American Philological Association

An Unfinished Inscription, IG II2 2362

Author(s): W. Kendrick Pritchett
Source: Transactions and Proceedings of the American Philological Association, Vol. 85 (1954),
pp. 159-167
Published by: The Johns Hopkins University Press
Stable URL: http://www.jstor.org/stable/283473 .
Accessed: 02/01/2014 18:37

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

American Philological Association and The Johns Hopkins University Press are collaborating with JSTOR to
digitize, preserve and extend access to Transactions and Proceedings of the American Philological
Association.

http://www.jstor.org

This content downloaded from 143.107.252.127 on Thu, 2 Jan 2014 18:37:40 PM

All use subject to JSTOR Terms and Conditions
Vol. lxxxv] An UnfinishedInscription,IG JJ2 2362 159

X.-An UnfinishedInscription,IG II2 2362

W. KENDRICK PRITCHETT
UNIVERSITY OF CALIFORNIA

The only extant catalogue of Attic demes was foundon March

20, 1840,1on the Acropolis west of the Parthenon. This unique
document,whichcontainsa list of demes arrangedaccordingto the
officialorder of theirrespectivetribes,was immediatelypublished
by Pittakys in the 'Eq 71wEp1s 'ApXatoXoYtK' of that year (No. 410).
A few years later, the document was completelyrestudiedby L.
Ross, who used it as the point of departureforhis book Die Demen
vonAttika (Halle 1846). Rangabe in 1855 reproducedthe text of
Ross withouta new examinationof the stone (AntiquitesHelleniques
No. 1258). In 1878, U. Koehler publishedas IG II 991 a new text
based on his own study of the stone. Later, R. Loeper (AM 17
[1892] 348, 364, 374, 390) reportedhis readingsfromthe stone of
certain lines with which he was concerned,but he did not give a
text of the document nor did he discuss the more difficultlines,
those containingthe names of the demes of Ptolemais and Hippo-
thontis. More recently(1931), the inscriptionhas been published
by Kirchner(IG 112 2362), who reportsin his criticalapparatus the
readingsof Lolling. Other scholars have made suggestionsabout
the restorationsin the text,2but so far as I can tell fromtheir
writings,these suggestionswere not based on a freshstudy of the
stoneitself. Kirchnerin theeditiominorindicatedthathe possessed
a squeeze, and the presentwriter,when he reportedon the inscrip-
tion in AJP 57 (1942) 423-25, likewise worked froma squeeze,
one in the collectionof the InstituteforAdvanced Study. So faras
one can tell, then,fromthe publishedliterature,only fourscholars
have reporteda complete text based on first-handexaminationof
the stone itself:Pittakys,Ross, Koehler,and Lolling.
We are concernedin this article firstwith the text of Column
IV, and secondly,with the demes of the tribe Ptolemais engraved
in Column III.3
1 The date is that reported by Pittakys. Some scholars (cf. M. N. Tod, ABSA
9[1902/3] 173) apparently translated Pittakys' dates in terms of the modern calendar.
2 In particular, Schebelew in
zrfqpavos in Honor of T. Sokolov (Petersbourg 1895;
in Russian).
I It may be noted here that there was anathyrosis on the left side of the block.

This content downloaded from 143.107.252.127 on Thu, 2 Jan 2014 18:37:40 PM

All use subject to JSTOR Terms and Conditions
160 W. KendrickPritchett [1954

Readings have been reportedforonly two lines of Column IV,

lines 67 and 68 respectively,as numberedin the editiominor. The
followingtable lists these readings.
Line 67
Pittakys nihil
Ross KL.I
Koehler nihil
Lolling K---
Line 68
Pittakys nihil
Ross E.A--
Koehler E - --
Lolling E A/
Several commentsmay be made on this table. Koehler and
Lolling,both noted epigraphists,whose highrecordof accuracy has
been proved in numberlesscases,4 saw only one and two letters
respectivelyin this column. The fact that Pittakys saw nothing
is not too significant,for thoughwe hold his work in the highest
underwhichhe labored,we have
esteem,and realize the difficulties
to admit that he has failed us on other occasions, and Rangabe
indeed thought it necessary to re-examineliterally hundreds of
inscriptionswhichPittakyshad published. The explorerRoss saw
the most. He reportedall, or parts, of fiveletters. Koehler,who
could see only one letter,an epsilon at the beginningof line 68,
was apparentlytroubledby the fact that Ross had seen traces of
fourothers. He wrotein his commentarythat the inscribedsurface
must have sufferedrecent damage, for letterswhich had formerly
been read had vanished.
The inscribedsurface of the slab is today solid, and squeezes
can be made withoutdamage to the letters. In some spots several
layers of the marble have been lost. But this disintegration,or
exfoliation,of the surface,broughtabout by contactwith moisture,
musthave ceased when the marblewas no longerexposed to water.5
Today the weatheredsurfaceis not at all flaky,nor do crystalsfall
4 For the present writer's opinion of the unusual accuracy of the readings of
Lolling, see W. K. Pritchett and 0. Neugebauer, The Calendars of Athens (Cambridge
1947) viii.
5 For an interesting study of the causes of disintegration in marble (including the
ancient). see A. Kieslinuer, ZerstdrunRenan Steinbauten (Leipzig 1932).

