T h e Date of the Older Parthenon

A n a l e c t a Gorgiana

240

Series Editor
George Kiraz

Analecta Gorgiana is a collection of long essays and short

monographs which are consistently cited by modern scholars but
previously difficult to find because of their original appearance in
obscure publications. Carefully selected by a team of scholars based
on their relevance to modern scholarship, these essays can now be
fully utili2ed by scholars and proudly owned by libraries.
The Date of the Older Parthenon

William Dinsmoor

gorgias press
2009
Gorgias Press LLC, 180 Centennial Ave., Piscataway, NJ, 08854, USA
www.gorgiaspress.com
Copyright © 2009 by Gorgias Press LLC
Originally published in
All rights reserved under International and Pan-American Copyright
Conventions. No part of this publication may be reproduced, stored in a
retrieval system or transmitted in any form or by any means, electronic,
mechanical, photocopying, recording, scanning or otherwise without the
prior written permission of Gorgias Press LLC.
2009

1
ISBN 978-1-60724-469-1 ISSN 1935-6854
Extract from The A^merican Journal of Archaeology, vol. 38 (1934).

Printed in the LTnited States of America

 THE DATE OF THE OLDER PARTHENON
UNDER this title, in 1 9 0 2 , was published an authoritative article by Professor Dorp-
feld, apparently closing for all time the debated question of the age of the prede-
cessor of the marble Parthenon which we see today. 1 Exactly a decade earlier
Dorpfeld had published his architectural restoration of the Older Parthenon, 2 a
study which was only slightly modified by his later article; and exactly a decade
afterwards this plan in turn was drastically revised in a brilliant article by Dr.
Hill.3 The date proposed by Dorpfeld, however, is still in favor; and it now seems
desirable to review his arguments in the light of evidence made available since his
publications.

1. ARCHITECTURAL E V I D E N C E

The various changes of opinion with regard to the date may be briefly summarized.
(I) Before 1902 there was no reason to suppose that there had been more than
one Older Parthenon, and the question was merely that of a single date.
(A) I t was originally suggested by Leake that the Older Parthenon was pre-
Persian, on the basis of the ambiguous reference by Hesychius (s.v. «aro^Trfoos!
and the existence of architectural remains in the north wall of the Acropolis (marble
drums and poros entablature blocks) which he assigned to this site, even before the
foundations were discovered.4 The pre-Persian date was verified by Ross, who exca-
vated south of the Parthenon in 1835-1836 and discovered that the terrace con-
tained burnt Persian débris, and that both the foundation in situ and the marble
drums in the north Acropolis wall showed traces of a conflagration.5 From this
moment the pre-Persian date of the Older Parthenon seemed to be an assured fact. 6
(B) A second stage was marked by the Acropolis excavation of 1885-1890, when
Dorpfeld identified the old poros temple of Athena and concluded that this must
have been the temple to which Hesychius referred, proving, moreover, that it had
been the true site of the poros entablature blocks in the north Acropolis wall. He
immediately perceived that the marble column drums must belong to a different
and later temple, namely, the foundation of the Older Parthenon, which he assigned
to the period of Kimon. 7 This date he regarded as confirmed by the excavation
of 1888 south of the Parthenon, revealing a terrace fill composed of fragments of
1 Dorpfeld, Ath. Mitt. 1902, pp. 379-416. 2 Ibid., 1892, pp. 158-189.

3 Hill, A.J.A. 1912, pp. 535-558. 1 Leake, Topography of Athens, 1821, pp. xii, 282-284.

6 Ross, Arch. Aufs. I, pp. 82-117, 127-141; I I , pp. 285-286.

6 Leake, Topography1, I, pp. 554-556; Penrose, Athenian Architecture1, pp. 17-18, 73-75, pl. I X ,
X X X I V , X L ; Beulé, Acropole d'Athènes, I I , pp. 5 - 1 3 ; Strack, Arch. Zeit. 1862, pp. 241-245, pl. C L X -
C L X I ; Michaelis, Parthenon, pp. 112-113, 119-123; Burnouf, Legende athénienne, pp. 53-61; Ville et
Acropole d'Athènes, pp. 164, 176, 179-181, 183-184; Penrose, Athenian Architecture2, pp. 18-20, 98-102,
116-118, pl. 9, 46; J.ll.S. 1891, pp. 275-297; 1892, pp. 32-47. All these writers, of course, were in
error in one respect; the Older Parthenon was not the old finished temple of Athena burnt by the Per-
sians, to which Hesychius seems to have referred. And Burnouf even claimed that the marble drums
in the north wall were extra blocks from the Periclean structure (Ville et Acropole, p. 184).
7 Already in his first discussions of the poros temple {Ath. Mitt. 1885, pp. 275, 277; 1886, p. 165;

1887, p. 45; Ant. Denk. I , pl. 1), DBrpfeld had outlined his Kimonian theory of the Older Parthenon.
408

T H E ARCHAEOLOGICAL INSTITUTE
OF AMERICA
 THE DATE OP THE OLDER PARTHENON 409

early buildings a n d sculptures which he regarded as those destroyed b y t h e Persians,

accompanied b y potsherds of which m a n y were red-figured a n d therefore agreed
with a d a t e early in t h e fifth c e n t u r y for this débris. 1 R e g a r d i n g all this material
as " P e r s e r s c h u t t , " it seemed obvious t h a t it m u s t h a v e been deposited a t some
t i m e a f t e r 480 B.C., t h o u g h before t h e age of Pericles, a n d so p r e s u m a b l y during t h e
political supremacy of K i m o n , while building t h e great south wall. 2 T h e physical
relationship showed t h a t t h e P a r t h e n o n f o u n d a t i o n m u s t in a n y case have a n t e -
d a t e d t h e south wall b y an u n k n o w n period of t i m e ; D o r p f e l d suggested t h a t it was
begun shortly a f t e r t h e b a n i s h m e n t of Themistokles in 471 B.C., a n d such a d a t e
won general acceptance. 3 M o r e exactly, F o u c a r t preferred 469 B.c. because of his
reading of t h e S t r a s b o u r g P a p y r u s . 4 K o p p descended even f a r t h e r , to 454 B.C.5
F u r t w a n g l e r offered merely a variation in t h a t he preferred t o regard Themistokles
as t h e builder, in 479 B.C.,6 a view which D o r p f e l d recognized as not impossible. 7
Of those who wrote a t this time, only Penrose adhered to t h e older view t h a t t h e
Older P a r t h e n o n was pre-Persian.
(II) I n 1902, however, Michaelis published his article on t h e 'Apxaîos News, arguing
t h a t a reference t o such a s t r u c t u r e a p p a r e n t l y d a t i n g f r o m a b o u t 506 B.c. implied
t h a t t h e r e were already at this early period two temples of A t h e n a on t h e Acropolis,
one of t h e m being a hypothetical Older E r e c h t h e u m . B u t Dorpfeld, who rightly
looked with suspicion u p o n t h e " O l d e r E r e c h t h e u m , " now reviewed t h e evidence
bearing on t h e Older P a r t h e n o n and argued t h a t it, too, could be t h r u s t back as
f a r as 506 B.C.8 His most i m p o r t a n t evidence was t h e resurrection of t h e d a t a
furnished b y Ross concerning t h e b u r n t Persian débris a n d t h e traces of fire on t h e
f o u n d a t i o n a n d column d r u m s ; a f t e r verifying these traces of a conflagration, D o r p -
feld r e t u r n e d t o t h e earlier t h e o r y t h a t t h e Older P a r t h e n o n h a d been destroyed b y
fire in 480 B.C. A reexamination of t h e old p h o t o g r a p h s a n d sketches of 1888—the
resulting sectional reconstruction of the s t r a t a (compare Fig. 1—after t h e lapse of
f o u r t e e n years, being in itself a r e m a r k a b l e feat) showed t h a t t h e fill c o n t e m p o r a r y
with t h e f o u n d a t i o n lacked b u r n t Persian débris a n d so a n t e d a t e d 480 B.C. F r o m
these facts it would h a v e been n a t u r a l , a n d probably sufficient, to draw t h e infer-
ence t h a t t h e Older P a r t h e n o n h a d been commenced a short t i m e before 480 B.C.,
n a m e l y , immediately a f t e r t h e b a t t l e of M a r a t h o n in 490 B.C. B u t Dorpfeld went
f a r t h e r . F o r he deduced two facts, either of which seemed a d e q u a t e evidence, b u t
when combined a p p e a r e d t o f o r m irresistible proof, t h a t t h e Older P a r t h e n o n h a d
passed t h r o u g h two building periods, with a change of plan which suggested a longer
d u r a t i o n a n d a more n o t a b l e interruption t h a n could be f o u n d within t h e decade
1
Dorpfeld, Ath. Mitt. 1888, pp. 106, 224, 431-435; 1892, pp. Ifi2, 187. 2 Ibid., 1892, pp. 188-189.
3
Wolters, quoted in Graef, Vasen der Akropolis, I, p. xxiv (written about 1888) ; Harrison and
Verrall, Mythology and Monuments of Ancient Athens, pp. 467-470; Botticher, Akropolis, pp. 88, 91,
97-99; Miller, A.J.A. 1893, pp. 520-521; Graef, Jb. Arch. I. 1893, Anz., 15; Curtius, Stadtgesch. v.
Athen, pp. 127-133; Beloch, Gr. Gesch.11, p. 583; Middleton, Plans and Drawings of Athenian Buildings,
pp. 4-7, nos. 55, 57, 83, 103, 112, 120, p. 12 E, pl. I, II, X I I I ; F r a z e r , Pausanias,H,p. 306; Ed. Meyer,
Forschungen zur alt. Gesch. II, p. 97, n. 1; Keil, Anonymus Argentinensis, pp. 81-107; Michaelis, Jb.
5
Arch. I. 1902, p. 12. > Foucart, Hev. phil. 1903, pp. 7-8. Kopp, Jb. Arch. I. 1890, p. 270.
6
Furtwangler, MeiMerwerke, pp. 162-168; Masterpieces, pp. 419-423; Busolt, Gr. Gesch.1 I l l 1, pp.
359-360; Diimmler, in Pauly-Wissowa, II, 1953-1954; Gardner, Ancient Athens, pp. 210-213.
7
Dorpfeld, Ath. Mitt. 1897, pp. 167-168. »Cf. A J . A. 1932, pp. 307 ff„ and p. 311, n. 3.
410 WILLIAM B E L L DINSMOOR

I ®
1 ?
a 55
W S

*
fc.«»
Of
0
S ^
H 53
§ «
02 .
» 5
g
<j -O
« s
1 &

H
T
P* w
THE DATE OF THE OLDER PARTHENON 411
412 WILLIAM BELL DINSMOOR

490-480 B.C. The first of these facts was the existence of two sets of temporary
retaining walls for the terraces south of the Parthenon, differing in technique and
date (one polygonal limestone and the other poros ashlar). 1 The second fact was
the incongruity between a poros and Kara limestone crepidoma supporting a marble
colonnade, an unsatisfactory contrast which argued a change and enrichment of
design at the level of the stylobate. 2 From these observations seemingly resulted
the following equations :
(a) polygonal retaining wall = poros limestone crepidoma;
(b) ashlar retaining wall = calcined marble drums.
Since the chronological succession of the two equations is assured, and since the
calcined marble drums (b) are manifestly pre-Persian, it appeared to be necessary
to conclude that both periods were pre-Persian.
(C) Consequently Dorpfeld assigned stage (a), assuming that the limestone
crepidoma was intended to support a poros limestone colonnade, to 510, 508, or
506 B.C.,3 and stage (b) to 490 B.C.4 These seemed to be conclusive arguments, and
the new chronology found general acceptance. 5
(D) Dorpfeld even considered, though with less favor, the possibility of earlier
dates, the poros Parthenon in the period of Hippias (514-510), and the change to
marble in that of Kleisthenes; but he rightly regarded this as less probable on ac-
count of the red-figured sherds which would then necessarily antedate Hippias, and
the resulting excessive length of the building period and the improbable use of
Pentelic marble under Kleisthenes. 6 This theory has been revived, however, by
Walter, who regards the decade 490-480 B.c. as an improbable one for monumental
building on account of the impending Persian danger, and assigns stage (b), the
marble Older Parthenon, to Kleisthenes, and stage (a), the poros Older Parthenon,
to the Peisistratidae, who thus showed Athena the same consideration that they
displayed toward Zeus in the Olympieum of the lower city. 7
(E) On the other hand, it is necessary to record another variant opinion, that of
Furtwangler, who likewise accepted the pre-Persian date for both stages of the
work, but set both in the decade after Marathon, on the basis of a general impression
1
The ashlar retaining wall had been ignored in Dorpfeld's article of 1892, though he had discussed it
earlier (Ath. Mitt. 1887, p. 386; 1888, pp. 433-434). = Dorpfeld, Ath. Mitt. 1902, p. 399.
8
The most probable date would have been 506 B.C.; for Kleisthenes was in power for a very short
period after the banishment of Hippias in 510, and was in his turn displaced by Isagoras with the help
of Kleomenes. Not until 507 B.C. did the capitulation of Kleomenes on the Acropolis permit the return
of Kleisthenes ; and not until the defeat of the confederated attack of the Spartans, Boeotians and Chal-
cidians on Attica in 506 B.C. would he have been free to undertake great structural embellishments.
« Dorpfeld, Ath. Mitt. 1902, pp. 379-416; 1911, pp. 49-51; Jb. Arch. I. 1919, p. 5.
!
Wiegand, Porosarchitektur, pp. 110-114; Judeich, Topographie1, pp. 64, 66, 225-228; D'Ooge,
Acropolis, pp. 66, 70, 78-92; Kavvadias and Kawerau, Ausgrabung der Akropolis, pp. 108-114; Hill,
A.J.A. 1912, pp. 536-537, 556-557; Collignon, Parthenon (1911), pp. 5-9; ed. min. (1914), pp. 22-37;
Dickins, Acropolis Museum, I, pp. 6-9; Weller, Athens, pp. 271-274; Lechat, Rev. Ét. anc. 1913, pp.
125-130; Judeich, Hermes, 1929, p. 406 (the allusion to resumption of work by Themistokles or Kimon
must be a clerical error); Topographie pp. 68, 70, 248-251 (with a possibility that the marble revision
was undertaken before 490); Bodenwaldt and Hege, Acropolis, p. 26. Also, in more general terms,
Beloch, Gr. GeschII 1, p. 205; Adcock, Cambridge Ancient History, IV, pp. 66-67; Seltman, Athens,
Its History and Coinage, p. 94. « Dorpfeld, Ath. Mitt. 1902, p. 411.
7
Walter, Athen. Akropolis, pp. 68-72.
 T H E D A T E OF T H E O L D E R P A R T H E N O N 413

that the sherds in the fill were too late for the decade of Kleisthenes. 1 This theory,
though close to the truth, has found no support.
( I l l ) Later investigation has shown that Dorpfeld's equations contain two nota-
ble fallacies which destroy their validity, and disrupt the coincidences on which he
based his interpretation. In the first place, the eventual publication of Kawerau's
plan of the excavations, with his accompanying remarks, shows that, with regard to
the west elbow of the so-called pre-Persian ashlar retaining wall (3), "its founda-
tions at the southwest corner and a part of the south face rest on the foundations of
the south citadel wall. The latter, therefore, at least in its lower courses, is older
than this poros wall. . . . In the foundation of the poros wall appear, in the lowest
courses immediately above the socle of the Kimonian south wall, some limestone
blocks and, furthermore, two fragments of marble column drums of the Older Par-
thenon with the same beginnings of the flutes." Kawerau, therefore, contended
that the eastern and western sections of the poros ashlar retaining wall (3), though
probably contemporaneous as Dorpfeld had assumed, were Kimonian rather than
pre-Persian, erected in order to support temporarily the earth terraces (perhaps
endangered by the collapsing of the polygonal retaining wall 2) during the erection
of the Kimonian south wall (5).2 On the other hand, we have the more subjective
but nevertheless conclusive argument that Hill's restoration of the Older Parthenon
has eliminated the second of the two deductions made by Dorpfeld. There never
was a discrepancy between a poros (and Kara limestone) crepidoma and the marble
colonnade, for the reason that the crepidoma was actually of marble like the colon-
nade, with a transitional Kara limestone bottom step analogous to that of poros in
the "Theseum " and that of blue marble at Rhamnus. The existence of the marble
step and stylobate, pointed out by Ross but overlooked by Dorpfeld, was independ-
ently demonstrated by Hill. We are now confronted, not with two projects, but
with a unified design and a single building period, and must restate the equation
(one rather than two) in the following terms:

(c) polygonal retaining wall = calcined marble drums.

