall 7 mm.
g. Fragment of a large vase, the surface of which is destroyed in places. Clay
fine reddish, slip same but lighter tone, paint orange-yellow. Preserved are
traces of a horizontal band, with a curved band above. LH IIIB. Max. dim.
46 mm., th. of wall 6,5 mm.
h. Fragment of the lip of a jug. Clay rough, yellowish green. Interior of lip
preserves trace of a band in dilute grey paint. Developed LH IIIBl. Max.
dim. 50 mm., th. of wall 7-8 mm.
283
LIST OF ILLUSTRATIONS
PLANS
1. The Acropolis rock.
2. The Acropolis in Neolithic, Early Helladic and Middle Helladic times.
3. Walls between the Propylaia and the Erechtheion area.
4. Walls in the area N and NE of the Erechtheion.
5. Remains of the LH I house N of the Erechtheion.
6. The Acropolis at the beginning of the LH period.
7. Mycenaean and Mediaeval walls NE of the Erechtheion.
8. Remains of terrace walls NE of the Erechtheion.
9. Foundation trench of the W terrace wall.
10. Terrace walls W of the Erechtheion.
11. The area of the Pandroseion and the Kekropion.
12. The complex of terraces S of the Erechtheion and around the Archaic
temple.
13. Terrace walls E of the Erechtheion.
14. The northeast ascent.
15. The area of the northwest descent.
16. The Acropolis at the end of the second LH building phase.
17. The west bastion of the fortification and the area around it.
18. The niche in the W fa~ade of the bastion.
19. Remains and traces of the west fortification wall.
20. The west fortification wall.
21. The north fortification wall from the Pinakotheke to the NW descent.
22. Remains of the north fortification wall beside the big Mediaeval buttress.
23. The north fortification wall from the NW descent to the House of the
Arrephoroi.
24. Remains of the north fortification wall next to the House of the
Arrephoroi.
25. The north fortification wall from the House of the Arrephoroi to the area
NE of the Erechtheion.
285
26. The north fortification wall in the area of the NE ascent.

27. The arrangement of the fortification wall at the point where the NE ascent
ends.
28. Trace of the north fortification wall over the end of the NE ascent.
29. The north fortification wall from the NE ascent to the Belvedere.
30. The east fortification wall.
31. The southeast corner.
32. The south fortification wall, up to the SW corner of the Parthenon, includ-
ing the remains preserved beside it.
33. The south fortification wall from the SW corner of the Parthenon to the
Propylaia.
34. The southwest corner of the fortification.
35. The arrangement of the west entrance.
36. The Pelargikon.
37. The remains of the little Mycenaean houses built over the NE ascent.
38. The Acropolis during the final years of the Mycenaean period.
39. The find-spots of the material dating the terraces and the wall.
FIGURES
(Photographs listed without credits are by the author)
1. The Mycenaean Acropolis according to Koster (Pelargikon pl. IVb ).

2. The arrangement of the area of the NE ascent according to Holland (AJA
1924, II, fig. 2).
3. The Acropolis of late Mycenaean times according to Travlos (I1oAt:o<5.
fig. 7).
4. The large piece of the rock broken off at the NW part of the Acropolis.
5. Wall 22 of Plan 3, from E.
6. The W side of the S column base, showing traces of working with a point.
7. Wall 5 of Plan 4, from E. To the left is 1, to the right, 2 (phot. DAI n° 48).
8. The N end of wall T3, which continues on the preserved top of the Myce-
naean fortification wall (right below the column drums). Above, right, col-
umn drum Q with traces of the wall construction on top of it.
9. Walls T4 and TS of Plan 7, from the S.
10. Walls 2P and 3 of Plan 8, from the N, at their point of junction. In front
286
the inner side of 2p. Wall 3 is next to the wooden stairway (phot. DAI
n° 46).
11. Wall lP of Plan 8 at left, and the foundation of the E porch of the
Erechtheion, from the N. (phot. DAI n° 742).
12. Foundation trench for the terrace wall behind the Promachos base, from
the south.
13. Working of the rock at 2 on Plan 9, from the W.
14. The stones of the terrace wall, in situ, next to S on Plan 9.
15. The west wall of the Erechtheion. At the level of the big marble beam and
to its left, the wall block with the lifting bosses incompletely removed.
16. Terrace wall 12 of Plan 13 from the NW (phot. DAI n° 50).
17. The cuttings on the rock W of the Propylaia, from the S.
18. The stones of the Cyclopean wall next to the Mediaeval buttress, from
above.
19. Successive courses of the Cyclopean wall beside the Mediaeval buttress,
from the W.
20. Stones of the Cyclopean wall on the edge of the rock outside the Classi-
cal fortification wall, W of the North Fountain.
21. The North Fountain within the natural cleft of the Acropolis rock. Sec-
tion looking E (Broneer, Hesperia VIII, 1939, pl. XII).
22. The descent to the North Fountain. Model (Broneer, Hesperia VIII, 1939,
fig. 17).
23. The stepped cuttings in the rock for the foundations of the fortification
wall S of the end of the NE ascent, from the W .
24. The outer face of the SE fortification wall, from the E (phot. DAI n° 80).
25. The E side of the rock, with the cuttings and walls of the Turkish domi-
nation.
26. Walls inside the SE corner of the Mycenaean fortification (phot. DAI
n° 51).
27. The inner side of the SE fortification wall and walls 1, 15, 13 and 14 of
Plan 31 (phot. DAI n° 54).
28. Interior of the SE corner of the Mycenaean fortification wall, from the S,
showing walls 16 and 6 and tomb 18 of Plan 31 (phot. DAI n° 55).
29. The outer, W face of the fortification wall that is preserved S of the Propy-
laia (phot. by P. Mylonas).
30. The Mycenaean column base, from above.
287
31. Side view of the Mycenaean column base.