This content downloaded from 143.107.252.127 on Thu, 2 Jan 2014 18:37:40 PM

All use subject to JSTOR Terms and Conditions
 tA'

( 'Ii~ 4

A,t

-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~-

41 L- L C

This content downloaded from 143.107.252.127 on Thu, 2 Jan 2014 18:37:40 PM

All use subject to JSTOR Terms and Conditions
PL. II. Fig. B. Latex Squeeze. Detail of lowerpart of "Column IV."

PL. II. Fig.Ces

PL. II. Fig. C. Latex Squeeze. Part of lines 49-56.

This content downloaded from 143.107.252.127 on Thu, 2 Jan 2014 18:37:40 PM

All use subject to JSTOR Terms and Conditions
Vol. lxxxv] An UnfinishedInscription,IG IJ2 2362 161

offto the touch, as they do, forexample, when marble is faultily

cut at the quarry.6
Judgingfromthe physicalconditionof the stone, then,thereis
no reason to believe that any of the surfacehas peeled offsince the
days of Ross. Our faithin his reportof this problematiccolumnis
furthershakenby the fact that he made manyerrorsin his readings
of the othercolumns- more than ten, indeed,withinthe limitsof
this one inscription. For example, in line 27 he reported 21A
where Koehler saw, and the stone clearly reads, BAA; in line 38,
OATM,where we now know the text is AAIM; and HP in line 52,
which Koehler and later editors rightlycorrectedto HE. More
importantly,in every place where Ross saw a letterwhichwe now
question, we believe we can discern the traces which misled him.
Thus, in the thirdletterspace of line 68 Ross read an alpha. This
appears today as nothingmore than the traces of two diagonal
scratchesforminga very obtuse angle, whichcertainlydo not con-
stitute evidence for the reading of an alpha. Taking all of these
considerationstogether,we must conclude that Ross's observations
were often unreliable,and that Koehler need not have made his
apologetic statementthat later damage to the stone had obscured
letterscorrectlyseen by Ross, but whichhe himselfcould not make
out. In all probabilitythe letterswere never there.
The presentwriter,when he came to restudyat firsthand the
inscription,which is now in the Epigraphical Museum, was struck
by the fact that the only letterin Column IV that could be looked
upon as at all likelywas the epsilon at the beginningof line 68.7
This, it should be emphasized,was the only letterseen by Koehler.
The evidenceforit can be seen in the photographof a latex squeeze
in Plate II, fig.B. This letterconsistsof a verticaland lowerand
upper horizontal hastas; there is no trace of the middle hasta.
This writercame to doubt, however,whetherthese strokesactually
did forma real letter. If theydo, several difficulties
arise.
In the firstplace, therewould be a discrepancyin the width of
the column. Columns I and II both measure 0.215 m. in width;
if we accept the so-called epsilon,Column III would be very much
6 For a study of the general structure of marble as related to
epigraphical matters,
see N. Herz and W. K. Pritchett, AJA 57 (1953) 71-83.
7 There is a vertical stroke, or scratch, in the space
below, but I do not believe
that it constitutes evidence for a kappa or any other ancient letter.