I t now remains to ascertain the date of this construction, which was in any case
earlier than 480 B.C. (any later date being excluded by the traces of fire).
(F) This unity of design was perceived by Heberdey, 3 who preferred the decade of
Kleisthenes on the ground that 490-480 B.C. was a period for strengthening rather
than weakening the fortifications; 4 before Marathon, he would have preferred about
500 B.C., to diminish the length of the building period as much as possible, but ad-
mitted the weight of Dorpfeld's arguments for Kleisthenes and 510 B.C.5 T h e
theory was complicated, furthermore, by diverse interpretations of the two ends of
the ashlar retaining wall (3). Observing Kawerau's evidence t h a t its west elbow
cannot have been earlier than Kimon's south wall (4),6 Heberdey argued that it
1
Furtwangler, Aegina, pp. 353, 496-497.
2 Kavvadias and Kawerau, Ausgrabung, pp. 118, 120; Heberdey, Purosshilptur, p. 233.
3
Heberdey, Porosskulptur, pp. 231-240.
4
I.e., the Old Propylon, which he rightly assigns to the same building program as the Older Parthe-
non, would have weakened the west defenses of the Acropolis (ibid., p. 230).
5
Ibid., pp. 229, 230, 240. s Ibid., p. 233.
414 WILLIAM BELL DINSMOOß

could not have been a patch facing the Pelasgian wall, which is particularly well
preserved at this point; 1 he could find no explanation for it other than a conjectural
enlargement of the masons' stoneyard during the Periclean period.2 On the other
hand, while admitting that the evidence for the east part of the ashlar retaining wall
is not contrary to a post-Persian date,3 Heberdey stressed Kawerau's observations
that it runs under the protruding upper part of the south wall (5) and, therefore,
must at least be pre-Periclean. He concluded that the east part of the ashlar wall
was pre-Persian, 4 and to explain its existence developed his theory of the great
ramps, at first supported on the polygonal retaining wall (2), and then extended

F I G . 3 . — V I E W OF SOUTHEAST C O R N E R OF O L D E R PARTHENON FOUNDATION

Phot. German Institute, Athens, Akr. 71

with the support of the east part of the ashlar retaining wall (3), both of these being
approximately of the same date. And as confirmation of the great ramps and of the
theory that the lofty basement was to have remained exposed to view on the south
flank, Heberdey cited the vertically drafted margin at the southeast corner (Fig. 3),
descending fourteen courses or 6.90 m. below the marble crepidoma of the present
Parthenon. His theory has been adopted by some of the most recent writers,5 but
is rejected by others.6
(G) Schräder, in reviewing Heberdey, raised the argument (later employed by

Ibid., p. 235.
1 2 Ibid., p. 238. 3 Ibid., p. 235. 4 Ibid., p. 234.
Fougères, Parthenon (1926), p. 9; Robertson, Greek and Roman Architecture, pp. 113-114; Körte,
s

Phü. Woch. 1931, 112; Langlotz, Vasen der Akropolis, I I , p. vii (preferring about 500 B.C. as the date
on account of the potsherds; see p. 439, n. 2).
s Judeich, Topographie,2 p. 250, n. 1; cf. Schede, Walter, and Rodenwaldt, as noted above.
 T H E DATE OF T H E OLDER PARTHENON 415

Walter, as noted above) t h a t it would be most unlikely t h a t t h e Peisistratids should

begin a great Olympieum in t h e lower city and remain contented with t h e modest
poros temple on t h e Acropolis. H e was also impressed by the " a r c h a i c " propor-
tions of t h e plan, with 6: 16 columns as in t h e Olympian Heraeum. And therefore,
without conclusive arguments, he was tempted to regard t h e Older P a r t h e n o n as a
Peisistratid creation of 530-520 B.C.1
(H) Heberdey's theory, however, arouses serious doubts. I n t h e first place,
Dorpfeld's original objection to placing t h e marble Older Parthenon as early as t h e
sixth century still holds its force: the building period would be excessively long in
proportion to t h e work accomplished, and "Pentelic marble suits better t h e time
after M a r a t h o n , Parian t h a t of Kleisthenes." 2 These objections would apply even
more strongly to a Peisistratid date. T h e arguments against monumental building
in t h e decade 490-480 B.C., or the assumption t h a t t h e Peisistratidae should have
displayed greater respect toward Athena, are purely subjective. There is no ade-
quate evidence for t h e great ramps: t h e stratification was horizontal, according to
all observers and photographs, with no detected inclination from east to west. 3
T h e theory t h a t t h e basement was to have been exposed to t h e great height of four-
teen courses, as suggested by t h e vertically drafted southeast corner, is controverted
by the appearance of t h e foundation itself. For t h e blocks are closely jointed only
in t h e three upper courses (under the recessed coping); below these t h e blocks often
touch only a t the upper anathyrosis band, and sometimes even this is lacking so t h a t
t h e joints are wide open from bottom to top; t h e gaps, according to Ross, are of 1 - 2
finger widths; 4 thus if t h e irregular faces of t h e blocks had ever been trimmed back
they would nevertheless have presented gaping joints. T h e true proposed ground
level is indicated, not by t h e b o t t o m of t h e vertically drafted corner (which was in-
tended merely as a guide to prevent contraction of t h e dimensions toward t h e top),
b u t by t h e horizontally drafted margin ten courses higher. 5 T h e differences be-
tween t h e west and east parts of t h e poros ashlar wall (3), namely, t h e greater thick-
ness (1.70 m. as against 1.20 m.) and t h e deeper foundations (going down to bed
rock instead of resting merely on t h e humus stratum) at the west, are in reality due
solely to t h e exigencies of t h e site. At t h e west, t h e poros wall (3) was built outside
rather than inside t h e Pelasgian wall (1); t h e humus s t r a t u m naturally did not exist
at this point, and t h e wall h a d to be founded on bed rock, and with this greater
depth was m a d e thicker. B u t in its upper portions the west p a r t of t h e poros wall
is reduced to 1.20 m. in thickness, like the east p a r t ; t h e construction of t h e two

1
Schrader, Deutsche Literatur-Zeitung, 1922, pp. 683-684; cf. 1930, p. 1562, where he complains t h a t
Weickert omits the Older Parthenon from the survey of archaic architecture.
2
Dorpfeld, Ath. Mitt. 1902, p. 411.
3
Kawerau, it is true, suggests t h a t since the natural rock, and therefore the base of the polygonal
terrace wall, rose in level from east to west, the earth might have followed this rock slope (Ausgrabung,
p. 114). B u t comparison of Dorpfeld's two sections at the middle and east end of the Parthenon
hardly permits the fine distinction between the levels of the top of stratum I I I suggested by Heberdey
(Porosskulptur, pp. 236-237).
4
Ross, Arch. Aufs. I, p. 89; Ziller, Zeitschr.f. Bauwesen, 1865, p. 40.
6
Dorpfeld, Ath. Mitt. 1892, p. 165; 1902, p. 395; Heberdey (Porosskulptur, pp. 237-238) dismisses
this sunken band too briefly. Judeich's objection to Heberdey's theory was based chiefly on this
improbable exposure of the basement.
 416 W I L L I A M B E L L DINSMOOR

portions is identical; both, therefore, must have been of the same date. And while
it is true that the east part of the ashlar retaining wall (3) runs under the overhang-
ing upper part of the south wall (5) and so must have been pre-Periclean, it is equally
certain that the west elbow of the ashlar wall (3), resting on Kimon's south wall (4),
cannot have been pre-Kimonian. In other words, the poros wall (8), instead of
forming two separate units, pre-Persian at the east and Periclean at the west, must
have been a homogeneous Kimonian structure. For such reasons, most of which I
had considered even before the publication of Heberdey's theory, it now becomes
necessary to adopt a different conclusion as to the period of equation (c), namely,
that the entire work was a unified design executed in the decade 490-480 B.C. " Now
the poros Parthenon disappears as a myth, its sole basis, an assumed incongruity
between a limestone stylobate and marble columns, and an assumed coeval terrace
wall, eliminated because there never was a limestone stylobate, and because the ter-
race wall contains fragments of marble columns of the very temple which Dörpfeld
thought it was intended to support. And the Older Marble Parthenon emerges as
the only predecessor of the present temple,1 the creation of Themistokles and Aris-
teides after the battle of Marathon, a memorial of victory over the Persians which
the Persians themselves demolished upon their return." 2
Thus we have six theories current at the present day:
(C) two Older Parthenons, poros ca. 510, marble 490 B.C. (Dörpfeld, 1902)
(D) " " ca. 530, " 510 B.C. (Walter, 1929)
(E) " " 490, " ca. 485 B.C. (Furtwängler, 1906)
(F) one Older Parthenon, marble ca. 510 B.C. (Heberdey, 1919)
(G) " " ca. 530 B.C. (Schräder, 1922)
(H) " " ca. 490 B.C. (Dinsmoor, 1922)
On architectural grounds we have found that there was but one Older Parthenon,
that of marble, so that theories C, D, and E should be abandoned. There remain
theories F , G, and H, differing only in that the marble design would be dated respec-
tively about thirty, fifty, or ten years before the Persian destruction. The archi-
tectural grounds, while not absolutely conclusive, point to the greater probability
of the latest of these dates (H).

2 . STRATIGRAPHICAL E V I D E N C E
In accordance with the chronological system proposed above, the terrace filling
south of the Parthenon should be divisible into the following eight well defined strata
from top to bottom:
V, the Periclean fill behind wall 5 (447-438);
IV, the later Kimonian fill behind wall 4 (469-461);
1 This is not the place for an examination of the problem of Buschor's " Ur-Parthenon " (cf. Buschor,

Ath. Mitt. 1922, pp. 96-98; Noack, Eleusis, pp. 298-300; Schräder, Jb. Arch. I. 1928, pp. 70, 85-89;
Deut. Literatur-Zeitung, 1930, p. 1562-1563; Weickert, Arch. Archit., pp. 19-21, 99,101; Walter, Akrop-
olis, pp. 46, 71), which is opposed by Dörpfeld {Phil. Woeh. 1929,1247); Judeich (Hermes, 1929, p. 395,
n. 1; Topographie', pp. 249 with n. 1, 458), and Körte (Phil. Woch. 1931, pp. 111-118).
2 Dinsmoor, Art and Archaeology, X I V , 1922, p. 237; cf. Anderson, Spiers, Dinsmoor, Architecture of
Ancient Greece, chronological table (489 B.C.), and A.J .A. 1932, p. 311, n. 3. The theory was first
presented in my lectures of 1920 at the Metropolitan Museum.
 T H E DATE OF T H E OLDER PARTHENON 417

I I I , the earlier Kimonian fill behind wall 3 (469-461);

l i e , the latest pre-Persian fill overflowing from wall 2 (490-480) ;
l i b , the main pre-Persian fill behind wall 2 (490-480) ;
l i a , the pre-Persian fill in the V-shaped trench (490-480) ;
lb, the pre-Persian accumulation behind wall 1 (before 490) ;
la, the Mycenaean and geometric accumulation behind wall l. 1
Of all the various objects contained in this débris,—architectural blocks, marble
and terracotta roof tiles, fragments of sculpture, inscriptions, bronzes, terracottas,
and potsherds,—the most vital perhaps are those of the last category. For pieces
of larger structures and works of art would presumably have been destroyed and
buried only after important alterations of design or cataclysms of war.2 But the
fragile painted vases were exposed to continual breakage. Hence it might logical^
be presumed that, while the larger objects buried in any given deposit might have
been created at a much earlier period, the potsherds on the contrary would have
accumulated (though in diminishing proportions) right down to the very date of the
deposit. For this reason we may adopt the potsherds as our scale, following the
principle that any stratum must be as late as its latest sherd.
Now for the first time the completion of the great publication of the Acropolis
vases, by Graef and Langlotz, 3 permits one to test, on fairly accurate stratigraphical
grounds, the various proposals as to the dates of the strata. 4 To do this, however,
our first task must be a reconstruction of an inventory of the sherds discovered
south of the Parthenon, on the basis of the very incomplete records now available.6
1 The subdivision of stratum I is in accordance with the sketch by Iiawerau, cf. Wolters in Graef,

Vasen der Akropolis, I, p. xix and fig. 2. The following concordance will explain the slight discrepancies
from Dorpfeld's and Heberdey's numbers :
Dorpfeld's I I I = Heberdey's I I I c = my I I I
f " ' nib \ T T

II \ " Ilia / l l c

[ " II " lib

lib " lib " lia
2 Heberdey has undertaken to examine the terraces on the basis of the fragments of poros sculptures

(.Porosskulptur, pp. 1-9, 231-240).

3 Graef and Langlotz, Die antiken Vasen von der Akropolis zu Athen (2 vol. text and 2 vol. plates),

1909-1933.
1 Such an investigation had been suggested by Wolters in 1925 (in Graef, op. cit., I, p. xxi). Langlotz
concludes, however, that " t h e sherds do not permit us to advance beyond the results obtained from the
building stones themselves" (II, p. ii).
s This analysis (for which the preliminary list of sherds of known provenance was prepared by

Miss Constance H. Curry), was commenced before the appearance of the third (and final) part of Vol.
I I of the Vasen der Akropolis (1933), with the valuable preface by Langlotz, who likewise analyzes the
sherds discovered south of the Parthenon. Though a second analysis might now seem to be super-
fluous, comparison shows that there were several omissions in the Langlotz lists (and, in a few cases,
inclusions of sherds for which the text gives no authority) ; and another reason for republication is the
fact that several areas of interest in this connection are not discussed by Langlotz (e.g., " D " and
" D 2 " ) , nor did he investigate thoroughly the older reports by Ross, Wolters, and Kavvadias. In
giving these revised lists, furthermore, it has seemed desirable to designate with " P " all sherds which
were noted as having been found in " Perserschutt " (though it must be recalled that this designation
may equally Well mean " T y r a n n e n s c h u t t " ) and also to indicate by asterisks (*) all sherds which are
mentioned as having been burnt in a conflagration. Again, the lists have been subdivided in accord-
ance with the days of discovery, on the principle that the sherds found on later days in a given area are
more likely to have been at lower levels.
418 W I L L I A M B E L L D I N S M O O R

Ross states that the surface of the ground in 1835 was at the level of the second
marble step of the Parthenon (and even higher at the west on account of the earth-
work of 1826). 1 He cleared away the late débris, wherever the masses of fallen
blocks left him free to do so, along the entire south flank, and here, furthermore, he
excavated two pits. One of these, in 1835, near the southwest corner (below the west
angle column of the Parthenon), 2 revealed natural rock at a depth of 5.50 m. or
twelve poros courses below the marble crepidoma. The other pit, in 1835-1836, at
the southeast corner,3 apparently did not go very deep at the corner itself (only five

II 20
F I G . 4 . — P L A T E BY THE BBYGOS M A S T E B , DISCOVERED BY ROSS
From Graef and Langlotz, Vaseti der Akropolis, I I , pi. 1

poros courses below the marble) ; but part of this trench, which extended to the south
wall, seems to have gone down to bed rock, and in any case it penetrated below the
Periclean stratum (V) and into the "Perserschutt." As Ross admits, "Unfortu-
nately at that time the potsherds were not so carefully observed as the larger finds." 4
He states that " t h e booty is very rich in sherds of very beautiful painted vases," 6
and alludes to "many sherds of vases with black figures on red ground," 6 but cites
only one of these, the foot of a "little-master" kylix with an inventory inscription of
Athena (I 1920). 7 Of more importance, because they were of a later period and,
nevertheless, were definitely found below or in the burnt débris ("Perserschutt"),
»Ross, Arch. Aufs. I, pp. 82-83.
2 Ibid., p. 88, pi V 2; the text disagrees with the plate in stating that it was between columns 1 and 2.

'Ibid., pp. 102, 104-106, 106-108, 109-112, 138. pl. V 3-4.

4 Ibid., p. 139. s Ibid.., pp. 107-108. « Ibid., p. 106.
7 Ibid., p. 108; the number (11920) is that of the publication in the Vasen der Akropoli.?, where, how-

ever, it was not identified by Graef.