32. The upper, visible surface of the first slab.
33. The lower, hidden surface of the first slab.
34. The upper surface of the second slab.
35. The lower surface of the second slab.
36. The Pelargikon as the area around the Acropolis, according to Botticher
(Akropolis p. 58, fig. 7).
37. The Pelargikon as the area W and S of the Acropolis, according to Har-
rison (M. and M. fig. 35).
38. The Pelargikon on the basis of the Archaic walls SW of the Acropolis,
according to Miller (AJA 1893, p. 489, fig. 1).
39. The top and S face of the projection of the rock, from above.
40. Cuttings and stones of the Pelargikon in situ, from the W (Plan 36, 2).
41. The oblique cutting at 7 on Plan 36, from the W. Below right, later work-
ing of the rock.
42. The rock above the Klepsydra, from the E. Discernible in the middle of
the photograph is the cutting 14 of Plan 36.
43. Preparation of the rock for the foundations of the terrace walls: the bed
of the trench W of the supporting wall of terrace III.
44. Foundation of the terrace wall directly on the uneven rock without cut-
ting: the NW corner of terrace IV from the W.
45. Block of the inner face of the SE fortification wall, showing traces of ham-
mering on the surface.
46. The foundation of the SW corner of the fortification wall by the Propy-
laia.
47. Reconstruction drawing of the construction of the upper sections of the
descent to the North Fountain (Broneer, Hesperia VIII, 1939, fig. 6).
48. Reconstruction of the third, fourth and fifth sections of the descent to the
North Fountain (Broneer, Hesperia VIII, 1939, fig. 18).
49. The two final sections of the descent to the North Fountain, built on the
rock (Broneer, Hesperia VIII, 1939, fig. 19).
50. Sherds of Group 1 (phot. by N. Tombazis).
288
ABBREVIATIONS - BIBLIOGRAPHY
I. PERIODICALS
AA Archii.ologischer Anzeiger
AbhBerlin Abhandlungen der Deutschen Akademie der WtSsenschaften
zu Berlin
AJA American Journal of Archaeology
AM Mitteilungen des Deutschen Archii.ologische Instituts, Athe-
nische Abteilung
ASAtene Annuario della Scuola Archeologi.ca Italiana di Atene
BCH Bulletin de Co"espondance Hellenique
Bd.A Bolletino d'arte
BerlPhilolWoch Berliner Philologische Wochenschrift
BSA Annual of the British School at Athens
Deltion 'AexawA.oyixov LIElriov
Ephemeris 'AexawA.oyixq 'Ecp17µEeir;
EpistEpetAth 'Emar:17µovixq 'EnH17eir; r:fjr; <PtA.oaocpixfjr; ExoA.ijr; r:ov
IlavEnwr:17µ{ov 'A817vwv
Ergon To "Eeyov rfjr; 'AexawA.oyixfjr; 'Eraiedar;
FuF Forschungen und Fortschritte
HSCP Harvard Studies in Classical Philology
Jdl Jahrbuch des Deutschen Archiiologischen Instituts
JHS Journal of Hellenic Studies
Mem.Aclnscr Memoires de l'Academie des Inscriptions et belles lettres
6Jh Jahreshefte des Oste"eichischen Archii.ologischen Instituts
in Wien
Praktika IIeaxrixa rfj~ iv 'AO~vm~ 'AexawA.oyix1]~ 'Eraieda~
PraktAkAth Ileaxnxa r:fjr; 'Axao17µ{ar; 'A817vwv
RhM Reinisches Museum fiir Philologi.e
SBBerlin Sitzungsberichte der Deutschen Akademie der WtSsenscha-
ften zu Berlin. Kl.asse fiir Sprache, Literatur and Kunst
VHAM Kong/. Vitterhets Historie och Antiquitets Manadsblad
289
II. BIBLIOGRAPHY
Apart from the published excavation reports, which are given in the relevant chapter
( ee pp. 25-29), there is, practically speaking, no special bibliography referring to the sub-
ject and restricted to it. There are, however, many publications that touch on some of the
details or on some of the broader subjects, one of these being the Acropolis of Athens.
These publications, which are of considerable assistance with the information they provide
or because of the viewpoints they expound, comprise the following:
Berard, Stud. Rob.: Berard, J., Le mur pelasgique de l' Acropole et la date de la descente
dorienne, in Studies Presented to D. M. Robinson I (Saint Louis 1951) pp.
135-159.
Beule, L 'Acropole: Beule, E., L 'Acropole d'Athenes I (Paris 1853).
Blegen, HSCP: Blegen, C. W., Athens and the Early Bronze Age of Greece, HSCP, suppl.
vol. I (Cambridge, Mass. 1940) pp. 1-9.
Bohn, Prop.: Bohn, R., Die Propyliien der Akropolis zu Athen (Berlin - Stuttgart 1882).
Botticher, Akropolis: Botticher, A., Die Akropolis von Athen (Berlin 1888).
Broneer, 0 ., Plato's Description of Early Athens in the Light of Archaeological Research,