This content downloaded from 143.107.252.127 on Thu, 2 Jan 2014 18:37:40 PM

All use subject to JSTOR Terms and Conditions
162 W. KendrickPritchett [1954

narrower,for the distance fromthe firstletter of Column III to

the "epsilon" is only 0.181 m.8
Furthermore,the "epsilon" is not in horizontalalignmentwith
the names in the other threecolumns. With the use of a straight
edge, it can be determinedthat lines of Columns I, II, and III
were in horizontalalignmentat least as faras the extentof Column
III (line 65). Lines 50-65 of Column III are exactly opposite the
correspondinglines of Column II and Column I, 27-44, 4-21, as
numberedby Kirchner. One would expect the "epsilon" to be on
the same line as that which includes lines 16, 39 and 60, or 17, 40
and 61 of the firstthreecolumnsrespectively. Instead, it is oppo-
site the interspacebetweenlines 60 and 61 of Column III.
I have returnedto the stone again and again in hopes of finding
otherlettersin Column IV. Repeated examination,however,has
not only failed,but the surprisingconclusionhas been forcedupon
me that there are in fact no other letters,nor were there ever.
Much of the originalsurfacestillsurvivesintact,uninscribed. Our
textwas nevercompleted.
Realizing the importanceof this observation,I soughtthe inde-
pendentjudgmentof otherscholars; so two epigraphists,Professor
Eugene Vanderpool and Dr. Markellos Mitsos, and one geologist,
Dr. N. Herz of the United States Geological Survey, kindly ex-
amined the stone with me. All concur in my opinion about the
physical conditionof Column IV. On parts of the stone we can
startat Column II and trace across to the rightmostedge and have
a continuityof the originalsurface. Where this originalsurfaceis
preservedin "Column IV" - and this area is ratherextensive-
thereare no letterswheretherecertainlywould be if a columnhad
been entered. The stonecutterleftthiscolumntotallyblank. This
polisheduninscribedsurfacecan be seen in the photographon Plate
I, fig.A.
A plausible explanationforcessationof workon the inscription,
which mirroredthe extinctionof the two Macedonian tribes,may
be foundin the suppositionthat the tribalorganizationset forthin
the list was in forcefora very short time,and this a time of great
turbulencein the lifeof the Athenianstate.
The document can be dated fairlyprecisely. Since Grotefend
(Zeitschriftfur Altertumswissenschaft 1857), it has been recognized
8 The fact that the stonecutter used abbreviations in Column I might lead us to
infer that he planned the spacing of his columns in advance.

This content downloaded from 143.107.252.127 on Thu, 2 Jan 2014 18:37:40 PM

All use subject to JSTOR Terms and Conditions
Vol. lxxxv] An UnfinishedInscription,IG IJ2 2362 163

that the tribal arrangementof demes reflectedin our inscription

was later than the formationof the tribe Ptolemais (223 B.c.)9 but
beforethat of the tribe Attalis. The order and the deme assign-
ment showed,too, that the tribesAntigonisand Demetrias,which
had officiallyoccupied the firstand second positions,wereno longer
in existence. There is no literaryevidence which tells us when
Antigonisand Demetrias, the two Macedonian tribes,were abol-
ished,but it is reasonableto seek a motivein some seriouspolitical
circumstance,for as A. H. McDonald has pointed out, "the two
tribes . . . had survived the revolt in 289/8, the Chremonidaean
War in 266, the secession in 229 and the alliance with Egypt in
224."10 For nearly thirtyyears preceding 201 B.C. Athens had
maintained a policy of strict neutrality. To judge from extant
evidence, the extinctionof the Macedonian tribes,which reflected
an abrupt change of policy, mightbe sought in any one of three
causes: first,theAtheniandeclarationofwar against Philip; second,
the Macedonian participation in the Acarnanian devastation of
Attica which followeda sacrilege at the Eleusinian mysteriesof
Boedromion,201/0 (roughly,September201 B.C.) ;11 and third,the
revelationof the secret Syro-Macedonian pact, which has been so
ably discussed by A. H. McDonald and F. W. Walbank.12
The firstpossibilitymust be rejected because Athens did not
declare war until afterAttalus had enteredthe city,and it was on
his entrance that the tribe of Attalis was created,13whereas the
Macedonian tribes,on the evidence of our unfinishedinscription,
were abolished beforethe creationof Attalis.
The second possibilitywas favored by W. S. Ferguson who
argued that the Acarnanian-Macedonianattack should be dated in
the autumnof 201 B.C. This date had the advantage,in Ferguson's
opinion,of allowingsome littletime to elapse betweenthe abroga-
tion of Antigonis and Demetrias and the creation of Attalis.'4
A. H. McDonald and F. W. Walbank, on the otherhand, conclude
that the Syro-Macedonian pact, directed against Athens' close
9 For the date, see W. K. Pritchett, The Five Attic Tribes after Kleisthenes (Balti-
more 1943) 13-23.
? JRS 2 7 (1937) 191.
"Livy 31.14.6-18.
12JRS 2 7 (1937) 180-207.
13 Polybius 16.25; Livy 31.15.
14 Quoted with slight modification from W. S. Ferguson,
Athenian Tribal Cycles in
the Hellenistic Age (Cambridge 1932) 141, note 1.