 T H E D A T E OF THE OLDER P A R T H E N O N 419

were three red-figured pieces, a plate by t h e Brygos painter (II 20), 1 a complete "owl
skyphos " with fragments of others, and p a r t of a terracotta many-spouted lamp. 2
" T h e s e valuable fragments were taken out from a great depth (10 to 12 feet) against
t h e foundation of t h e temple, from an earth stratum which was mixed with charcoal,
broken archaic bronze figurines, broken painted roof tiles, and other débris; this was
t h e rubbish from the sanctuaries and buildings destroyed by the Persians on t h e
Acropolis." 3 Any argument t h a t these sherds might have been mixed with Persian
débris at a later date, toward 460 B.C.,4 during t h e erection of the south wall, is con-
troverted by the b u r n t condition of the plate, which had suffered in the conflagra-
tion. As Ross points out, "Special importance seems to reside in the skyphos, the
lamp, and the b u r n t sherd [of the plate] ; for they prove (even though in the same
excavation were found sherds with black figures, which might have belonged to
earlier decades) t h a t at the time of the burning of the old temple, and so during t h e
Persian wars, ceramics with red figures on black ground were in use, and they show
to what power the drawing had attained." 5 T h u s the identification of the " P e r -
s e r s c h u t t " in turn thrust back t h e beginnings of the red-figured style far earlier t h a n
460 B.C.,6 where it had been placed by contemporary archaeologists. B u t these ob-
servations were ignored by Gerhard and J a h n , and were regarded by K r a m e r as
paradoxical; 7 in consequence, the study of Attic vase-painting was retarded for half
a century, 8 and Ross was not fully justified until the great excavation of 1888.
E v e n as late as 1920 Langlotz wrote: "Of all the red-figured vases of the ripe severe
style t h a t have come to light on the Acropolis, only the plate made known by Ross
can be dated with practical certainty before 480 B.C."9 And the "owl skyphos,"
which was equally certainly found below Persian débris, has been so completely for-
gotten t h a t the numerous examples of this type have been most recently assigned to
the fourth century, the late fifth century, the third quarter of the fifth century, or
t h e second half of the fifth century. 1 0 Langlotz alone comes to t h e right conclusion
when he assigns them " w i t h o u t exception from the period about 500, datable by
means of the shape." 1 1 A comparison of the few known facts of discovery in-
dependently brought me to t h e same conclusion. Several such " g l a u k e s " from the
1
Ross, op. cit. I, p. 140, pl. X. The plate was easily identified; cf. Diimmler, Bonner Studien, p. 74;
Klein, Euphronios, pp. 38, 52; Hartwig, Meisterschalen, pp. 338-339; Tonks, "Brygos," in Mem. Amer.
Acad. Sc. XIII, 1904, p. 113, no. 42; Hoppin, R. F. Vases, I, p. 141, no. 103; Langlotz, Zeitbestimmung
der strengrotfigurigen Vasenmalerei, p. 99; Beazley, Att. Vasenmaler, p. 182, no. 85 ; Graef and Langlotz,
II 20.
2
Ibid., pp. 139-140, pl. IX. The lamp seems to be otherwise unpublished, and the skyphos appears
to have vanished from the Acropolis (it is not included in the Graef-Langlotz publications).
3
Ibid., II, p. 330; cf. I, p. 139.
* Furtwangler formerly even argued that the date ante quern was merely the beginning of the Peri-
clean temple in 447 B.C. ("Bronzefunde au,S Olympia," Abh. Berl. Akad. 1879, iv, p. 6 = Kleine Schriften,
I, p. 341, n. 3; cf. however, 50. Berl. Winckelmannsprogramm, 1890, p. 162).
6 5
Ross, op. cit., I, p. 140. Ibid., I, pp. vi-vii, 139-141; II, pp. 330-331.
» Kramer, Arch. Zeit. 1852, Anz. 198. » Cf. Graef, Jb. Arch. I. 1893, Anz. 13-14.
s
Langlotz, Zeitbestimmung, p. 99.
"Pottier, Vases antiques du Louvre, III, p. 294, no. G 618; Orsi,Mon. Ant. XXIII, p. 914, n. 1; Robin-
son, Harcum and Iliffe, Catalogue of the Greek Vases in Toronto, I, p. 183, nos. 373-375 ; Pellegrini, Vasi
dipinti dette necrop.felsinee, p. xlvi; Beazley, CVA. Oxford, I, p. 40, pl. XLVIII 9, and II, p. 114, pi
L X I I 1 - 2 ; Miss Lamb, CVA. Cambridge, I, p. 35, pl. X X X I V 3.
n
Langlotz, Gr. Vasen in Wiirzburg, p. 119.
420 WILLIAM B E L L DINSMOOR

Acropolis were listed by Graef and Langlotz (II 529-537, and about 30 uncata-
logued fragments), and, while none of them can be identified with the Ross material,
the only one of recorded provenance (II 529) was found on October 8, 1888, in
"Perserschutt" (presumably, according to the date, toward the southwest corner of
the Parthenon), while another (II 531) was injured by fire. Again, in the Corin-
thian excavations of 1928, in a group of graves otherwise containing only material
of the last quarter of the sixth century and the first quarter of the fifth, Shear found
two of these owl skyphoi, one of them in grave X X X with two black-figured Attic

1! ¿30 it m
F I G . 5 . — O W L S K Y P H O S F O U N D B Y R O S S , AND F R A G M E N T S F R O M " P E R S E R S C H U T T

From Ross, Arch. Aufs. I, pi. I X , and Graef and Langlotz, Vasen der Akropolis, I I , pi. 40

skyphoi.1 Two other owl skyphoi were also found together with black-figured ware
in graves V I and CLVI at Corinth.2 An example found at Marion has been assigned
to the fourth century merely because this was the greatest period of importation of
Attic pottery into Cyprus; 3 but this seems an inadequate reason, and the graves
excavated at Marion date from various periods, archaic and classical.4 While some
of the more loosely executed examples may be fairly late, 5 it seems not unreasonable
to infer that the type was established and the better examples painted at about 490
B.c.,6 perhaps with some relationship to the introduction of the olive wreath en-

1 Shear, A.J.A. 1928, pp. 493-495.

! I am indebted to Dr. and Mrs. Shear for permission to mention these yet unpublished pieces.
3 Gjerstad, Illustrated London News, 1929, Nov. 9, p. 808. 1 Woodward, J.U.S. 1929, p. 237.
6 For Apulian imitations, see CFA. Italia, pi. 295, 6-7.
6 The most complete bibliography, with a list of about 106 examples, has been compiled by Robinson

(I.e.). To these we may add, besides the Acropolis pieces, 35 others : two at Angers (Valotaire, Rev.
Arch. X V I I , 1923, pp. 64-65, no. 26-27), two others at Oxford (Beazley, CVA. Oxford, p. 114, pl. L X I I
1-2), twelve in Berlin (Furtwängler, Vasensammlung im Antiquarium, pp. 729, 733-734, nos. 2587,
2595-2605; cf. Schöne, Comm. Mommseni, p. 654, no. 12; Hackl, "Merkantile Inschriften," in Münch.
Archäol. Studien, p. 53, no. 594, cf. p. 73), five at Würzburg (Langlotz, Gr. Vasen in Würzburg, p. 119,
nos. 614-618, pl. 217), one in Vienna (Masner, Antike Vasen im K.K. Oester. Museum, p. 59, no. 379),
two at Goluchow (Beazley, Greek Vases in Poland, p. 63; Bulas, CVA. Pologne, p. 27, pi. 40,4-5), three
at Copenhagen (Blinkenberg and Johansen, CVA. Danemark, p. 123, pi. 161. 3-5), fragments at the
Hague (Scheurleer, CVA. Pays-Bas, pl. 87, 1, 3), one from Camarina (Mon. Ant. X I V , p. 848, fig. 59),
one from Caulonia (Mon. Ant. X X I I I , p. 914, fig. 154), two others at Corinth (graves VI and CLVI,
unpublished), one from Marion in Cyprus (Gjerstad, III. London News, 1929, Nov. 9, p. 808), one at
Michigan (Van Ingen, CVA. Michigan, p. 31, pi. 15, 2), and one in Providence (Luce, CVA. Provi-
dence, p. 31, pi. 23, 3).
 T H E D A T E OF T H E O L D E R PARTHENON 421

circling Athena's helmet on the Athenian coins, likewise a symbol of the victory at
Marathon. 1
Other excavations were carried out along the south flank of the Parthenon in 1845
by Pittakis; but these were limited to a general lowering of the surface level,2
continued in 1859-1860. 3 In 1864 Ziller sank two pits against the Parthenon foun-
dation, one at the southeast corner (where Ross had not reached the bottom) and
the other between columns 5-6 from the east, both descending 10.77 m. or twenty-
two poros courses below the marble crepidoma;4 no record was made of any sherds
found. The surface level, between 1860 and 1888, was between six and eight poros
courses below the marble crepidoma; 5 and some stones of the Ergasterion were even
then visible.6 In other words, this part of the Acropolis presented in 1888 the same
general appearance that it had received by 1860; and our knowledge of the chro-
nology of red-figured vase-painting had hardly advanced beyond the stage that it
had reached before 1835.
It was with great surprise, therefore, that Kavvadias found advanced red-figured
vases in 1886-1888 mingled with the Persian debris.7 The finds on the Acropolis
now definitely established the fact that the severe (ripe archaic) red-figured style
was already in full development by 480 B.C., and that the origins of the style must
recede far back into the sixth century; other comparisons have established the date of
the beginning as about 530 B.C. In short, the chronological development of red-fig-
ured vases is now well established, with few exceptions; 8 and we may now reverse the
process, employing the chronological scale thus ascertained for the dating of the strata.
For the sherds found during the earlier stages of the great excavation of 1888,
however, we are dependent upon the brief reports by Wolters 9 and Kavvadias. 10 It
seems almost unbelievable, in these days of carefully recorded observations of all
minor finds, that during the great Acropolis excavations of 1885-1890 the potsherds
were so neglected. Of the 4313 vases (including pinakes) described by Graef and
Langlotz,—and these include a vastly greater number of individual sherds,—by far
the majority are totally without provenance. Apart from 220 sherds found in
earlier excavations,11 the recorded yield of the excavations of 1885-1890 consists
merely of 550 sherds.12
1 Cf. Seltman, Athens, Its History and Coinage, pp. 102-109.
2 Kavvadias and Kawerau, Ausgrabung, p. 12. »Ibid., pp. 14, 16.
1 Ziller, Zeitschr. f. Bauwesen, X V , 1865, pp. 33-54, pi. A, B .

6 This is shown by Ziller's sections (pi. B ) and by Dorpfeld's photographs (Ath. Mitt. 1892, pi. I X ;

1902, pi. X I I I ) , in which the surface grass plainly appears three to four courses below the proposed
Older Parthenon grade level (as indicated by the strongly marked ridge at the bottom of the fourth
course below the marble crepidoma). 6 Dorpfeld, Ath. Mitt. 1888, p. 224.

I Studniczka, Jb. Arch. I. 1887, p. 159; Winter, ibid. p. 229; Kavvadias, AeXr. 'Apx- 1888, pp. 43, 44,

83; Kavvadias and Kawerau, Ausgrabung, p. 38; Graef, Jb. Arch. I. 1891, p. 43; 1893, Anz. 14-15;
Wolters, in Graef, I , pp. xi, xxi. 8 One such exception is the "owl skyphos" discussed above.

' Wolters, Ath. Mitt. X I I , 1887, pp. 386-388; X I I I , 1888, pp. 107-110, 225-229, 437-442.
» Kavvadias, ATXR. 'A PX . 1888, pp. 10-13, 3 0 - 3 2 , 4 3 - 4 5 , 54-55, 8 1 - 8 3 , 1 0 1 - 1 0 5 , 1 2 2 - 1 2 5 , 1 5 3 - 1 5 4 , etc.
I I These include the 2 identified above as found in 1836 by Ross, 4 east of the Parthenon in 1863/4,

44 northeast of the Propylaea in 1 8 6 4 , 1 others of 1863/4, 100 under the Museum in 1867, 1 east of the
Parthenon in 1873, 8 south of the Propylaea in 1875, 1 dated 1877 and 2 from the north wall in 1877
identified by Wolters (Graef, I , p. xvi), 48 east of the Parthenon in 1882/3, and 8 uncertain.
12 These include 1 near the north wall in 1885, 6 merely dated 1886, 62 east of the Parthenon in 1886,

1 east of the Parthenon in 1887, 3 at the " t h o l o s " in 1887, 223 south of the Parthenon in 1888 (as dis-
 422 WILLIAM B E L L DINSMOOR

For these reasons it becomes imperative to scan even the summary reports for
casual allusions to sherds. Thus we find that Wolters mentions for 1888 the following
twenty items: January-March (op. cit. pp. 387-388), II 439, 195, 1051, 1367, 1354;
April-May (pp. 108-110), 1 I 202, 222, 345A, 601b, 601a right piece, II 325, 439;
June-July (pp. 227-228), I 732, II 504, 1296, an unidentified onos, and a complete
Mycenaean cup;2 August-December (p. 441), I 2560, II 3, 1345. Similarly the
monthly reports of Kavvadias for 1888 mention the following twenty-five items:
Jan. 13-Feb. 12 (op. cit. pp. 12-13), 3 a) II 1051, 0) possibly I 2509 or 2585, y)
II 352d, 5) possibly I 2450 or 2458, «) I 601a right piece; Feb. 13-March 12 (pp. 30,
32), I 202 and 222, a) II 439, 0) II 325, 7) II 1354, S) II 1367, e) I 2570a; March 13-
April 12 (p. 44), e) II 1295, s) II 762; May 13-June 12 (p. 83), I 1408 and 732; June
13-July 12 (pp. 102-103), a) I 1220, 0) I 2519, 7) II 1296, ó) I 607r, e) I 2582; Aug.
13-Sept. 12 (p. 154), a) II 589 left piece, 0) II 352a, 7) I 2560, 5) II 238. Of these,
eleven are duplicated notices; thus we are actually concerned with thirty-four
sherds. Combining the two lists, and inserting the statements as to provenance
contained in these reports, we obtain the following list:
Jan. 13-Feb. 12, above the poros stratum southeast of the Parthenon and also in the fill behind the
south wall (without proper observation of the strata so that they probably belong both to the Kimonian
and to the Periclean strata I V and V ) : 4 1 601a right piece, possibly I 2450 or 2438, possibly I 2509
or 2585; I I 352d, 1051.
Feb. 13-March 12, in the fill southeast of the Parthenon, behind the south wall but beyond the poros
stratum (i.e., south of the polygonal wall, and so in strata I I I , I V or V ) : 12570a; I I 3 2 5 , 4 3 9 , 1354, 1367.
Feb. 13-March 12, in the humus stratum (la) at the same place: I 202, 222.
March 13-ApriI 12, in the fill southeast of the Parthenon, between the retaining wall and the south
wall (and so in strata I I I , I V or V): I I 762, 1295.
January-March, in uncertain location southeast of Parthenon: I I 195.
April-May, in the fill behind the south wall, southeast of the Parthenon or east of the Museum
(uncertain strata): I 345A, 601b; I I 439 fragment.
M a y 13-June 12, in the fill south of the east end of the Parthenon, between the poros stratum and
the south wall (i.e., south of the polygonal wall, and so in strata I I I , I V or V ) : I 732, 1408.
June 13-July 12, under the poros stratum south of the middle of the Parthenon and also beyond this
toward the south wall (i.e., in strata l i b , l i e , I I I or I V ) : I 607r, 1220, 2519, 2582.
June-July, in the fill south of the middle of the Parthenon: I I 504.
Id., in the humus (stratum I b ) : I I 1298, an unidentified onos.
Id., in the humus (stratum l a ) : a complete Mycenaean cup.
Aug. 13-Sept. 12, between the west half of the Parthenon and the south wall: I 2560; I I 238, 352a,
589 left piece.
August-December, in uncertain locations along the south wall: I I 3, 1345.
cussed below), and 254 west of the Parthenon in 1888. Wolters mentions a loutrophoros found at
the north wall in 1886 (Graef, I, pp. xi, xxi), 119 sherds east of the Parthenon in 1886 (ibid., p. xiv),
28 sherds at the " t h o l o s " in 1887 (ibid., p. xii); but not all of these can now be identified.
1 For the identification of I 2 0 2 and 2 2 2 with the finds mentioned in Ath. Mitt. pp. 1 0 8 - 1 0 9 , A Í X T .

'Apx- 1 8 8 8 , pp. 3 0 - 3 1 , and Kavvadias-Kawerau, Ausgmbung, p. 1 0 4 (cf. pi. E 1 , at level 1 4 1 . 0 0 at lower

left corner), see Wolters, in Graef, I , p. xxxiv with note 1.
2 Wolters describes the onos fragment as a so-called cover tile, but refers, for the form, to Jb. Arch. I.
I I , 1887, p. 69, thus showing that it was an onos. The Mycenaean cup, illustrated by Wolters, does
not appear in Graef and Langlotz (cf. Wolters, in Graef, I, p. xxxiv).
3 Allowance is made for the reports dated by the Julian calendar.