AJA 45, 1941, p. 92.
» What Happened at Athens, AJA 52, 1948, pp. 111-118.
Broneer, A ntiquity 1956: Broneer, 0., Athens in the Late Bronze Age, Antiquity 1956, pp.
9-18.
Bundgaard, Mnesikles: Bundgaard, J. A., Mnesikles, A Greek Architect at Work (Copen-
hagen 1957).
Cavvadias, P., Note sur les fouilles de l'Acropole d'Athenes, BCH 1896, pp. 382-383.
Curtius, E., Eleusinion und Pelargikon, Sitzungsberichte der Preussischen Akademie der Wis-
senschaften , Berlin 1884, pp. 499-512, and Gesammelte Abhandlungen I
(Berlin 1894) pp. 435-450.
Curtius, Stadtgesch.: Curtius, E., Die Stadtgeschichte von Athen (Berlin 1891 ).
Curtius, E. - Kaupert, J. A., Atlas von A then (Berlin 1878).
Daniel, J. F., A Mycenaean Fountain on the Athenian Acropolis, by 0. Broneer (book
review), AJA 44, 1940, pp. 552-559.
Dinsmoor, W. B., The Date of the Older Parthenon, AJA 38, 1934, pp. 408-448.
Dirlmeier, Die Pelasgermauer: Dirlmeier, F., Die Pelasgermauer der Akropolis, in J. D.
Plassmann, Kleine Kostbarkeiten der Kunst (Berlin 1940) pp. 37-43.
D'Oodge, Acropolis: D'Oodge, M. L., The A cropolis of Athens (New York - London 1908).
Dorpfeld, W., Der Alte Athenatempel auf der Akropolis, AM 1876, pp. 337-351.
» Die Propylaen der Akropolis zu Athen, AM 1885, pp. 131-144, pl. II.
» Uber die Ausgrabungen der Akropolis, AM 1886, pp. 162-169.
» Das Alte Athen von Theseus, Rheinisches Museum far Philologie 1896, pp.
127-137.
290
Dorpfeld, W., Zu den Bauwerken Athens, VIIL Die pelasgische Mauer der Akropolis,
AM 1911, pp. 71-72.
Dorpfeld, Alt Ath.: Dorpfeld, W., Alt Athen und seine Agora, 1 (Berlin 1937).
Elderkin, G. W., The Cults of the Erechtheion, Hesperia X, 1941, pp. 113-124.
Erechtheum: Stevens, G. P., Paton, J. M., Caskey, L. D., Fowler, H. N ., The Erechtheum
(Cambridge, Mass. 1927).
Frazer, Paus.: Frazer, J. G., Pausanias 's Description of Greece II (London 1913) pp. 355-
359.
Furtwangler - LOschke, M. V.: Furtwangler, A. - LOschke, G., Mykenische Vasen (Berlin
1886),
Furumark, MP: Furumark, A., The Chronology of Mycenaean Pottery (Stockholm 1941)
p. 77.
Furumark, A., The Mycenaean III C Pottery and its Relation to Cypriote Fabrics, Opu-
scula archaeologi,ca 3, 1944, pp. 194-265.
Graef, B., Uber die allgemeinen Ergebnisse der Vasenfunde von der Akropolis zu Athen,
Ber!Phi!Woch 1893, pp. 254 ff.
Graef-Langlotz: Graef, B. - Langlotz, E., Die antiken Vasen van der Akropolis zu Athen I
(Berlin 1925).
Harrison, P.A.: Harrison, J. E., Primitive Athens as Described by Thucydides (Cambridge
1906).
Harrison, M. and M.: Harrison, J. E. - Verrall, M., Mythology and Monuments of Ancient
Athens (London 1890).
Hauser, Strena Helbigiana: Hauser, E., Der Bauder Akropolismauer, Strena Helbigiana pp.
115-121.
Heberdey, R., Das Westtor der Pelasgerburg von Athen, 6Jh XIII, 1910, pp. 1-4.
Hill, Athens: Hill, I. T., The Ancient City of Athens (London 1953).
Holland, I. B., The Strong House of Erechtheus, AJA 1924, pp. 142-169.
» The Hall of the Athenian Kings, AJA 1939, pp. 289-298.
lakovidis, S. E., Al MVXrJVaixai axeon6At:l~ (Athens 1973) 113-140.
» Vormykenische und Mykenische Wehrbauten, Archaeologi,a Homerica , E 1
(Gottingen 1977) 193-204.
» Late Helladic Citadels on Mainland Greece, Monumenta Graeca et Romana
IV (Leiden 1983) 73-90.
Jahn - Michaelis, A1X Ath.: Jahn, 0. - Michaelis, A., A1X Athenarum a Pausania descripta
(Bonn 1901).
Judeich, Top.: Judeich, W., Topographie van Athen (Handbuch der Altertumswissenschaft
III, 2, 2, 2a, Milnchen 1931).
Karo, G., Archaologische Funde im Jahre 1908, Griechenland, AA 1909, p. 106.
Kavvadias, P., 'Avaox.acpil £v i:n 'Ax.gmt6A.Et, Ephemeris 1886, pp. 73-82.
Kavvadias, IIeofrn:oetx~ 'Aexaw).oyia : Kavvadias, P., IIeoiar:oetx~ 'Aexaw).oyia (Athens
1909).
291
Kavvadias-Kawerau: Kavvadias, P. - Kawerau, G., 'H avaaxacp~ rfj~ 'Axeon6A.Ew ~ a.no rov
1885 µ t xet rov 1890 (Athens 1906).
Kontoleon, N.: Kontoleon N., To 'Eet xfJEwv w~ olxoo6µr]µa xOov[a~ A.areda~ (Athens 1949).
Koster, Pelargikon: Koster, A. , Das Pelargikon (lecture), AA 1909, pp. 549-551.
Leake, Topography: Leake, W. M., The Topography of Athens2 (London 1841).
Lechat, H ., Les fouilles de l'Acropole, BCH 1888, pp. 244-245.
Lolling, Topographie: Lolling, H. G., Topographie van Athen (Handbuch der Altertum-