4 In fact, Wolters (op. cit. p. 388) expressly says that some of the sherds were found in the upper

strata mixed with marble chips; and I I 1 0 5 1 of about 450 E.C. could not have been in an earlier deposit
than that of Pericles.
 THE DATE OF THE OLDER PARTHENON

Of these t h i r t y - f o u r sherds, only eight (I 202, 222, 732, 1408, 2560; I I 3, 439, .504)
were identified b y Wolters, Graef and Langlotz. 1 I t has now been possible to iden-
t i f y t w e n t y - t w o additional sherds, 2 of which only five, as we shall see, were more
exactly designated a t t h e m o m e n t of discovery. 3 I n other words, f r o m these old
reports we o b t a i n new clues as to t h e provenance of eighteen of t h e published frag-
ments, one f r o m Ross and t h e others f r o m Wolters and Kavvadias. 4
T h e more careful observation of t h e sherds was initiated by Wolters a n d Graef
on J u n e 4, 1888, when, f o r t u n a t e l y for our purpose, m o s t of t h e south flank of t h e
P a r t h e n o n yet remained unexcavated. W o r k h a d already been s t a r t e d at t h e east
end of this area, a n d h a d proceeded as far as t h e east wall of t h e Ergasterion, to a
considerable d e p t h ; 5 indeed, t h e site of Ziller's great pit, extending 714 m . f r o m t h e
corner of t h e foundation, h a d been cleared down to t h e rock. 6 T h e eastern half of
t h e length of t h e P a r t h e n o n was subdivided into lettered sections, " A " a n d
" C " being inside, and " B " a n d " D " outside, t h e polygonal retaining wall (2)
which nearly bisects t h e length of t h e area (Fig. 7). T h e western half was not so
subdivided; b u t (since " E " was reserved for an area still f a r t h e r west) we m a y
a p p l y " C 2 " and " D 2 " respectively to t h e strips inside a n d outside t h e polygonal
wall along t h e west half of t h e P a r t h e n o n ' s flank. A special section is t h e triangle
near t h e southwest corner of t h e P a r t h e n o n , composed b y t h e Pelasgian wall (1)
a n d t h e two arms of t h e poros ashlar retaining wall (3). E a c h of these seven sec-
tions will first be considered separately; a f t e r w a r d s we m a y combine t h e resulting
evidence for t h e successive s t r a t a .
A t t h e t i m e of t h e excavation, because of t h e prevailing theory t h a t t h e Older
P a r t h e n o n was K i m o n i a n , all t h e s t r a t a above I a n d below V were regarded as
K i m o n i a n , with, a t most, a few years difference between those n o r t h a n d south of
t h e polygonal retaining wall (2). As Dorpfeld a f t e r w a r d s stated, "All s t r a t a with
f r a g m e n t s of sculpture and building stones were t h e n designated as ' P e r s e r s c h u t t , '
regardless of whether they were f o u n d behind t h e polygonal retaining wall or behind
t h e poros ashlar retaining wall or behind t h e south citadel wall; regardless of whether
t h e f r a g m e n t s showed or did not show traces of fire; regardless of whether t h e y con-
sisted solely of poros or whether m a r b l e f r a g m e n t s were intermingled with t h e m .
E x a c t segregation of t h e finds in accordance with the various terraces was u n f o r t u -
nately not carried out, because it was regarded as superfluous; only in t h e case of a
few objects is t h e provenance so exactly known t h a t t h e a t t r i b u t i o n to one of t h e
various periods is now possible." 7 I t is no wonder, a f t e r such a revelation, t h a t t h e
1
For I 2560 they give no provenance.
Namely, I 345A, 001a right, 601b, 607r. 1220, 2519, 2570a, 2582; II 195, 238, 325, 352a, 352d, 439
fragment, 589 left piece, 762, 1051, 1295, 1296, 1345, 1354, 1367. These do not include the four un-
certain pieces, namely the Mycenaean cup, the onos, I 2450 or 2458, or I 2509 or 2585. The apparent
discrepancy in the case of I 601a, which Graef reports as found on Sept. 5 in "C2," can only mean that
3
his reference is to the left piece. These are I 607r, 2519, 2582; II 238, 1296.
4
Of these, five are mentioned only by Wolters (I 345A, 601b; II 195, 439 part and 1345), seven only
by Kavvadias (I 1220, 2570a; II 352a, 352d, 589 left, 762, and 1295), and five by both (I 601a right
piece; II 325, 1051, 1354, and 1367). In addition, we now have the provenance of I 1920 found by
Ross.
5
According to Wolters (in Graef, I, pp. xviii, xxiii, xxiv), Dorpfeld's photograph {Ath. Mitt. 1902,
pi. XIII) represents the state of the excavation at this moment. ' Wolters (ibid., p. xxiii).
7
Dorpfeld, Ath. Mitt. 1902, p. 408.
F I G . 0 . — P L A N OF E X C A V A T E D SECTORS SOUTH OF W E S T P A R T OP P A R T H E N O N

From Giaef and Langiotz, Vaxen der Akropolis, I, p. xxvi = I I , p. vii

F I G . 7 — P L A N OF E X C A V A T E D SECTORS SOUTH OF E A S T P A R T OF PARTHENON

From Graef and Langiotz, Vasen der Akropolis, I, p. xviii = I I , p. v
 T H E D A T E OF T H E O L D E R PARTHENON 425

value of observations as to the discoveries in the "Perserschutt" has frequently

been doubted.1 But, as Wolters remarks, the observations were in most cases so
carefully made that a reconstruction of the stratification is not an impossibility.2
We must continually bear in mind, however, that the designation "Perserschutt" is
by no means to be taken as literal evidence that the find was made in post-Persian
débris; the original designation, even when applied under the mistaken impression
that the Older Parthenon and its fill were Kimonian, was retained in the final publi-
cation on the principle, as Wolters put it, that "one cannot exchange horses while
crossing a ford." 3 Wre have only to remember that the notice "Perserschutt" as
applied in 1888 means that the sherd was found in a deposit of débris, and we must
depend on other evidence for the decision as to whether the deposit is later or earlier
than 480 B.c. To distinguish the earlier débris the term "Tyrannenschutt" was
coined by Dickins. 4 The term was an apt one so long as the date 510-506 B.C. was
accepted for the deposit; 5 but it becomes less satisfactory if we ascertain that the
débris was thrown here twenty-two years after the expulsion of the tyrants.
The lower portion of section " A , " approximately the first quarter of the length
of the south flank of the Parthenon (from the west edge of Kawerau's pit to the east
wall of the Ergasterion), and north of the polygonal retaining wall, was excavated
June 4-8. 6 Thirty-seven sherds found in this area were dated and, in twenty-six
instances, marked "in débris near Parthenon foundation." Ten others lack this
descriptive phrase but clearly belong to this area because of their dates of discovery ;
these latter are distinguished in the following table by marks of interrogation (?);
and seven of them bearing the additional designation of "Perserschutt" (which, as
noted above, may equally well mean "Tyrannenschutt") are here marked " P " .
Observations were begun in stratum l i b (as is shown by the fact that the working
level was then twelve courses below that with the projecting ridge and so sixteen
courses below the lowest marble step of the Parthenon, 7 that is, two courses above
stratum I), and were continued down through stratum I. But in view of the fact
that Graef includes in the "Perserschutt" with absolute certainty finds made be-
tween June 4 and July 4 with merely the date and no other notations,8 it seems clear
that all the thirty-six sherds in question came from strata l i b or l i a , i.e., above
stratum I. As for the exception, I I 731, which was found "in the lowest stratum
above the natural rock" on the very first day, the mere fact that the excavation was
conducted by clearing first in each layer a strip along the Parthenon foundation,
makes it possible that I I 731 was found in the V-shaped trench Ha. And to this, in
accordance with our conservative principle of attributing all ambiguous sherds to
the upper of the possible strata, we shall assign it : 9

1 Cf. Furtwangler, Aegina, p. 353; Langlotz, Zeitbestimmung, p. 99. 2 Wolters, in Graef, I p xx

3 Ibid., p. xix. 4 Dickins, Acropolis Museum, I, p. 9.
5 Heberdey also objects that this name depends on purely subjective interpretation, though, in effect,
he agrees with it {Deut. Literatur-Zeitung, 1913, p. 171; Porosskulptur, pp. 3-4).
6 Graef and Langlotz give June 4-7; but two fragments are definitely dated June 8.
7 Ath. Mitt. 1902, pl. X I I I . s Wolters, in Graef, I, p. xxx.
9 This list differs from that of Langlotz (II, p. vi) in the omission of 1387a and 838, for which neither

date nor provenance are mentioned in the text, and in the addition of the certain fragments I 357b,
360b, 716, 798, 856, 1045a, 1165, 1322, and of the probable fragments I 2516, 2562, and I I 1156. I t is
426 W I L L I A M B E L L DINSMOOR

Geometric and VI Cent.

Proto-Attic First Half Late B . F . ca. 490
Strata l i b
or I l a :
June 4 I I 731
June 5 I 357b, 360a I 645a I 615m, 1165, 1295a, 2010,
2028a, 2038, 2192 (?),
2260 P (?), 2271c P (?),
2462 (?)
June 6 276, 286 798, 811, 856, 1045a,
360b 1669a, 2045, 2047, 2048,
2549 P (?), 2562 P (?),
I I 1156 P (?)
June 7 597c, e, I 1164, 2046, 2271b P (?),
2683 (?) 2516 P (?)
June 8 716, 1322

The lower portion of section " B " , corresponding to " A " in that it lay east of the
Ergasterion, but south of the polygonal retaining wall, was excavated June 11-20, 1
Work was at this time far below the level of stratum V. Nineteen sherds were
listed, besides some unspecified Mycenaean and geometric material. The earlier
finds, dated June 11-12, were marked "outside inner retaining wall" or "between
middle retaining wall and south wall"; 2 one piece showed traces of fire (*). The
sherds of June 15-29 were marked " d e e p " 3 ; in alternate days they seem to have
been taken from the north and south edges of the area, as those of June 15, 18, and
20 were marked " deep between retaining wall and south wall" (once with the addi-
tional indication "Perserschutt," " P " ) 4 , while those of June 16, 19, and 29 were
marked "deep near south wall." B u t all the sherds found on June 15-29 " d e e p "
outside the retaining wall, unless otherwise designated, were clearly in "Perser-
s c h u t t " according to Graef13. Thus eight of the sherds of June 15, 18, 20 must have
been in strata l i e or I I I , while all six sherds of June 16, 19, 29 must have been in
stratum I V . As for the exceptions, I 1911 was found on June 15 close to the poly-
gonal wall; " t h e stratum is older than the beginning of the Parthenon"; this must
have been stratum l b , which was penetrated on that day, as noted by Wolters, 6 while
Mycenaean and geometric sherds were in stratum la. 7

conceivable that we might include also I I 1252, of which the date 5:VI:1886 suggests a mistake for
5:VI:1888; but this would not alter our conclusions. Fragment I 2260 is wrongly numbered 2960 in
the text.
1 Graef and Langlotz give the dates June 15-25; but the sherds from this section bear the dates

June 11-29. 2 In the case of I 2094 the position outside the retaining wall is " n o t certain."
3 In the case of I 345B the designation " d e e p " was omitted; but it was clearly found at a low level

together with I 437 and 1371.

4 Also I 345B was found on June 20, " i n Perserschutt with charred remains."
6 Wolters, in Graef, I, p. xxx.
6 Wolters, in Graef, I, pp. xix, xxv, with sketch by Kawerau (ibid., fig. 2).

' This list differs from those given by Langlotz in the inclusion of I 475, 737, 985, 1371, 2094, 2523,
and I I 1185.
 THE DATE OF THE OLDER PARTHENON 427

Proto-Attic VI Cent,
and V I I Cent. First Half Late B . F . ca. 500 ca. 470
Strata l i e ,
I I I , or I V :
J u n e 11 I 2094 (?)
J u n e 12 I 475 I 2523 * 737
Stratum I V :
June 16 985, 1275d, I I 738
I I 1185
J u n e 19 I I 214
J u n e 29 I 1273
Strata l i e
or I I I :
J u n e 15 1543, 1944 P,
2511
J u n e 18 1299, 2567
J u n e 20 I 345B P, 437 P 1371
Stratum l b :
June 15 1911
Stratum l a :
J u n e 22 I, p. xix (Myc. & geom.)
J u n e 23 I, p. xix (Myc. & geom.)

Section " C " occupies t h e second q u a r t e r of t h e length of t h e south flank of t h e

P a r t h e n o n , f r o m t h e east wall of t h e Ergasterion t o t h e middle (ninth) column of t h e
P a r t h e n o n . T h e excavations were conducted here J u n e 2 6 - J u l y 17.1 Of t h e twen-
ty-eight sherds listed, I 591a bears merely t h e d a t e a n d t h e n o t a t i o n t h a t it c a m e
f r o m section " C " ; this a n d I 593 might have been in s t r a t u m V. 2 T h e finds of
J u l y 6 occurred lower, in s t r a t a l i b or l i e , fill c o n t e m p o r a r y with t h e building of t h e
P a r t h e n o n foundation. 3 T h e next f r a g m e n t s bear t h e notation t h a t t h e y were
1
Langlotz gives the dates June 26-July 15, stating t h a t the finds of July 16-17 were made during t h e
refilling operations of the latter half of the month. But the circumstantial accounts of their discovery
(I 2582 " i n C in black earth," I 2526 "close by Parthenon in black e a r t h " ) would indicate t h a t they
were found during excavation, which was therefore continued until July 17.
2
Langlotz states t h a t 1591a comes from the " T y r a n n e n s c h u t t , " and 1593 bears the notation " i n C
in black earth." B u t Wolters notes t h a t the finds as late as July 3-4 were still in late strata (in Graef,
I, p. xxvi), so t h a t stratum V would be a possibility. As noted below, the " P e r s e r s c h u t t " was not
certainly reached until July 5, the black earth not until July 7. Langlotz also assigns I 2691 to the
" T y r a n n e n s c h u t t " ; but this, an antefix (?) of the third quarter of the VI century, bears merely t h e
date July 3, 1888, and so might equally well have come from section " D " . Also ambiguous, in t h a t
they might have come either from " C " or from " D " , are I I 749 (July 3) and I 988 (July 4), both " i n
uncertain s t r a t a . " We therefore eliminate I 988, 2691, and I I 749 from our table.
3
Graef says t h a t the finds of July 5-14 in " C " , unless specifically stated to be in black earth, were
in " Perserschutt " ; and of those catalogued only the sherds of July 6 fall into this category. Even here
we find ambiguity: I 506 is dated 6: VI: 1888, which must be a mistake for 6: V I I : 1888 (so Langlotz,
II, p. vi), and is noted merely as coming from " C " ; the same is true of I 1994 and 2002, with regard
to which Graef surmises t h a t they were found in débris near the Parthenon foundation or in the black
earth beneath (the year 1882 given for 1894 is clearly a misprint); and I I 45, on which the notation
" C " was accidentally omitted (though required by the date of discovery), is specifically stated to have
been found in poros strata contemporary with the Parthenon foundation. Langlotz (II, p. vi) is
clearly in error in stating t h a t I I 45 was found in the black earth. Hence the attribution of the finds
of July 6 to the black earth depends on the attribution of I 593 of July 4, which is in itself doubtful.
Certainty occurs first on July 7, when I 2481 was found " i n C in the black earth under the poros
s t r a t u m , " labelled in such detail as to suggest t h a t it was t h e first fragment so found.
428 WILLIAM BELL DINSMOOR

found "in C, in black earth" (i.e. stratum I), which was excavated July 7-17. 1
Fragments injured by fire are marked with an asterisk (*). In this area was the
second pit excavated by Ziller in 1864; but care was taken to avoid deceptive
material from its refilling,2 and in any case the material in the black earth could not
be subject to doubt.3
VI Cent.
First Half ca. 540-520 Late B.F. ca. 510 ca. 500
Stratum l i e
or V:
June 28 I 591a
July 4 593
Strata l i b
or l i e :
July 6 506 I 1994, 2002 I I 45
Stratum l b :
July 7 I 2481
July 9 1069
July 10 914*, 1767,
1835, 2482,
2519
July 13 912, 2535 1369, 2137,
I I 1296
July 14 607r, 793a*, I 1466, 1547, I 2570b*
1140b
July 16 456
July 17

Section " D " corresponded to "C" in that it lay opposite the second quarter of
the south flank of the Parthenon, east of the central column, but south (outside) of
the polygonal retaining wall. Only seven sherds are specifically noted as having
come from this region;4 and only those of Aug. 4-9 were found "deep" and so de-
serve confidence; but as they lay in "Perserschutt" they probably were in strata
lie, III or IV.5
1
Wolters gives the dates July 6-15; but the dates July 7-17 are derived from the fragments, as noted
above. 2 Wolters, in Graef, I, p. xxv.
3
This list differs from that given by Langlotz (II, p. vi) in the exclusion of I 2691 (see note 2, p. 427),
and the addition of I 607r, 793a, 1069, 1767, 2682, and I I 1296 (all, except I I 1296, bearing specific
allusions to the black earth; and this single doubtful sherd I I 1296 is specifically stated by Wolters
(Ath. Mitt. 1888, p. 228) to have been found in the humus stratum). The omission of the designation
" C " on I 2519 and 2526 was clearly accidental, as both were found in the black earth, the latter "close
by Parthenon," and the former a t a date when " C " was being excavated. Errors in dates occur on I
1767, 2224, and I I 4 5 6 , noted as 10: VI, 14: VI, and 16: VI respectively; since all three were found " i n
C in black earth," while I I 456 bears the additional note " 2 m. from Parthenon," it is clear that VI was
a mistake for VII (cf. I 506 as discussed above, p. 427, n. 3).
1
These are not listed by Langlotz. To them we might possibly add I 2255a, found between the
middle retaining wall and the south wall on July 12, a date too early for section " D 2 , " and too late
for " B . " See p. 427, note 2, with regard to I 988, 2691, and I I 749, which, if admitted in section " D , "
would belong to the upper strata.
s
Thus I 635 and 1773 are both marked " d e e p " ; of the former, Graef notes that it was found just
west of the east wall of the Ergasterion, and he states that the latter was in "Perserschutt." This
would apply also to 1622a. Graef noted that sherds found " d e e p " Aug. 4-29 were in " P e r s e r s c h u t t "
(I, p. xxx). Of the other sherds listed, I 1057 was found " i n uncertain strata," and 1263 "in D in
 T H E DATE OF T H E OLDER PARTHENON 429

Older B.F. Late B.F. R.F.

Strata III,
IV or V:
July 3 / 4 I 1057
July 5 1149 I I 1056
July 7 I 1263
Strata l i e ,
I I I or I V :
Aug. 4 635 622a
Aug. 9 1773 P

As section " C 2 " we m a y designate t h e prolongation of t h e excavation of " C "

along t h e western half of t h e south flank of t h e P a r t h e n o n , between Aug. 2 a n d
Sept. 5. 1 T h i s section received n o special designation a t t h e time, most of t h e six-
teen sherds listed being noted merely as " i n s i d e retaining wall," 2 a n d generally
with t h e additional n o t e " o p p o s i t e P a r t h e n o n columns nos. 2 - 8 (or 3 - 7 , or 3 - 4 )
f r o m west." 3 Little is s t a t e d in regard t o t h e s t r a t a in which these were f o u n d : for
m o s t (i.e., I 1260, 1734; I I 96, 99, 114, 238, 551) only t h e position in plan is given;
others (I 607, 914, 1784) are mentioned as occurring in t h e fill near t h e P a r t h e n o n
f o u n d a t i o n (to I 914 is a d d e d " i n uncertain s t r a t u m " ) , and two (I 2560, 2592) even
lack t h e p r o v e n a n c e ; all these m i g h t therefore h a v e been in s t r a t a l i b or l i e . Wol-
ters states t h a t n o t until Aug. 28 was t h e " K i m o n i a n " s t r a t u m (II) reached, 4 a n d
Graef adds t h a t sherds f o u n d on Aug. 30 a n d a f t e r w a r d s were certainly in " P e r -
s e r s c h u t t . " 5 B u t I 601a and 612 were f o u n d "close above r o c k , " which Graef
rightly interprets as in accumulation older t h a n t h e " P e r s e r s c h u t t . " 6
Ca. 540-530 Late B.F. ca. 510 ca. 500
Stratum V:
Aug. 3 I 2119c
Strata l i b
or l i e :
Aug. 29 607 I 914
Aug. 30 I I 99 12592 P
Aug. 31 1260,1734, 96,114, II238
1784 551a
Sept. 1 I 2560 P I 2515
Stratum l b
or l i a :
Sept. 3 612
Sept. 5 601a
the late cistern," so t h a t neither would seem to be of stratigraphical value. This would apply also
to I 1149 and I I 1056, which were found at the same time.
1
This list differs from t h a t given by Langlotz (II, pp. vi-vii) in the addition of I 607, 914, 1784,
2560, and I I 551a. In his list, I I 2515 is a misprint for I 2515.
2
I n the case of I I 238 the word " u n t e r h a l b " is probably a misprint for " i n n e r h a l b " with reference
to the retaining wall.
3
For I 2592 we have only the date Aug. 30 " i n Perserschutt," whence Langlotz rightly concludes
t h a t it was found in the section here known as " C 2 , " and includes it in his list; no other section was
being excavated at this moment. Likewise 12560 has only the date and the notation " P e r s e r s c h u t t " ;
but in this case the provenance is supplied by the reports of Wolters and Kavvadias (see pp. 422, 423).
4 5
Wolters, in Graef, I, p. xxvi. Ibid., I, p. xxx.
6
Graef infers also t h a t 12515 was found in a s t r a t u m older than the "Perserschutt," but the grounds
for this assumption are not evident and the sherd cannot be used for dating s t r a t u m l b .
430 WILLIAM B E L L DINSMOOR

The section corresponding to " C 2 , " outside the polygonal retaining wall and
opposite the west half of the south flank of the Parthenon, though likewise originally
without designation, may here be known as " D 2 . " This area was excavated im-
mediately after " C 2 , " from Sept. 3 to Oct. 27. Of the twenty sherds listed, the
first (Sept. 5-15) were marked "outside retaining wall," 1 sometimes with the
additional phrase "before SW. corner of Parthenon" (I 681b; I I 305), or "before
Parthenon column no. 4 (or 5-6 or 5-11) from west" (II 268, 344, 355, 1019), or
"near (polygonal wall) steps" (I 2090b). To this same section outside the poly-
gonal retaining wall belong fragments marked " t " ("deep"); 2 only one sherd
(I 837) seems to have been so marked, on Sept. 27; and Graef adds that it was found
at the level of the "Perserschutt." Here, too, we must include a series marked
" h , " found between the Pelasgian wall and the south wall on Oct. 2-27, and there-
fore at low levels. Of these, I 613 and 717 are said to have been found in uncertain
strata, I I 1116 deep in the "Perserschutt," and I 1691 " d e e p " ; the last, though
Graef concludes that it must be in a stratum earlier than the polygonal retaining
wall, may also have been in "Perserschutt" like the Dipylon fragments found on the
same day.3 We can only conclude that the sherds of Sept. 27 Oct. 27 must have
been in strata l i e , I I I or IV, while those found before Sept. 27 might have been
in stratum V. 4
Geometric and Middle
VII Cent. VI Cent. Late B.F. 500/490 490/480 480/470 470/460
Strata IV
or V:
Sept. 5 I I 344
Sept. 6 I 2073 777 (?)
Sept. 11 I 681b I I 305 P I I 1019
Sept. 13 I 2090b P 268 I I 355
Sept. 15 2496
Strata lie,
I I I or IV:
Sept. 27 837
Oct. 2 1432
Oct. 4 613
Oct. 6 305 P
Oct. 17 I , p. 23
Oct. 18 717
Oct. 19 I 290 717
Oct. 20 I, p. 23
Oct. 22 638
Oct. 25 I I 1116 P (?)
Oct. 27 I, p. 23 (P) I 1691

The "triangle" composed by the Pelasgian wall (1) and the two arms of the poros
retaining wall (3) was excavated Sept. 18-Oct. 6, 1888. The sherds, necessarily
belonging to stratum I I I , were in many cases marked merely with the date (other
1 To I 2073 is added, "not certain." * Graef, I, p. xxvii, note 2.
3 Graef, I, p. 23. Graef's general statement (p. xxx) that fragments found in " h " after Oct. 9 must be
older than the polygonal wall, and so in stratum I, conflicts with his detailed statements of provenance
and particularly with the presence of the late piece I I 638. 4 These are not listed by Langlotz.
 THE DATE OF THE OLDER PARTHENON 431

areas excavated at this same period having special designations) ; 1 but some were
specially marked "Dreieck"; some were also noted as found in "Perserschutt"
(here marked " P " ) . Some sherds were found on Sept. 11-12 at a level just above
that which later revealed the triangle, and so apparently in stratum V. A few
sherds dated after Oct. 6 are of uncertain origin; but it happens that they include
solely fragments of vases found in stratum III. Fragments injured by fire are
marked with an asterisk. On account of the possibility of doubt, the forty-five
certain "Dreieck" pieces are here listed first:2
560-540 Late B.F. ca. 510 ca. 500 ca. 490 ca. 480 ca. 470
Stratum V:
Sept. 11 II 564
Sept. 12 I 2336 P
Stratum III :
Sept. 19 I 912 P 815 P,
897, 925
Sept. 21 2211c II 461c P H 758*
Sept. 22 883 910* 641*, 778*,
925
Sept. 24 641*, 704 P, I I 761 P
969
Sept. 25 295, 730*, 641*, 947*
758*
Sept. 27 815 P
Sept. 28 295
Oct. 1 623*, 730*,
758*, 898*
Oct. 2 I 2511
Oct. 3 773b P 1884 P* 800 761
Oct. 4 2203b
Oct. 6 758*
Oct. 9 758*
Oct. 10 758*
Oct. 11 758* 641*
Oct. 12 641*, 815 P
Oct. 17 461c P 758*
Oct. 27 897

Many of the sherds found in the triangle were not so specified, however, being
marked merely with the date, as contrasted with those found on the same dates in
other areas and distinguished by special designations. This rather dangerous
method of procedure leaves a slight amount of doubt with regard to the authen-
ticity of the evidence offered by these fragments, so that they are here grouped in a
separate list, which includes forty-eight sherds dated Sept. 18 to Oct. 6, 1888, but
without provenance: 3
1
Graef, I, pp. xxvii, note 2, xxx.
2
This list differs from that of Langlotz in the omission of I 967a, 2256, 2391b, 2434, 2455, 2489, and
II 62, 227, 608, 683, 736, 813, 831 and 978, which lack the provenance and so are included in the follow-
ing list. We also omit II 81c and 987, which were found west of the Parthenon and are included in the
Langlotz list by error. We add, however, 1912, and I I 461c, 564, 730, and 947, all found in the triangle.
3
This list includes I 967a, 2256, 2391b, 2434, 2455, 2489, and II 62, 227, 608, 683, 736, 813, 831, and
978 of the Langlotz list, as well as twenty-four others which were found between Sept. 18 and Oct. 6.
432 WILLIAM BELL DINSMOOR

560-540 Late B . F . ca. 510 ca. 500 ca. 490 ca. 480 ca. 470
Stratum I I I :
Sept. 18 II ' pa
Sept. 19 I I 813
Sept. 20 I 1236b P,
1845 P
Sept. 21 I 645c P 669 P, I 967a P
1066a P ,
! P
Sept. 24 ì P, 967c P 608, I I 22*, I I 626,
>P 813 978* 831
Sept. 25 742d P , I 2434 813 22* 626
980a, b, c
Sept. 27 2391b 665a P , 967a P 977* 625
2489
Sept. 28 2356 P 736*
Oct. 1 2278 P
Oct. 2 I I 7*
Oct. 3 2124 P
Oct. 4 662 P , 448
2255c P,
2455
Oct. 5 432 P 723 P 350
Oct. 6 2455 II 813

We now turn to the second part of our investigation, a discussion of the contents
of the successive strata. 1 And, in order to obtain a clearer idea of the relationship
of sherds to strata, we may begin with the latest, 2 which at the same time provides
a definite chronological basis, that of the Kimonian south wall. For the lower part
of this wall (4), antedating the Periclean addition (5), dates from the period im-
mediately after the battle of the Eurymedon about 469 B.C. (Plutarch, Kimon, 13,
6; Kimon and Lucullus, I, 5; Glor. Ath. 7; Nepos, Cimon, 2, 5; Pausanias, I, 28, 3). 3
Consequently the fill (IV) behind this lower part of the south wall should contain
material ranging down to about 469 B.C., but no later.
Of the fragments recorded as found before June 4, along the south wall of the
> I t will be noted t h a t certain fragments tabulated as found in different locations belong t o the same
vases. This is true of I 607, 645, 912, 914, and 2511. In t h e last case there is no difficulty; pieces
found in section " B " (June 15) and t h e triangle (Oct. 2) could both belong t o stratum I I I . But pieces
of I 607 and 914 were found in s t r a t u m l b ( " C " , July 14 and 10), others in l i b or I l e ( " C 2 " , Aug. 29);
similarly a piece of I 912 was found in s t r a t u m l b ( " C July 13) and another in I I I (triangle, Sept. 19) ;
and one piece of 1645 was found in s t r a t u m H a or l i b ( " A " , June 5), another in I I I (triangle, Sept. 21).
I n these four instances it is now impossible to decide whether the provenance should be doubted, or
whether some old débris was actually employed in filling the upper strata. We are not confronted by
these problems, however, since none of the sherds in question is important for dating purposes.
2
Literally, the latest would be t h e Periclean wall (5) and stratum V; but so much of the latter had
been removed before 1888 t h a t we cannot now secure adequate evidence.
3
T h e exact date of the battle of the Eurymedon is uncertain. Diodorus (XI, 60) gives 470/69 B.C.,
and is followed by Beloeh (Or. GesckII 1, p. 67, I I 2, p. 185) ; Busolt (Or. GeschIll 1, p. 143 with
note 2) made out a strong case for 468; Perrin (Plutarch's Cimon and Pericles, pp. 8, 191-192) gives
467; Walker (Cambridge Ancient History, V, p. 53, table I) prefers 467 or 466; Judeich (Topographie 2,
pp. 74, 209) accepts 465 B.c. I have adopted W 7 ade-Gery's preference for 469 B.c. for t h e reasons cited
by him (J.H.S. 1933, p. 81 and n. 87), agreeing with Diodorus.
 THE DATE OF THE OLDER PARTHENON 433

Acropolis, those of J a n u a r y were gathered from above the poros stratum southeast
of t h e Parthenon, and also from the fill behind the south wall, without proper
segregation; some must have been in Periclean deposit (V), such as I I 1 0 5 1 of about
450 B.C., and others in t h e upper strata mixed with marble chips. 1 Those listed in
February, March, and M a y , however, were found in the fill behind t h e south wall
(those of March and M a y being specified as south of the retaining wall); b u t again
we cannot be certain as to t h e levels. T h e latest pieces were t h e following: 2
12570a (pi. 107), X 76a; edge of black-figured pinax
dedicated by des (Lolling's restoration
Andokides is purely hypothetical); "first dec-
ade of fifth c e n t u r y " (Graef).
I I 3 2 5 (pi. 20-22), B 76; kylix of Makron, of Glau-
kon period, about 480 or 480-470 (Langlotz;
Beazley, Att. Vasenmaler, p. 212, no. 10).
I I 439 (pi. 35-36), C 16; white-ground kylix of
Euphronios, by the Pistoxenos painter, with
kalosname Glaukon, about 480-470 (Langlotz;
Beazley, Att. Vasenmaler, p. 260, no. 5).
I I 762 (pi. 62), G 389; volute krater of Oreibelos,
b y the painter of the Deepdene amphora, about
470 (Langlotz; Beazley, Att. Vasenmaler, p.
295, no. 19).
I I 1295, N 14; edge of column krater with name of
Sostratos (the same name on I I 1294 dated
end of sixth century by Langlotz).
I I 1 3 5 4 (pi. 93), N 104; lamp dedicated to Athena,
about 500 (Langlotz).
I I 1367 (pi. 91), F 97; fragment by Kallis, dedi-
cated t o Athena Hygieia, about 470 (Lan-
glotz).

If we could assume t h a t these were all in

t h e lower stratum (IV), they would be in
II 75»
perfect conformity with t h e date of the
F I G . 8 . — L A T E S T S H E R D S FROM STRATUM I V ,
south wall, since the latest pieces (II 325,
B A R E L Y A N T E D A T I N G 4 6 9 B.C.
439, 762, 1367) descend only to about From Graef and Langlotz, Vasen der Akropolis,
470 B.C. B u t t h e upper stratum (V) would II, pi. 20-22, 52, 62
be equally plausible and equally satisfac-
tory as to date. I n any case, it would seem t h a t I I 325, which Graef believed to
have come from "Perserschutt, " 3 was in stratum IV.
T h e sherds definitely known to have come from stratum IV, inventoried after
J u n e 4, range from Dipylon ware (in " D 2 " ) to about 470 B.C. T h e latest, selected
from those found by Ross in stratum IV (or I I I ) , and those tabulated for sections
" B , " " D , " and " ¿ 2 , " are the following: 4
i Wolters, Ath. Mitt. 1887, p. 388.
!
T h e numbers of the sherds are those of the final publication, to which are added the provisional
numbers (X 76a, B 76. etc.) as used by Beazley and others.
•< Graef, Jb. Arch. I. 1891, p. 43.
4
The late sherds of section " B " were definitely found in stratum IV. The only late piece in " D "
434 W I L L I A M B E L L DINSMOOR

I I 20 (pl. 1), B 14; plate by " t h e Brygos master in his middle period" (Langlotz; Beazley, Att. Vasen-
maler, p. 182, no. 85), and so about 490-480 B.c.
I I 214 (pi. 12), A 38; kylix of the "Panaitios period, about 5 0 0 " (Langlotz).
I I 638 (pi. 32), H 12; loutrophoros fragment of about 480 (Langlotz).
I I 738 (pi. 62), G 33; bell krater, not dated, but those in its vicinity (e.g., 732, 734, 737, 739, 740, 742)
assigned by Langlotz to about 470 B.c.
Unidentified owl skyphos; about 490 B.c. (?)