swissenschaft III, Nordlingen 1889).
Marinatos, S. - Hirmer M., KefJrrJ xai MvxrJvaix~ 'EA.A.a~ (Athens 1959).
Martin, R ., L 'urbanisme dans la Grece antique (Paris 1956).
Middleton, JHS Suppl.: Middleton, J. H . (ed. by E. A. Gardner), Plans and Drawings of
Athenian Buildings, JHS Suppl. 3, 1900.
Miller, W. , A History of the Acropolis of Athens, AJA 8, 1893, pp. 473-554.
Montelius, VHAM: Montelius, 0., Ett fynd fr an Athens Akropolis, Kong/. Vitterhets His-
torie och Antiquitets Manadsblad 1889, pp. 49-60.
Montelius, La Grece: Montelius, 0 ., La Grece preclassique I (Stockholm 1924).
Morgan, C. H., The Terracotta Figurines from the North Slope of the Acropolis, Hespe-
ria IV, 1935, pp. 189-213.
Mountjoy: Mountjoy, P. A., Mycenaean Athens (Jonsered 1995).
Mylonas, G. E. , Athens and Minoan Crete, HarvStClPhil, Suppl. vol. I (1940) pp. 11-36.
Mylonas, G., ot f3amf..LXOL -cacpm: ot f3amf..LXOL 'tacpOL 'tWV Muxrivwv xal ~ 'A8riva·i:xi] Jta-
gaoomc;, 'Enli'uµf3wv Xe. Taovvra (Athens 1941) pp. 415-422.
Oikonomos, G . P., 'H enl -cfjc; ".Axgono/..ewc; /..a-cgda -cfjc; ".A8rivac; Ntxric;, Ephemeris 1939-
1941, pp. 97-110.
Pantelidou: Pantelidou, M. A., Al fleoiaroetxai 'AOfjvat ('A8fjvm 1975).
Penrose, F. C., On Some Traces Connected with the Original Entrance of the Acropolis
of Athens, JHS XV, 1895, pp. 248-250, pls XI-XIV.
Pfuhl, E., "A. Koster, Das Pelargikon (book review)", Berliner Philologische Wochenschrift
1911, pp. 299-307.
Philippson, Die gri.echischen Landschaften: Philippson, A., Die gri.echischen Landschaften
I, 3 (Frankfurt a.M. 1952).
Picard, L 'Acropole: Picard, Ch., L 'Acropole I (L 'enceinte, l'entree, le bastion d'Athena Niki,
les Propylees) , II (Le plateau superieur, l'Erechtheion, les annexes) (Paris 1930).
Picard, Ch., L'anastylose du temple de la Victoire Aptere, Revue archeologique XV, 1940,
pp. 256-258.
Robert, Aus Kydathen: Robert, C., Der Au/gang zur Akropolis, Aus Kydathen (Berlin 1880)
pp. 173-194.
Schede, Die Burg van Athen: Schede, M., Die Burg van Athen (Berlin 1922).
Skias, A. N. , ".Avaaxacpal ev ".Axgonof..eL, Praktika 1910, p. 144.
Stevens, Hesperia , Suppl. III: Stevens, G. P., The Setting of the Periclean Parthenon, Hes-
peria , Suppl. III (1940) pp. 1-88.
292
Stubbings, F. H., The Mycenaean Pottery of Attica, BSA 42, 1947, pp. 1-75.
Tamaro, B., Culto miceneo sull'Acropoli, ASAtene IV-V, 1921-1922, pp. 1-11.
Thompson, H. A., Activities in the Athenian Agora, 1959, The Eleusinion, Hesperia 29,
1960, pp. 334-338.
Travlos, J., 'H rtaf...moxQLO"navtx."1 j3aotf...tx."1 -coi.i 'Aox.f...1']rtt£l.ou i:&v 'A81']v&v, Ephemeris
1939/1941, pp. 34-68.
Travlos, llo).wo.: Travlos, J., 'H 7WAEOOOµtxi/ t;at;u; rwv 'A017vwv (Athens 1960).
Tsountas, 'Axeo.noA.t~: Tsountas, C., 'H 'Axeo.no).i~ rwv 'A017vwv (Athens 1900).
Unger, G. F., Enneakrunos und Pelasgikon, Sitzungsberichte der philosophisch-philologi-
schen und historischen Classe der k B. Akademie der Wissenschaften zu
Munchen 1874, pp. 263-351.
Vanderpool, E., The Route of Pausanias in the Athenian Agora, Hesperia 18, 1949, pp.
128-137.
Vermeule-Townsend, E., The Fall of the Mycenaean Empire, Archaeology 13, 1960, pp.
66-75.
Wachsmuth, Die Stadt Athen: Wachsmuth, C., Die Stadt Athen im Altertum (Leipzig 1874).
Wachsmuth, C., Neue Beitrage zur Topographie von Athen - Der Konigspalast auf der
Burg und die Pelasgische Mauer, Berichte uber die Verhandlungen der k.
Siichsichen Gesellschaft der Wissenschaften zu Leipzig 39, 1887, pp. 399-405.
Walter, Athen, Akropolis: Walter, 0., Athen, Akropolis (Wien 1929).
Weller, Ch. H., The pre-Periclean Propylon, AJA 8, 1904, pp. 35-70.
Welter, G., Vom Nikepyrgos, AM 1923, pp. 190-201.
White, J. W., To IIeA.aQyLxov btt IlEQLxA.Eouc;, Ephemeris 1894, pp. 25-62.
Wide, Ausonia 1912: Wide, S., II pomerium e ii Pelargicon, Ausonia VII, 1912, pp. 177-
197.
Wilamowitz, Burg u. Stadt: Wilamowitz-Mollendorff, U. v., Burg und Stadt von Kekrops
bis Perikles, Philologische Untersuchungen 1880, pp. 97-172.
Wrede, Attische Mauem: Wrede, W., Attische Mauem (Athens 1933).
Wright, J., The Mycenaean Entrance System at the West End of the Acropolis of Athens,
Hesperia 63, 1994, 323-360, pls. 77-80.
BCH 1936, Chronique des Fouilles , p. 455.

BCH 1937, Chronique des Fouilles , p. 443.
BCH 1938, Chronique des Fouilles, p. 448.
BCH 1939, Chronique des Fouilles, p. 289.
AM 1887, Funde, pp. 141-142.
AM 1888, Funde, p. 107.
AM 1888, Miscellen, p. 228.
Praktika 1909, p. 60.
293
INDEX
Agrippa base 122

Agrolas 252
Anakeion 198
Androutsos bastion 203
Apollodoros, Bibliotheka 271
Archaeological Society, excavates S Slope 25, excavates Acropolis 26
Archaic W approach 122, 208, remains SW of entrance 56
Areopagus 197, 199
Aristion 202
Aristophanes 21, 258, 262, 267
Aristotle 21
Arnobius 272
Ascents to Acropolis, (second phase) 113, NE ascent 27, 29, 79, 84, 109, 206, NW
ascent 45, W ascent 25, 56, for animals 122
Asklepieion 25, 48, 197, spring 141, 202.
Balanos, N., 28, 29, 56, 115, 116, 118, 119, 120, 207
Bastion, mycenaean, beneath Athena Nike 25, 28, 56, 115, 116-118, 182, 183,
cross-wall 23 7, Archaic rebuilding 115
Bekker, I., 259, 263
Belger, Chr., 32
Belvedere 159, 160
Berard, J., 20, 38
Beule, E., 25, 31, 122
Blegen, C. W., 253
Bohn, Rich., 25, 122
Botticher, A., 31
Brauroneion 177
Broneer, 0., 27, 28, 29, 37, 39, 105, 144, 222, 228
Building, LH I, 147
295
INDEX
Buildings, mycenaean, at SE corner 165, 167-171, of historical times 165, 167