Thus, while we have comparatively little recorded material from this fill, such evi-
dence as it yields is in conformity with the date of the wall which supported it. I t
was the débris of 480 B.C., the true "Perserschutt," with A few accretions of the next
decade. The latest sherds (II 638 and 738) were painted only a few years before
the Kimonian wall was constructed and the fill (IV) thrown in behind it.
We turn next to the poros retaining wall (3) and to its fill (III), which Dorpfeld
had assigned to the period immediately after the battle of Marathon. The contents
of the strata behind the eastern portion of this wall, opposite the southeast corner of
the Parthenon, can no longer be distinguished. But at the southwest corner of the
Parthenon the elbow of poros retaining wall composed, with the Pelasgian wall as
the hypotenuse, a right triangle of which the contents were very carefully segre-
gated. I t will be noted from the preceding tabulation (p. 431) that this fill was of
uniform character from top to bottom. 1 For our purpose we need to consider only
the sherds which have a bearing upon Dijrpfeld's date of the wall, namely, those of
490 B.C. or later.
At least six vases (II 295, 623, 730, 758, 898, and 910), and probably also five
others (II 350, 608, 683, 736, and 813), found within this fill, are definitely assigned
to about 490 B.C. :
I I 295 (pi. 15), B 69; kylix "of the first decade of the V century by a painter closely related to the
Brygos painter" (Langlotz).
I I 350 (pi. 23), B 117; kylix of about 490-480 (Langlotz).
I I 608 (pi. 45), G 362; amphora of 490 (Langlotz, p. vii).
I I 623 (pi. 48), G 223; amphora, "early work of the Tyskiewicz painter" (Langlotz; Beazley, Att.
Vasenmaler, p. 116, no. 38), of 490 (Langlotz, p. vii).
I I 683, G 334; hydria, placed by Langlotz between 681 (ca. 500) and 684 (ca. 480).
I I 730 (pi. 58), G 11; kalyx krater of the "Xleophrades painter, ca. 4 9 0 " (Langlotz; Beazley, Att.
Vasenmaler, p. 72, no. 29).
I I 736 (pi. 58), G 36; kalyx krater of 490 (Langlotz, p. vii).
I I 758 (pi. 63), G 25; volute krater of the "Syriskos painter, ca. 4 9 0 " (Langlotz; Beazley, Att. Vasen-
maler, p. 158, no. 3).
I I 813 (pi. 73), G 342; column krater of 490 (Langlotz, p. vii).
I I 898 (pi. 77), G 224; fragment of uncertain form, by the "Kleophrades painter" (Langlotz; Beazley,
Att. Vasenmaler, p. 75, no. 58), of 490 (Langlotz, p. vii).
I I 910, G 337; fragment of uncertain form, by the "Eucharides painter" (Langlotz; Beazley, Att.
Vasenmaler, p. 96, no. 33).

(II 1056) was at so high a level that it might have been in stratum V, and so is here omitted. And in
section " D 2 " we confine our observations to the material found on or after Sept. 27 (see p. 430), yield-
ing only one late sherd.
1 Cf. Wolters, in Graef, I, p. xxvii.
 THE DATE OF THE OLDER PARTHENON 435

E v e n if w e h a d n o t o t h e r e v i d e n c e t h a n these, t h e r e f o r e , it w o u l d be h a r d l y possible
t o d a t e t h e fill i m m e d i a t e l y a f t e r t h e b a t t l e of M a r a t h o n ; f o r it w o u l d b e a l m o s t
impossible t o i m a g i n e t h a t as m a n y as eleven n o t a b l e v a s e s w o u l d h a v e been b r o k e n
up and buried immediately after their manufacture.
I n a d d i t i o n t o these, f u r t h e r m o r e , w e h a v e a t least eight v a s e s ( I I 6 4 1 , 7 0 4 , 7 7 8 ,
8 1 5 , 8 9 7 , 9 2 5 , 9 4 7 , a n d 9 6 9 ) , a n d p r o b a b l y also t h r e e o t h e r s ( I I 2 2 , 9 7 7 , 9 7 8 ) , f o u n d
w i t h i n this fill a n d n e v e r t h e l e s s d e f i n i t e l y assigned t o a b o u t 4 8 0 B.C.:

II 22 (pi. 2), B 18, plate by a "backward painter of the years about 480" (Langlotz).
II 641 (pi. 51), G 371; loutrophoros with "drawing of about 480" (Langlotz).
II 704 (pi. 54) F 90; oinochoe (?) of 480 (Langlotz, p. vii), with details characteristic of 480-470 (Lang-
lotz).
II 778 (pi. 68), G 194; stamnos by the "Syleus painter, about 480-470" (Langlotz; Beazley, Att.
Vasenmaler, p. 162, no. 28).
II 815 (pi. 74), G 122; column krater by the "Harrow painter" (Langlotz; Beazley, Att. Vasenmaler,
p. 471, no. 55), of 480 (Langlotz, p. vii).
II 897 (pi. 77), G 140; fragment of uncertain form, of 480 (Langlotz, p. vii).
I I 925 (pi. 76), G 274; fragment of uncertain form, undated, but those in its vicinity (e.g., 911, 936,
941, 949, 954) assigned by Langlotz to about 480 B.C.
I I 947 (pi. 79), G 278; fragment of uncertain form, cf. 925.
I I 969, G 18; fragment of uncertain form, about 480 (Langlotz).
II 977 (pi. 83), G 336; fragment of uncertain form, undated, but those in its vicinity (e.g., 972, 974)
assigned by Langlotz to about 470 B.C., but, like 974, probably to be placed before 480 because
of traces of fire.
II 978, G 361; fragment of uncertain form, cf. 977.

T h e s e e l e v e n vases, a t t h e earliest, could n o t h a v e been b r o k e n u p u n t i l t h e P e r -

sian d e v a s t a t i o n of t h e A c r o p o l i s in 4 8 0 B.C.; a n d t h e i r e v i d e n c e is sufficient t o
p r o v e t h a t w e a r e dealing w i t h a " p o s t - P e r s i a n " wall.
B u t t h e sherds tell us e v e n m o r e . F o r we h a v e r e m a i n s of a t l e a s t one v a s e ( I I
7 6 1 ) , a n d p r o b a b l y also of f o u r o t h e r s ( I I 4 4 8 , 6 2 5 , 6 2 6 , 8 3 1 ) , w h i c h w e r e f o u n d in
t h e fill a n d n e v e r t h e l e s s belong t o a b o u t 4 7 0 B.C.

I I 448 (pi. 33), C 25 ; white ground kylix, placed at end of series by Langlotz.
I I 625 (pi. 48), G 79; amphora of the "Harrow painter, about 470" (Langlotz; Beazley, Att. Vasen-
maler, p. 120, no. 45).
I I 626 (pi. 49), G 170; amphora, not dated, but those in its vicinity (e.g., 625, 627, 628) assigned by
Langlotz to about 470 B.C.
I I 761 (pi. 66), G 114; volute krater by the "Copenhagen painter about 460" (Langlotz; Beazley, Att.
Vasenmaler, p. 156, No. 5). But this date seems rather late for the Copenhagen painter, whom
Beazley (Attic, Red-Figured Vases in American Museums, p. 63) associates with the Syriskos painter
and the Syleus painter; and Miss Swindler (Ancient Painting, p. 192) likewise places the Copen-
hagen painter in the decades 490-470; therefore we may prefer 470 (or earlier) rather than 460 B.C.
I I 831, G 63; column krater, not dated, but the preceding example (830) dated by Langlotz about
470 B.C.

I t is n o t e w o r t h y t h a t t h e fill w i t h i n t h e t r i a n g l e c o n t a i n s a l m o s t t h e s a m e in-
g r e d i e n t s as t h a t i m m e d i a t e l y behind t h e s o u t h w a l l : considering o n l y t h e m a t e r i a l
of 4 9 0 B.C. or l a t e r , w e h a v e 4 1 % ( 1 1 vases) of a b o u t 4 9 0 , a n d 4 1 % ( 1 1 vases) of
a b o u t 4 8 0 , a n d 1 8 % (5 vases) of a b o u t 4 7 0 B.C., s h o w i n g a g r a d u a l r e d u c t i o n t o w a r d
t h e t e r m i n a l d a t e , as is t o be expected. I t will be n o t e d , f u r t h e r m o r e , t h a t t r a c e s
436 WILLIAM BELL DINSMOOR

II HB

I! ri-.'.",
!( liifl«

4 Wi
A *

Ii 701
FIG. 9.—LATEST SHERDS FROM STRATUM I I I , BARELY ANTEDATING 4 6 9 B.C.
From Graef and Langlotz, Vasen der Akropolis, II, pl. 33, 48, 49, 66
 T H E DATE OF T H E O L D E R PARTHENON 437

of fire were observed on very many of the sherds from stratum I I I in the triangle
(I 1884; I I 7, 22, 623, 641, 730, 736, 758, 778, 898, 910, 947, 977, 978), clear evidence
that we are here concerned with débris of the Persian conflagration of 480 B.C.1
Again we are dealing with the true "Perserschutt," with some additional débris
of the next decade. In other words, the potsherds agree with the architectural
evidence in determining the date of the elbow of the poros retaining wall (3) as
immediately after 470 B.C.; it is neither pre-Persian (Dorpfeld) nor Periclean
(Heberdey), but dates from the period midway between, that of Kimon, coeval
with the great south wall.2
As for the fill contemporary with the erection of the Parthenon foundation,
strata l i a , l i b , and lie, there now seems to be no way of distinguishing their con-
tents. In section " A , " as we have seen, all the sherds must have come from
strata l i a or l i b (with the possible exception of I I 731 which is here included for
conservative reasons, the only alternative being the older stratum lb) .3 In section
" C " there is less ambiguity, since the sherds can easily be distinguished from those
found in stratum l b " i n the black earth." Thus we may conclude that all sherds
found on July 6 in section " C " lay in the upper strata l i b or lie. 4 As for section
" C 2 , " where the records were not very carefully kept, we may assume that all
sherds listed for Aug. 29-Sept. 1 were in the upper strata l i b or l i e , as contrasted
with those found "close above rock" on Sept. 3-5 and so presumably in the black
earth. 5
The sherds listed as coming from the strata 11 a c contemporary with the Parthe-
non foundation, as excavated in these three sections, are of all styles from the
geometric to the red-figured. For our purposes, we are concerned only with those
of the end of the sixth century and later, that is, in the period during which Dorp-
feld assumed that the foundation was in course of erection. In this connection it
would be desirable to make a more careful analysis of the later black-figured sherds
(I 615m, 716, 798, 811, 856, 914, 1045a, 1164, 1165, 1260, 1295a, 1322, 1669a, 1734,
1784, 1994, 2002, 2010, 2028a, 2038, 2045-2048, 2192, 2260, 2271b-c, 2462, 2515,
2516, 2549, 2560, 2562, 2592, and I I 1156) than I am capable of producing. We
may, however, select the following significant pieces:
I 2515 (pi. 103), X 16; natural clay ground pinax, 500 (Langlotz, p. vii).
I 2560 (pi. 107), X 75; natural clay ground pinax, the drawing of the head recalls Oltos (Graef) and
so of about 510 B.C.
I 2562 (pi. 106), X 47; natural clay ground pinax of "end of the VI century" (Graef).
I 2592 (pi. I l l ) , X 100; white ground pinax "probably of beginning of the V century" (Graef).

The dates of the red-figured sherds are more easily defined, as follows:
I I 45 (pi. 3), A 182; kylix of Oltos " o f the time about 5 1 0 " (Langlotz, p. vi; Beazley, Att. Vasenmaler,
p. 17, no. 77).
I I 96 (pi. 7), A 53; kylix of the school of Epiktetos, 510 (Langlotz).
1 The same conclusion may be drawn from the fact that calcined terracotta simas identical with those

found in the triangle (Graef, in Wiegand, Porosarchitektur, p. 112) were also found by Ross in stratum
IV (Ross, Arch. Aufs. I, pp. 105-106, 109-110, 139; Buschor, Tondacher der Akropolis, Simen, p. 24).
2 Langlotz, with this same material, reached the conclusion that "the triangle was not filled up with

débris until after the departure of the Persians, and perhaps the wall (3) was not built until then"
(II, pp. vii-viii). 3 See p. 425. 4 See p. 427. 5 See p. 429.
438 WILLIAM B E L L DINSMOOR

I I 99, A 95; kylix of the school of Epiktetos, " a b o u t 5 1 0 " (Langlotz).

I I 1 1 4 (pi. 7), A 99; kylix of the school of Epiktetos, Leagros period, 510 (Langlotz).
I I 238, B 18; kylix of the "Panaitios period, about 5 0 0 " (Langlotz), with an inscription of Smikros
dedicating it to Athena, possibly the painter Smikros (for whom see Beazley, Alt. Vasenmaler,
p. 62).
I I 551 (pi. 41 and text fig.), F 6; kantharos of the "Kerberos painter, about 5 1 0 " (Langlotz; Beazley,
Att. Vasenmaler, p. 30, no. 8).
I I 731 (pi. 58), G 31; kalyx krater, " a b o u t 4 9 0 " (Langlotz), " n o t painted before 500" (Langlotz,
p. vi).

i
F I G . 1 0 . — L A T E S T S H E E D S FROM S T R A T U M I I , B A K E L Y A N T E D A T I N G 4 9 0 B . C .
From Graef and Langlotz, Vasen der Ab-opolis, I, pi. 103, 111; II, pi. 58

Among these easily datable sherds of the three last decades (and thorough
analysis of the late black-figured pieces would undoubtedly add others), therefore,
we obtain the following proportions: 65% (5 vases and 2 pinakes) of about 510,
and 2 6 % (1 vase and 2 pinakes) of about 500, and 9 % (1 vase) of about 490 B.C.
This gradual diminution in the numbers of sherds from each decade, terminating
abruptly with 490 B.C., is parallel to the similar diminution toward 470 B.C. among
the sherds found near the Kimonian walls, and in the same way forms an indication
that we are approaching the date of the foundation. We are obviously dealing,
not with the true "Perserschutt," but with the so-called "Tyrannenschutt," 1 of
which the date, however, must be two decades later than has generally been as-
sumed. I t would be difficult to imagine that as many as seven important pieces
1 The significant fact that none of the sherds from strata I l a - c has traces of fire (an absence which

agrees with that noted for the poros fragments: Dorpfeld, Ath. Mitt. 1902, p. 408; Heberdey, Poros-
slculptur, p. 6) is cited merely as a contrast to stratum I I I .
 T H E D A T E OF T H E OLDER P A R T H E N O N 439

manufactured about 510 B.C. h a d been broken up and buried four or five years
later; b u t when we add to these t h e testimony of t h e four pieces of 500-490 it
becomes increasingly clear t h a t the fill could not have been deposited as early as
506 B.C. or immediately thereafter. If we assume for t h e latest vase ( I I 731) a
date midway between t h e two offered by Langlotz, about 495 B.C., it would be
reasonable to assume t h a t this single example had been accidentally broken within
seven or eight years and was used as part of t h e fill in the V-shaped trench (stratum
I l a ) . I n other words, I I 731, found in t h e earliest layer of fill contemporary with
t h e foundation, directly above rock, 1 gives us t h e date post quern for t h e beginning
of t h e Parthenon foundation as about 495 B.C. or rather a few years later. 2
Lastly, we turn to s t r a t u m l b , t h e upper half of the black earth or humus which
covered this p a r t of t h e Acropolis to a depth of about 2 m. before t h e time of t h e
erection of t h e P a r t h e n o n foundation. Unfortunately, in the first section observed
( " A " ) , we cannot be certain t h a t the only sherd known to have come from a low
level belonged to this stratum. 3 I n section " B , " I 1911 must have come from
this stratum. I n t h e case of section " C " t h e sherds found in " t h e black e a r t h , "
as s t r a t u m I was then designated, were carefully noted; all of t h e twenty-two
sherds (with one exception, I I 1296) listed on July 7-17 bear such notes; and t h e
single exception is authenticated by other sources. 4 Of these sherds, one is red-
figured (II 456), found " i n C in the black earth about 2 m. from t h e P a r t h e n o n . " 5
I n t h e case of section " C 2 " t h e observations were not very careful; only t h e t w o
sherds listed on Sept. 3 - 5 were noted as having been found "close above r o c k "
and so presumably in stratum I ; but these are not red-figured.
T h u s t h e records show t h a t a t least one red-figured sherd was found in stratum
I. This agrees with t h e statements of Kavvadias, Wolters, Graef, and Dorpfeld
t h a t some red-figured sherds were found in this humus stratum, 6 —a statement
which Dorpfeld embodied in the following significant sentence: "All objects which
it contains, including some red-figured sherds, must be older t h a n t h e t e m p l e . "
I n itself this fact would not disagree with Dorpfeld's date of 506 B.C. for t h e begin-
ning of t h e temple, since the beginning of t h e red-figured style has been pushed
back to 530 B.C. B u t closer examination of t h e recorded sherd betrays a discrep-
ancy:

II 456 (pi. 38), E 12; skyphos, "fine drawing by the Euergides painter (?), about 510" (Langlotz).