Building, mycenaean, in area of Mediaeval buttress 133-135
Building phases 249, at Mycenae 249, at Tiryns 249
Bursian, K., 31
Buttress, Mediaeval 130-131
Caskey, L. D., 141

Cave of Agraulos 141, of Apollo 27, 45, 107, 130, 132, 133, 215, of Pan 26, 30, 45,
49, 107, 109, 130, 133, 197, 215, of Zeus Olympios 27, 45, 107, 215, East
cave 46, 223
Chapel of Ag. Ioannis Alaniares 109
Child graves, MH 53, 54, 145, 182, beneath LH I house N of Erechtheion 75. See
also graves, tombs.
Cleidemos 19, 30, 49, 113, 204, 206, 208
Clemens 272
Cleomenes 21
Column base, mycenaean 244, 252
Column bases, Archaic 65-68, at Tiryns 252
Column drums in N Classical wall 147, 149
Constructions, mycenaean, behind NW Cyclopean wall 128-130
Cratinos 21, 264
Curtius, E., 31, 32
Davidson, Th., 31
Descent to caves 26, 57, 210-214
Descent to North Fountain 26, 57, 140
Dinsmoor, W. B., 37
Dionysios Halicarnasseus 258
Dionysos precinct 51
Dittenberger, W., SIG 266
D'Oodge, M. L., 32
Dorpfeld, w., 32, 36, 178
Drain, Archaic 129, 130
Early Helladic habitation 27, 51

Eleusinion 30, 197, 198, 199, 210
296
INDEX
Enneapylon 30, 204, 205, 208

Erechtheion 23, 26
Etymologicum Magnum 260
Eustathios of Thessalonika 270
Figurines, mycenaean 25, submycenaean 207

Fortification wall, traces 123-125, 153, 157, 160-161, 161-163, remains 25, 57, 107,
123, 128, 131-132, 135, 136, 138-139, 144, 145, 146, 147, 148, 150, 158, 177-
180, 182, stretches 160, 165, 174-175, 177-180, foundation 235, construction
235, 252 (Tiryns 252). Height 239. Thickness 237-238.
Fowler, L. D., 27
Gottling, c. w., 31
Graef-Langlotz (Wolters) 27
Grave mound, MH 26
Grave of Hippolytos 55, of Kekrops 23, 24, 55, of Talos 55, 198
Graves, LH 171, 174
Hansen, H., 25
Harrison, J ., 20, 32
Hearth, LH IIIB-C in precinct of Dionysos 223
Heberdey, R., 34
Hekataios 20
Hekatompedon 20
Herodotos 20, 21, 22, 32, 257-258, 260-262, 271
Hesychios 20, 259, 270
Hesychos (hero) 204
Hill, I. T., 38, 183
Hippias 21
Hoard of mycenaean bronzes 175
Holland, L. B., 27, 34, 35, 36, 37, 40, 58, 59, 105
Homer 271
House, LH I 70-75, 109, 113
House of Arrhephoroi 140
House walls, LH 172-175
Houses, floors 244, on NE pathway 27, 107, 216-217, 223
297
INDEX
Hyperbios 252
Italian School of Archaeology 27

Inner gate (W entrance) 184, 186
Judeich, W., 36, 57
Kawadias, P., 26, 33, 57, 105, 107, 131, 140, 160, 214
Kawerau, G ., 26, 34, 59, 77, 101, 116, 122, 128, 130, 159, 237
Kekropion 95, 97
Kekrops, tomb, see Grave
Keramopoullos, A. , 23, 27, 56, 107, 206, 208, 209
Kimilia 45
Kirnon 22
Klepsydra 28, 141, 198, 202, 210, 217, 218, 219
Kolbe, W., 28, 29, 39, 174, 227, 228
Koster, A. , 37, 40, 57, 116, 131, 202, 209, on fortification wall 26, 32, 33, on
entrances 33, on Pelargikon 34
Kylon 21 , 129 n. 238, 202
Leake, W. M., on Pelargikon 30, 198

Lion Gate, Mycenae 182, 184, 186
Lolling, H. G. , 32
Lucian 23, 267-269
Main Gate 183

Mardonios 22, 153
Middle Helladic habitation, S. Slope 27, 53
Miller, W., 21 , 32
Moschophoros base 165
Mudbrick house walls 243
Muller, K., FGH III 270
Muller, K. 0., 30
Museum, Acropolis 25, 165
Neolithic remains on S Slope 27, on N Slope 29, NL house 50
298
INDEX
Niche in W bastion 119-121

North Fountain 28, 29, 39, 46, 202, 203, wooden steps 239-241, stone steps on
mud plaster 241-243
Obsidian 51
Odeion Herodes Atticos 199
0 live tree 24
Orlandos, A. K., 28, 116
Palace 109
Palace of Erechtheus 24
Panathenaic Way 199
Pandroseion 95, mycenaean fill 96
Parian Chronicle 21, 263
Pathway, W entrance, for pedestrians 187, for animals 189
Paton, J. M., 26
Pausanias 258-259, 272
Peisistratids 129 n. 238, 202
Peisistratos 19, 21, 22, 129
Pelasgians 19
Pelargikon 19, 20, 21, 22, 23, 24, 25, 30, 41, extent 220, date 221
Pelasgikon 19, 20, 21
Pelastikon 20
Pelekides, E., 26, 51
Peripatos 106, 107
Persians (destruction of Cyclopean wall at NE descent) 152, 153
Pfuhl, E., 35
Photios 260
Picard, Ch., 20, 36
Pittakis, K., 79
Pits, containing LH IIIA-C pottery, 216, 222
Pollux 269
Poseidon 24
Pottery. Provenance: from W bastion 228, from NW beside Mediaeval buttress
228, from LH I house 109, from NE pathway houses 223, 227, 229, from S
wall of Terrace IV 102, from brick wall SE of Parthenon 230, with LH
299
INDEX
bronze hoard 175, from S fortification wall 227, from graves S of Parthenon
230, from graves next to SE comer of fortification 230, from N Fountain
144, 206, 228-229, from SW slope 223-224, from Odeion of Pericles 223.
Shapes: goblet 277, 280, 282, pithoid vessel 277, kylikes from N Fountain
144, pithoid amphoriskos 171, platter 280, shallow bowl 276, 279, skyphos
280, stirrup jar with bronze hoard 176. Period: Prehistoric pottery (pub-
lished by Wolters in Graef-Langlotz), Prehistoric from N slope 28, NL 279,
NL from N slope 50, EH 276, 277, 279, 280, incised ware 51, MH 53, 54,
279, 282, matt painted 53, 281, minyan 53, dating phase II 113-114, dating
Terrace III 92, LH I 277, 278, 280, 282, LH II 277, 280, 282, LH III 277,
280, LH IIIA 277, 280, 282. LH IIIBl 280, 283, LH IIIB2 280, 282, 283,
LH IIIC 229, 230, byzantine sherds 132
Pnyx 199
Pythion 197, 198, 216
Reservoir, Archaic 129, 130

Retaining walls of ramp to NW descent 108, to W ascent 121, 187
Ross, L., 25, 31
Sanctuary, of Aglauros 46, of Eros and Aphrodite 27, 141, 222

Schaubert, E ., 25
Scholiast in Aristophanes 259, 264, 269, in Lucian 270, in Sophocles 271
Sea (in Erechtheion) 24
Skias, A., 26, 53
Spring at Asklepieion 49. See also Kl.epsydra.
Stairway to the plateau of the caves, Classical 136, mycenaean 136-137, to palace
of Mycenae 195-196, to Terrace III 94
Stairwell, inside Cyclopean wall at landing of NE ascent 150-152
Steps of descent to the plateau of the caves 108-109, between gates of W entrance
189, of NE ascent 107, sandstone steps from palace 244, poros steps near
Hekatompedon 191-195, from House Mat Mycenae 196, from megaron at
Tiryns 195
Stevens, G. P., 26, 28, 29, 37, 86, 87, 121, 123, 183
Strabo 21
Temple, Archaic 109, of Athena Nike 115
300
INDEX
Themistocles 22
Theodoretos 272
Thucydides 21, 202, 209, 264-266
"Tokens" in Erechtheion 23, 95, 100
Tombs 109
Travlos, J., 38, 183
Trident, of Poseidon 24
Terraces, complex 109-113. Traces 85, 86, 86-91. Remains 79-80, 80-81, 82-84, 92,
101, 101-104, W wall of terrace III 28, 86-94, walls E of Erechtheion 74-
86, S of Erechtheion and around Archaic temple 98-100, S of NE ascent
101-105, foundation trench 232, foundation underpinning 232-233.
Tsountas, Chr., 32
Turkish walls E of Erechtheion 76, 78
Tyrrhenians 19
Unger, G. F., 31
Urfirnis 51
Wachsmuth, C., 31
Walls between Propylaia and Erechtheion 59, 61-65, E of Erechtheion 69-70, N
and E of Erechtheion 57, of Pelargikon, cuttings on rock 217, 218, 219
War of Independence 141
Welcker, F., 31
Weller, Ch. H., 122
Wells NL 28, EH 28, MH 28, 53, LH 28, 203, near Klepsydra 222, in area of
Odeion of Herodes Attikos 225, well full of sherds 216
Welter, G., 28, 37, 56, 116, 183
West Classical tower 115
White, J. W., 178, 179
Wilamowitz-Mollendorff, U. v., 31, 199
301
THE BOOK
OF ATHENS
BY SPYROS E. IAKOVIDIS
No 240 OF THE ARCHAEOLOGICAL SOCIETY
AT ATHENS LIBRARY
WAS PRINTED IN JULY 2006
BY «GRAPHIKES TECHNES EKDOSEIS PERPINIA
ANTONIS EVAG. BOULOUKOS & Co»