I t might seem rash to attach too much importance to a single red-figured sherd.
B u t it will be noted t h a t t h e black-figured sherds from t h e same context are in com-
plete agreement. Some of these, to be sure, are fairly early (I 601, 607, 612, 793,
1
If, indeed, it was not already imbedded in the earlier stratum I (see p. 425).
2
With the same material Langlotz concludes that this fill "is chiefly of sherds earlier than 500 " and
consequently that the foundation dates from about 500 (II, p. vii).
3
See p. 425, note 2, concerning II 731. « See p. 428, n. 3.
5
See p. 427, note 3, for the exclusion of I 593, 1994, 2002, and II 45 from stratum I.
8
Kavvadias, A¿KT. 'A Px . 1888, p. 44; Kavvadias and Kawerau, Ausgrabung, p. 38; Wolters, Aih.
Mitt. 1888, p. 228, and Graef, I, pp. xviii-xix, xxvi; Dorpfeld, Aih. Mitt. 1902, pp. 386, 411. At least,
one of these sherds had been found before June 4 (Wolters, in Graef, I, pp. xviii-xix), possibly that
mentioned by Kavvadias.
440 WILLIAM B E L L DINSMOOR

912, 1140, 2481, 2526, 2535). Others, however, are late black-figured (I 914, 1069,
1369, 1466, 1547, 1767, 1835, 1911, 2137, 2224, 2482, 2519, 2570b, 2582, and II
1296): and it might be possible that a detailed analysis would yield some informa-
tion of significance. At present, I can cite only some fragments of votive tablets,
for which the black-figured technique was retained down to the beginning of the
fifth century.
I 2570b (pi. 107), X 76b; natural clay ground pinax of the "first decade of the V century" (Graef).
Graef suggests that this joined a piece (I 2570a) mentioned above; but the border has a different
pattern so that the junction is uncertain.
I 2582 (pi. 109), X 89; white ground pinax of the "end of the VI century" (Graef).

Thus, in the case of I 2570b, at least, we apparently have a fragment which was
not even painted until after the date to which Dorpfeld assigns the erection of the
foundation; and yet the fragment lay in the gradual and natural accumulation of
soil which antedated that foundation. Again, II 456 was painted at about 510
B.C.; it is most improbable that it would have been broken up and buried within
four or five years of its execution, and this improbability becomes an impossibility
when we combine it with the testimony of I 2570b. Finally, when we remember
that the superposed stratum II has been shown to be as late as 490 B.C., it becomes

I 4i7<rti I -¿.iS'i
F I G . 1 1 . — L A T E S T SHERDS FKOM STRATUM I, B A R E L Y ANTEDATING 4 9 0 B.C.
From Graef and Langlotz, Vasen der Akropolis, I, 107, 109; II, pi. 38
 T H E B A T E OF T H E OLDER PARTHENON 441

apparent that the underlying stratum I must have formed the exposed surface of the
Acropolis as late as 490 B.C., and that at this date the Older Parthenon had not yet
been commenced. 1
Thus the stratigraphical evidence agrees with that of the architecture in demand-
ing a post-Marathonian date for the very beginning of work on the basement of
the Older Parthenon. In other words, it compels us to abandon theories C, D, F ,
and G. Of the survivors, E has already been eliminated on architectural grounds.
We are left, consequently, with theory H as the sole possibility, a single Older
Parthenon, of marble, planned after 490 B.C.

3 . ASTRONOMICAL E V I D E N C E

In view of our acquired knowledge that not only the marble peristyle, but like-
wise its great foundation, must be assigned to the decade 4 9 0 - 4 8 0 B.C., and pre-
sumably to the early part thereof, we may seek greater precision by drawing upon
a third source of information. I refer to the orientation of the temple. This is an
aspect of the problem which may be approached with a certain amount of diffidence,
particularly in view of the fact that similar methods had yielded to Burnouf the
date 554 B.C. and to Penrose 1150 or 1495 B.C.2 NOW, however, with the restric-
tion of a single decade, such fantastic calculations may be eliminated, and we may
ascertain whether it is possible to obtain a result more in accordance with the
architectural and stratigraphical requirements.
T h e orientation of the Parthenon, and of the basement on which it stands, was
obtained most accurately by Penrose, carefully checked astronomical observations
giving 12° 53' north of true east. 3 T h e axis, prolonged at this angle, would pass
the crest of M t . Hymettus at a distance of 9075 m. and a height of about (ilO m.
above sea level, about 4,50 m. above the Parthenon basement, 4 the angle rising to
the visible horizon then being 2° 50'. B u t the sun " c a n be reckoned upon to throw
a strong illumination as soon as a quarter of his refracted orb is above the visible
horizon," which "allows the reduction of the apparent height of the mountain by
0° 2 2 ' . " 6 Thus we may estimate the angle to the horizon as about 2° 25'. 6 At
1 I t is to be noted that a few of the fragments found in the black earth show traces of fire (I 793a,

914, 2570b), which could hardly have been, in this stratum I. connected with the Persian conflagration.
I t is probable that they were injured in some earlier accidental fire, a form of destruction to which the
early temples were constantly exposed. As Wolters observes, the traces of fire are not exclusively of
Persian origin (in Graef, I, p. xx).
2 Burnouf, Legende Athénienne, pp. 5 3 - 6 1 ; Ville et Acropole d'Athènes, p. 179; Penrose, Trans. Royal
Soc. London, C L X X X I V , 1893, pp. 809, 827, 828; C X C , 1897, p. 45. A. Mommsen more conserva-
tively obtained Aug. 81, 458 B.c. (Bursian's Jahresbericht, L X X I I I , 1892, pp. 22-25; Feste der Stadt
Athen, p. 55).
3 Penrose obtained 12° 52 4 1 " west of true north for the east front of the Parthenon, and 12° 53' 5 9 "

north of true east for the north flank (Athenian Architecture,1 p. 8; cf. Trans. Royal Soc„ 1893, p. 809).
This is sometimes expressed as 257° 7' from south (Penrose, op. cit., Nissen, Orientation, p. 169). The
older and less accurate magnetic observations vary from 13° 0' to 14° 22' north of east (sec table given
by Nissen).
4 The measurements of distance and height are taken from the map of the Hymettus district by

Steffen and Kaupert (Curtius and Kaupert, Karten von, Attika, pl. IV), prolonging the axis of the
Parthenon to the crest and reading the level of the crest by means of the contour lines.
5 Penrose, Trans. Roy. Soc. 1893, p. 818.
6 This is the result obtained through direct observation by Penrose (ibid.., p. 809).
442 WILLIAM BELL DINSMOOR

this point the sun would rise, as seen from the Parthenon, on two occasions in t h e
year, sixty-three days before or after the summer solstice. 1 T h e earlier date, in
April, would be meaningless; 2 but the later date, in August, may well have had some
connection with the great day of the Panathenaic festival, the fiscal new year, the
third from the last day (rpirri tpdivovTos) of Hekatombaion. 3 To test this possibility,
we compare in the following table the theoretical Panathenaic dates (b), as de-
rived by orientation from the summer solstice (a), with the theoretical Panathenaic
dates (e) as derived from the lunar months (c, d). For convenience, the orienta-
tion of t h e rising sun (f) is added for each of these Panathenaic dates (e).
(a) 1 (b) (c) ! (d)= (e) (f) 4

Summer Solstice (>3 D a y s Later Nearest New Moon vo-jfirivia tjydivovTOs Orientation
Tp'iTï]
490 J u n e 29, 4 P.M. Aug. 31 Aug. 25, 7:45 A.M. Aug. 27 Aug. 24 16° 4'
489 " 28, 9 P.M. Sept. 12, 1:45 A.M. Sept. 14 Sept. 11 7° 34'
488 " 29, 4 A.M. " 1, 7:45 A.M. 3 Aug. 31 12° 53'
487 " 29, 9 A.M. Aug. 21, 9:00 P.M. Aug. 23 " 20 17° 46'
486 " 29, 2 P.M. Sept. 9, 10:00 P.M. Sept. 11 Sept. 8 9° 2'
485 " 28, 9 P.M. Aug. 29, 2:30 P.M. Aug. 31 Aug. 28 14° 16'
484 " 29, 2 A.M. Sept. 17, 3:00 P.M. Sept. 19 Sept. 16 5° 5'
483 " 29, 9 A.M. 7, 0:15 A.M. 9 6 10° 0'
482 " 29, 2 P.M. Aug. 27, 2:45 A.M. Aug. 29 Aug. 26 15° 10'
481 " 28, 9 P.M. Sept. 13, 8:15 P.M. Sept. 15 Sept. 12 7° 4'
480 " 29, 2 A.M. 2, 9:45 P.M. 4 1 12° 25'
1
The solstice dates are computed from Ginzel, Chronologie, II, p. 578, by transforming the tenths
of days into hours and adding 1 hr. 35 m. for the difference between Greenwich and Athens.
2
The dates of the new moons are computed from Ginzel, Chronologie, I, pp. 550-551, by transforming
t h e hundredths of days into hours and minutes, and adding 1 h. 35 min. for the difference between
Greenwich and Athens.
3
The corresponding first day of the Attic month is here tentatively assumed to have begun, not at
the time of the astronomical new moon (as I think was the case after Meton's reform; cf. Ardions of
Athens,^ pp. 313-315), but on the evening of the new light, when the crescent first became visible.
Beginning on t h a t evening, the next day would have been the first "business d a y " of the month, and
is t h a t recorded in the table (d).
4
Taking the orientation 12° 53' north of east on Aug. 31, 488 B.c. as the standard, the variations for
each of the preceding and following Panathenaia (e) are computed by means of the daily differences
obtained from Nissen's table (Templum, p. 246). I t must be pointed out t h a t the theoretical Pana-
thenaic dates (e) are modified in six instances by study of the calendar (table given below), and fall one
month earlier t h a n here shown. Consequently we should correct as follows : 489, Aug. 13, 20° 36'; 486,
Aug. 8, 22° 24'; 484, Aug. 18, 18° 87'; 483, Aug. 7, 22° 45'; 481, Aug. 14, 20° 13'; 480, Aug. 3, 24° 3''.

I t is possible, of course, t h a t some latitude should be allowed for any unknown

vagaries of the Attic calendar at this period; but it seems unlikely t h a t it would
have varied m a n y days from the moon. 4 Therefore, if we find a correspondence in
1 Nissen (Orientation, p. 169) says 64 days before or after the solstice, 30 days from the equinox.
But since this total must be 93 days or slightly less, it is permissible to rectify his calculation to 63 days,
particularly in view of the fact t h a t his own results are based on a calculation of 63 days (i.e., April 27
to Aug. 30 = 126 days inclusive).
2
I t was the adoption of this earlier date, for the purpose of coinciding with the heliacal rising of the
Pleiades, t h a t led Penrose astray.
3
This has been suggested b y Burnouf (Legende Athénienne, pp. 53-61), Nissen (Rh. Mm. X L ,
1885, p. 336; Orientation, pp. 169-171), A. Mommsen (Bursian's Jahresbericht, L X X I I I , 1892, pp.
22-25; Feste der Stadt Athen, p. 55), and Ginzel (Woch. kl. Phil. 1908, p. 259).
4
Among the evidences of such vagaries, we m a y cite t h a t of the battle of Plataea in 479 B.C. on
Boedromion 3 (Plutarch, Camillus, 19, 3; Glor. Ath. 7) or 4 (Plutarch, Aristeides, 19), equivalent to
 T H E D A T E OF T H E O L D E R P A R T H E N O N 443

the days between columns (b) and (e), we have reasonable assurance that we have
found the correct year. An exact correspondence occurs only in 488 B.C., which
is approximately the year required by the architectural evidence. Other years
which would be architecturally suitable, 489 and 487 B.C., show impossible astro-
nomical discrepancies of eleven days, equivalent to 5 degrees of orientation. Also
490 B.C. yields a discrepancy of seven days, and would be impossible in any case
since the Panathenaic festival occurred before the battle. 1 And 485 B.C., which
shows an admissible astronomical discrepancy of three days, is architecturally too
late.
Before we can accept this astronomical evidence with full confidence, it is de-
sirable to check it with reference to the Athenian calendar of this decade. The
single fixed date hitherto known was the solar eclipse of Oct. 2, 480 B.C., at 3 P.M.,
which according to Herodotus (IX, 10) was the cause of the Spartan retirement from
the Isthmus, apparently just after the battle of Salamis. The battle itself was

F I G . 12.—ORIENTATION o r THE OLDEE PARTHENON

Panamos 26/27, a difference of 7-8 days on which Plutarch dwells at length. But this was the local
Boeotian calendar, which was irregular as late as the second century, when we have Thouios 1 =
Homoloios 16 Kara 0c6v (I.G. VII, 517; Dinsmoor, Archons of Athens, p. 416 n. 2), a difference of 44-45
days. The discrepancies in the dates of the battles of Marathon and Salamis, on the other hand, must
be due to errors of other sorts, as discussed below.
1
Seltman (Athens, Its History and Coinage, p. 103) suggests that the battle, which he follows Busolt
in equating with Aug. 10,490 B.C., occurred before the Panathenaia. But this, of course, is impossible,
because, whatever the Julian date of the battle, the Attic date (in Metageitnion or Boedromion) must
have been later than Hek. 27/28. We may, I think, disregard Unger's proposal (Jb. kl. Phil. CXXVII,
1883, pp. 387-389) to locate the battle on Thargelion (1)6, 490 B.C., on the basis of intricate calcula-
tions and a corrupt passage of Aelian (Var. Hist. II, 25). On the other hand, the Panathenaia of 490
could have been later than the battle if we accepted Munro's date of 491 B.C. for this event (Cambridge
Ancient History, pp. 233, 245 n. 1, 253 n. 1). But apart from the inherent improbability of such a year
for Marathon, which disagrees with all our evidence (particularly that which specifies that it was ten
years before Salamis in 480 B.C., as fixed by the Olympic festival and a solar eclipse), the astronomical
discrepancy of seven days in Aug. 24-31 would be an obstacle against dating the Older Parthenon in
490 B.C. Conversely, the date of 488 B.C. for the Older Parthenon would add one more objection to
placing Marathon in 491, as much as three years earlier.
444 WILLIAM BELL DINSMOOR

fought on Boedromion 20 (Polvaenus, I I I , 11, 2), or " i n the twenties of Boe-