Spyros E. Iakovidis - The Mycenaean Acropolis of Athens - 2006

Uploaded by

Copyright:

Available Formats

Spyros E. Iakovidis - The Mycenaean Acropolis of Athens - 2006

Uploaded by

Document Information

Original Title

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

Spyros E. Iakovidis - The Mycenaean Acropolis of Athens - 2006

Uploaded by

Copyright:

Available Formats

THE MYCENAEAN ACROPOLIS

© The Archaeological Society at Athens

Editing: Ilektra Andreadi

THE MYCENAEAN ACROPOLIS

Preface to the English edition 9-10

The Greek edition of the Mycenaean Acropolis of Athens was published in

The fortified citadel, a characteristically Mycenaean creation, has always

The ancient sources contain a fair number of references to the Mycenaean

* See infra, Appendix I. 4. Herod. VI 137 (Hekataios ).

indiscriminately,8 and Hesychios adds the form nEAOCJTlKOV. 9

14. AJA 1893, p. 482. 19. I. 60 (ed. Jacoby).

23. VIII 51-53. 26. Herod. VIII 51-53.

37. See also Skias, Ephemeris 1902, p. 123, n. 1.

Fowler, The Erechtheum, Cambridge, Mass. 1927, pp. 13-14, 122,

40. See also Bd.A 1922-1923, p. 278 f. , 1924-1925, p. 88 f.

43. Ersch u. Gruber 1, Sect. VI, 229 f. , A then.

The existence of a fortification wall similar to other Mycenaean fortifica-

55. Topographie p. 337. 61. RhM 1896, p. 131.

of an entrance, and he categorically rejected the possibility that there was a

65. Pelargikon, especially pp. 5-16, 27 66. flQoiaro(}txi/ 'AQxawA.oyia p. 300.

Koster's observations are in general sound but inadequate. His conclusions

72. See also Erechtheum pp. 137-142.

73. L 'Acropole I pp. 11, 19-20. 75. Alt Ath. p. 3.

The discoveries of the period 1932-1939 added much important material.

lem of the historical sequence of events with arguments based on a thorough

Fig. 3. The Acropolis of late Mycenaean times according to Travlos

THE MYCENAEAN ACROPOLIS

The Acropolis rock consists of a large ellipsoidal mass of cretaceous -

83. See especially Philippson, Die griechischen Landschaften I, 3, pp. 894-895.

0 10 50 100 150 200 250 300 350

Plan 1. The Acropolis rock.

0 10 50 100 150 200 250

Located on the Acropolis by the Athenian myths were the grave of

117. He is referred to as T aAws in practice before historical times (see also

I. THE LATE HELLADIC REMAINS ON THE ACROPOLIS

Odeion of Herodes Atticus. They had been attributed to the Mycenaean

124. D'Oodge, Acropolis p. 25, Botti- 129. Judeich, Top. p. 116, n. 2.

Other remains of buildings that have been ascribed to Mycenaean times

Plan 3. Walls between the Propylaia and the Erechtheion area.

Acropolis excavation and were inaccessible. Wherever conditions permitted,

Fig. 5. Wall 22 of Plan 3, from E.

149. Holland, AJA 1924, p. 162.

Preserved to the E and NE of the Erechtheion are complexes of walls.

158. See Holland, AJA 1924, pp. 142-157.

Plan 4. Walls in the area N and NE of the Erechtheion.

Plan 5. Remains of the LH I house N of the Erechtheion.

It represents the remains of a four-sided area, the N part of which is not

0 10 50 100 150 200 250 300 350

Plan 6. The Acropolis at the begi,nning of the LH period.

measurement of 0,30-0,50, and standing to a height of 0,40-0,80 from the rock,

THE AREA W OF THE TERMINUS OF THE NE ASCENT

Preserved in the area N and NE of the Erechtheion (see also Plan 4) is

Plan 7. Mycenaean and Mediaeval walls NE of the Erechtheion.

Fig. 9. Walls T4 and TS of Plan 7, from the S.

Plan 8. Remains of terrace walls NE of the Erechtheion.

166. Kavvadias - Kawerau pl. A.

Preserved within the cella of the Erechtheion, in the corner formed

Plan 9. Foundation trench of the W terrace wall.

Fig. 13. Working of the rock at 2 on Plan 9, from the W