dromion" (Plutarch, Camillus, 19, 6), which agrees with the story of Herodotus
( V I I I , 65) with regard to the phantom Iacchic procession preceding the battle, and
with our more general knowledge that it was in the autumn (Herodotus, V I I I , 113,
130). The most probable date is Boedromion 24 or 25. 1 Plutarch's alternative
date, Mounichion 16 (Lysander, 15; (¡lor. Ath. 7), cannot be explained except as an
error; for this would have been about 145 days earlier, in the spring.2 With our
knowledge, therefore, that Boedromion 20 ( + ) was shortly before Oct. 2, we find
that the new year must have coincided with the new moon of July 5 (9:30 P.M.), the
actual new year's day probably being July 7, 480 B.c. With our new synchronism
Hek. 27/28 = Aug. 31, 488 B.C., we obtain a new year Aug. 4 / 5 , probably Aug. 5
(new moon Aug. 4, 8:15 A.M.).3 The interval comprised 2893 days, that is, 98 lunar
months, including six ordinary and two intercalary years. 1 T o avoid three ordinary
years in succession, it is necessary to assume that the two intercalary years were
486/5 and 483/2 B.C. Consequently 489/8 and 480/79 B.C. would likewise have
been intercalary, and 490/89 B.C. ordinary, placing the new year of 490 B.C. twenty-
five lunar months before Aug. 5, 488 B.C., and so on July 29, 490 B.C. The battle
of Marathon, according to Plutarch (Camillus, 19, 5; Mai. Herodot. 26; Glor. Ath. 7)
was fought on Boedromion 6, a date which cannot be reconciled with Herodotus,
as Plutarch himself admits (Mai. Herodot. 26). For Herodotus tells us that Pheidip-
pides, after traveling to Sparta in two days, received his answer on the 9th day of a
month (VI, 106) while the Spartans, starting as soon as it was full moon, reached
Attica in three days, just after the battle (VI, 120), or, as Plato says (Laws, I I I ,
p. 698 E ; Menexenos, p. 240 C), one day after the battle. And Herodotus, though
accused by Plutarch of deception, is approved by modern authorities. 5 There
1 Busolt (Gr. Gesch. I I , pp. 702, n. 2, 704 note; followed by Hauvette, Hérodote, p. 414) calculates
that Sept. 27 or 28, when the moon did not rise until 12:58 or 1:55 A.M. respectively, would have
fitted the accounts of the preliminaries of the battle. As Boedromion began 59 days after July 7 and
so on Sept. 4, Busolt's dates would be equivalent to Boed. 24-25. Others have confined Polyaenus and
Plutarch too literally to Boed. 20 or 21 (Bôckh, Mondcyklen, p. 73; Beloch, Gr. Gesch. I I 2, p. 48;
Macan, Herodotus VII-IX, p. 293) equating this with Sept. 20, 23, or 24 (Munro, Cambridge Ancient
History, IV, pp. 304, 313).
2 Clinton (Fasti Hellenici, I I , p. 32) regarded Moun. 16 as an error; Beloeh {Gr. Gesch.'1 I I -2, p. 48)
suggests that Moun. 16 was the day of celebration, a festival of Artemis, and that Plutarch confused
the two dates. Munro suggests that both dates are right, that Moun. 16 is Old Style and Boed. 20 is
New Style (Cambridge Ancient History, IV, p. 313), and that the new year according to the old calendar
occurred on Dec. 5, 480 B.C. He therefore deduces that the battle of Marathon, on Boed. 16, was on
Thargelion 11 in Old Style (ibid. p. 245). But we have no authority for such a tremendous divergence;
and, in detail, the date Boed. 20 is not explicit; Plutarch apparently implied Boed. 20-29. I t seems
preferable to regard Moun. 16 as a simple error.
3 With the new year on Aug. 5, Hekatombaion would have been a hollow month, with the Pana-

thenaia on Hek. 27.

4 One of the ordinary years would have contained 355 instead of 354 days (if this calculation is cor-

rect), thus fitting octaeteris no. 3 as described in my Archons of Athens, p. 305. In default of evidence
as to the location of this augmented year, I have tentatively placed it in the middle of the series
(485/4).
s Clinton (Fasti Hellenici, I I , p. 28) merely assumed that the civil calendar (Plutarch) disagreed with
the lunar month (Herodotus). I t is generally argued, however, that Plutarch's date Boed. 6 is merely
that of the celebration and that the battle was on the second day after the full moon of the preceding
month Metageitnion, and so on Met. 17 (Bockh, Kl. Schriften, IV, pp. 85-97; Jb. tel. Phil. Suppl. I I ,
 T H E DATE OF THE OLDER PARTHENON 445

seems to be a way of explaining the discrepancy. For Boedromion 6 may well

have been the day on which, Miltiades having the turn as chief general, the decree
was piissed to engage the Persians in battle at M a r a t h o n and a corresponding vow
made to Artemis Agrotera, the patroness of t h a t day. Pheidippides would likewise
have set out on t h a t day for Sparta, arriving on the 8th. After nine days of inaction
at M a r a t h o n , the battle took place when Miltiades again received his turn in rota-
tion on t h e 16th. T h e Spartans, starting on the 14th, arrived on the 17th. 1 T h e
festival was thereafter celebrated on the day of Artemis Agrotera, Boed. 6, as
Plutarch expressly states (Glor. Ath. 7); and he confused it with t h e actual date of
t h e battle, ten days later. 2 T h e date of the actual battle, Boed. 16, would have
been Oct. 11, 490 B.C.3 T h e resulting calendar of the period of the Persian Wars
is as follows:

Hek. 1 Days Events

490 July 29 O 354 Marathon, Boed. 1 6 = O c t . 11
489 July 17 I 384
488 Aug. 5 O 354 Parthenon, Hek. 2 7 = Aug. 31
487 July 25 0 354
480 July 14 I 384
485 Aug. 1 0 355 (?)
484 July 22 0 354
483 July 11 I 384
482 July 30 0 354
481 July 18 0 354
480 July 7 I 384 Salamis, Boed. 24/25 = Sept. 27/
479 July 26 0 354 Plataea, Boed. 3 / 4 = Sept. 25/26
pp. 64-73). H a u v e t t e (Herodote, p. 270) preferred Met. IB for the reason discussed below, and was
followed by Meyer (Gesch. Alt. I l l , p. 334). The return to Boedromion, and the placing of the battle
on the 10th of that month (ten days after rather than twenty days before Plutarch), was sponsored by
Grundy (Great Persian War, pp. 173, 182-183, 193) and Munro (Cambridge Ancient History, IV, pp.
243, 245).
1
Assuming t h a t the full moon could not have appeared until the 15th, it has been generally assumed
t h a t the Spartans arrived on the 18th (Clinton, Fasti Hellenici, II, p. 28; Bockh, Jb. kl. Phil. Suppl. I,
1855, pp. 64-73; etc.). Beloeh (Gr. Gesch.1 I I 1 , p. 22, n. 2; I I 2, pp. 56-57), furthermore, entirely dis-
cards the tradition on the ground that it would have taken the Spartans at least five, and probably
seven days to cover the distance, arriving on the 22nd. But with Boed. 1 = Sept. 26, the full moon of
Oct. 8, 4:15 P.M. would have occurred on Boed. 13, if the Spartan calendar then coincided with the
Athenian; so that the Spartans might in any case have started on Boed. 14. And the fact that the
distance of 1160 stadia =222 km. (Pliny, H.N. VII, 84; Isocrates, Paneg. 87; cf. Busolt, Gr. Gesch? II,
p. 580, n. 1) was traversed in the time indicated seems possible for the special small force of 2000 men.
2
This follows the explanation given by Hauvette, who applied it, however, to Metageitnion of 490
B.C.; his assumption t h a t the vow was made on Met. 6 (ten days before Miltiades resumed command
and delivered battle), and t h a t the celebration was held exactly a month after the vow and so on Boed.
6, lays too much stress on the senseless observance of an exact month for the interval. Grundy applied
the same explanations to Boedromion of 490, and Munro to Boedromion of 491 B.C.
3
The date Metageitnion 17 was equated by Bockh with Sept. 12, that of M e t . 16 with Sept. 10 by
Meyer. Busolt argued that this was too late for a campaign begun in the spring and so preferred
Aug. 12/13 (Gr. Gesch.2II, pp. 580 n. 3, 596, n. 4). Beloch, abandoning most of the literary tradition,
assumes that the battle took place about four days after the full moon of Aug. 10 or Sept. 9 (Gr. Gesch. 2
I I 2, pp. 56-57). Munro equates Boed. 16= Sept. 21, 491 B.C. (which he assumes to be a reformed
Metonic date, and t h a t the contemporary calendar date would have been Thargelion 11); he argues
that the full moon of Oct. 8, 490 B.C., would also have been too late in the season; and that this was one
of the reasons for his preference for 491 B.C. But, as H a u v e t t e points out, Busolt's citations from
446 WILLIAM BELL DINSMOOE

In order to confirm the calendar arrangement for this decade, it would be desirable
to connect with other known dates both before and after it. Unfortunately, we
have no other definite dates until we reach July 15, 432 B.C., the initial day of the
Metonic calendar; 1 and we cannot be certain as to the precise relationship between
the Metonic cycle and the preceding octaeteris. I t has been noted that between the
new year of 480 B.C. (July 7) and that of 432 B.C. (July 15) would have elapsed 594
lunar months and hence exactly six octaeterides of 99 months each. 2 But we have
no assurance that 432 B.C. marked the end of an octaeteris; nor was Meton more
likely to have awaited such an event than did Kallippos in 330 B.C. The intercala-
tions within the octaeterides were undoubtedly irregular, so t h a t the interval of 594
months may be regarded as a mere coincidence, of no more significance than if there
had been 593 or 595 months. 3 A more reasonable system for the beginnings of
octaeterides would have been their coincidence with Panathenaic years. If, for
instance, we take 486 B.C. as such an initial year, the connection with the Metonic
new year in 432 B.C. would be as follows: 4
New Moon Hek. 1 Length of Octaeteris
486 July 13, 5:30 A.M. July 14 2923 days
478 " 14, 0:45 P . M . " 15 " "

470 " 15, 8:15 P . M . " 16 " "

462 " 17,5:00 A.M. " 17 " "

454 " 18,5:15 P.M. " 18

440 " 20, 8:15 A . M . " 19 " "

438 " 22, 3:00 A . M . " 20 2187 "

I t is apparent, in this scheme, that the initial day of the octaeteris was gradually
falling behind the moon, a defect which must have been one of the causes of Meton's
reform. I t is also apparent that the unfinished octaeteris beginning in 438 B.C.
would have comprised 2187 days or 74 lunar months, corresponding to two inter-
calary and four ordinary years, with two or three extra days inserted to correct the
lunar calendar. 5 While this tabulation may in itself seem very hypothetical, its
probability is greatly strengthened when we carry it back in the opposite direction.
For exactly ten octaeterides or 990 lunar months earlier than 486 B.C. bring us to the
notable year 566 B.C., when the Panathenaic festival was reorganized on a quad-
rennial basis (Pherecydes 6 and Eusebius')• And in this notable year, furthermore,
Herodotus (VI, 43,48, 95) do not prove that the campaign was begun very early; and the same observa-
tion would apply to Munro's objections to the October date. There seems to be no necessity, there-
fore, for adopting the Bockh-Hauvette expedient of assigning the battle to the preceding month,
Metageitnion (thus dating it Sept. 11/12).
1
For the establishment of this date, see my Archons of Athens, pp. 311-317.
2
West, A.J.A. 1934, p. 7.
s
I n other words, there is no need to regard 488 and 480 B.C. as the beginning and end of an octaeteris,
for which the interval of 98 months, according to my scheme, would in any case have been insufficient.
1
The dates of the astronomical new moons arc computed from Ginzel's tables as noted above.
5
In other words, it is possible to adopt the conclusion reached by West {A.J.A. 1934, p. 9), that
Meton did not alter the arrangement of the months, and that 433/2 was an ordinary year, a question on
which I had previously been unable to reach a decision (Archons of Athens, p. 317).
6
Fherecydcs, in Didymus, as quoted by Marcellinus, Thucydides, 3 (cf. Muller, Frg. Hist. Or. I, p.
73; Jacoby, Frg. Gr. Hist. I, pp. 59-60): " i n the archonship of Hippokleides."
7
Eusebius, trans. Hicronymus (ed. Helm, 1913, p. 102; Fotheringham, 1923, p. 181), Abr. 1451 = 01.
53, 3; Armenian trans, (ed. Karst, 1911, p. 188), Abr. 1451 = 0 1 . 53, 4 (the Olympic date being as usual
one year too late); cf. Georgius Syncellus (ed. Dindorf, 1829, p. 454, 8).
 T H E DATE OF T H E OLDER PARTHENON 447

the initial day of the octaeteris, the day of the visible new moon, would coincide with
the summer solstice, an appropriate occasion for beginning this luni-solar system.
T h e exact distribution of the individual days must remain slightly uncertain; I have
assumed that the series began with the more normal octaeteris of 2922 days, and
that this by observation was immediately found to be inadequate and was replaced
successively by those of 2923 and 2924 days, 1 eventually returning to 2923 days in
the time of Kleisthenes; but there are other possibilities.

New Moon Hek. 1 Length of Octaeteris

566 June 27, 11:15 P.M. June 29 2922 days
558 " 29, 6:30 A.M. " 29 2923 "
550 " 30, 1:15 P.M. " 30 2924 "
542 July 1 , 1 0 : 4 5 P.M. July 2
534 " 3 , 1 0 : 4 5 A.M. " 4
526 " 5, 2:00 A.M. " 6
518 " 6, 9:00 P.M. " 8
510 " 8, 3:30 P.M. " 10
502 " 10, 7:45 A.M. " 12 2923 "
494 " 11, 8:30 P.M. " 13
486 " 13, 5:30 A.M. " 14

In view of this agreement of the resulting calendar with all the required condi-
tions, we may conclude that the axis of the temple, and the perimeter of the founda-
tion, were staked out at sunrise on Aug. 31, 488 B.C.2
* * * * *

According to this chronology, it was nearly twenty-three months after the victory
of Marathon that the great temple intended to commemorate it was laid out. T h e
interval, however, is easily explained. As Dorpfeld pointed out, 3 Aristeides, the
archon of 4 8 9 / 8 B . C . , was éxi/ieXijn)? TS>V B-QIXOAÍOIV IRPOOÓDWV (Plutarch, Aristeides, 4, 3),
and as such is most fittingly to be regarded as the sponsor of the marble pre-Persian
Parthenon. B u t the design could not have been executed, the finances diverted
toward this purpose, and preliminary operations carried out at the quarries, in the
few days between his accession to office and the festival of 489 B.C. T h e actual
beginning of the work was postponed, therefore, until the following Panathenaic
festival of Aug. 31, 488 B.C. I t is upon Aristeides, nevertheless, that we are to
look, not merely as the continuator of an earlier project of Kleisthenes, but as the
originator of the plan for replacing the old poros temple of the Peisistratidae by a
monumental structure crowning the south flank of the citadel.
This determination of the date of the great temple destroyed during the burning
of the Acropolis on Sept. 2 5 / 2 6 , 480 B.C., two days before the battle of Salamis, is
of more importance than the mere dating of a single building. I t marks the be-
ginning of an epoch, that of construction in Pentelic marble, and was apparently the
project for which the Pentelic quarries were first worked on a large scale. 4 We now
1 On the various lengths of octaeterides, see Archons of Athens, pp. 303-305.
2 On the Panathenaic procession starting at sunrise, see I.G.2 I I , 334; Himerius, I I I , 16.
3 Dorpfeld, Ath. Mitt. 1902, p. 411.

' This forms an additional reason for dating Athenian buildings in island marble, such as the treasury
and stoa at Delphi, before rather than after 490 B.C.
448 WILLIAM BELL DINSMOOR

see that it was part of a grandiose scheme for the reconstruction of the Acropolis, a
scheme of which the technically similar old Propylon formed another unit. Thus
we may clarify also the problem of the Old Propylon, which is now likewise fixed in
the period 4 8 9 - 4 8 0 B.C, and probably, in view of a slightly less developed technique,
in the first half of this decade, the work of the same group of stonecutters who pro-
duced the Marathon base.1 And the first two lines of the latter, in turn, are ap-
parently by the same scribe who carved the "Hekatompedon inscription" (I.G. 21,
3/4) ,2 at the time of the building of the Propylon,3 so that we may retain the name of
the archon as $[IXOKP<1TO?] ( 4 8 , 5 / 4 B.C.) instead of some unknown individual before
506 B.C. as has been recently suggested.4 And we now see that it was probably to
this great building program of Aristeides that Demosthenes referred (XXII, 13),
when he somewhat ambiguously stated that the Parthenon and Propvlaea were built
from the spoils of Marathon. Furthermore, the new system of dimensions and pro-
portions, and even the very materials, prepared for this Older Parthenon, were in-
corporated in, and thus influenced the design of the masterpiece by Iktinos which we
see to-day, completed and dedicated at the Panathenaia of 438 B.C., exactly on the
fiftieth anniversary of the laying out of the basement.
WILLIAM B E L L DINSMOOR
COLUMBIA UNIVERSITY
New York
1
I.G.' I, 763; Hesperia. I I , 1933, p. 480.
2
Wilhelm, Ath. Mitt. 1898, p. 489; Miss Spaulding, A.J.A. 1906, p. 404; Oliver, Hesperia, 1933, p.
486.
3
T h i s discussion of t h e d a t e of t h e Propylon will be amplified in m y m o n o g r a p h on t h e P r o p y l a e a .
4
Luria, Hermes, 1927, pp. 270-272; Dorpfeld, Phil. Woeh. 1929, 1247.