This content downloaded from 143.107.252.127 on Thu, 2 Jan 2014 18:37:40 PM

All use subject to JSTOR Terms and Conditions
164 W. KendrickPritchett [1954

friendEgypt, became known in the summerof 201 B.C., and they

believe that the news of the pact resultedin the immediateabolition
of the two tribes.15 They state that subsequentlythe incidentat
the Eleusinian mysteriestook place, and the Acarnanian invasion
is put back until the springof 200.16 This date for the abolition
of the Macedonian tribesseems to be based on the inferencethat
some considerabletime must be allowed for the period of eleven
tribes."7
Now, however,that the incompletecharacterof our inscription
is manifest,we need no longerfollowFerguson,or McDonald and
Walbank, in postulatingany lengthyperiod for the existence of
eleven tribes. The earlier interpretationof Holleaux,'8 to which
Fergusontook exception,'9has the advantage of makingthe interval
a very shortone. The invasion took place early in 200 B.C., with
the immediateabolition of the Macedonian tribes; the creation of
Attalis ca. May, 200 B.C.20 An overt attack, not an unfriendly
pact, aroused the populace to the drasticstep involvedin the aboli-
tion of the two A1lacedoniantribes.
* * *

We turnnow to the second part of our paper,whichis concerned

with the partially preserved registerof the demes of the tribe
Ptolemais.2' The names of these demes were cut at the end of
Column II (which is now lost) and the beginningof Column III.
I am concernedhere with these latter lines,49-56 as numberedby
Kirchner. Eight lines still remain in which traces of lettershave
15 JRS 27 (1937) 191; Walbank, Philip V of Macedon (Cambridge 1940) 124-25,
340.
16 See Walbank, op. cit. 129. That the Acarnanian attack with the co-operation
of Macedonian ships took place in the spring of 200, not the autumn of 201, can be
inferred from Polybius' statement (16.26.9) that the Athenian crews, first captured
by the Macedonians, then rescued by the Rhodians, were still with their ships when
returned by the Rhodians. The Rhodians reached Athens while Attalus was receiving
his new honors.
17 Cf. A. H. McDonald, JRS 27 (1937) 191, note 70.
18 CAH 8.161-63.
19Op. cit. (above, note 14) 141, note 1.
20 CAH 8.162.
21 The "Egyptian" tribe comprised more demes than any other. Twenty-four
were listed by me in The Five Attic Tribes afterKleisthenes 29-32, and it is by no means
certain that the list is complete. On the other hand, some of the evidence is tenuous,
and much of it depends on inscriptions dated as late as the second century after Christ.
It should be emphasized that in the present study no new re-examination of the
evidence for tribal affiliation has been attempted; restorations have been presented
on the basis of the list published in The Five Attic Tribes.

This content downloaded from 143.107.252.127 on Thu, 2 Jan 2014 18:37:40 PM

All use subject to JSTOR Terms and Conditions
 Vol. lxxxv] An UnfinishedInscription,IG II2 2362 165

been reported. The text of threeof these lines seems certainand

does not need restudy. In lines 54 and 55, the names of thedemes
'TwrwpEaand EvvonL6bat have been read withoutdifficulty since the
editioprincepsof Pittakys. In line 49, the firsttwo lettersare (3T,
whichof the knowndemes of Ptolemaiscan be completedonlywith
the name evpyvL6at.
For the remainingfive lines, 50-53 and 56, there is conflicting
testimony. I offerbelow a criticalapparatus, givingfirstthe vari-
ant readings,and followingthis with a discussionof what can be
seen on the stone today. With the help of the photographof a
translucentlatex squeeze, Plate II, fig.C, the reader has a means
of checkingthis discussion.
Line 50
Pittakys nihil
Ross nihil
Koehler LA/
Lolling EAQ
Pritchett(1942; frompaper squeeze)22 KA
Pritchett(1952) .AU
In the firstletter space, there is no sure stroke of an original
letter.23 In the second letterspace, a lambda seemsmostprobable,
a delta possible,an alpha improbable. In the thirdletterspace an
omega is certain. Of the known demes of Ptolemais, KXwIrlbaLis
the only possibility.24
Line 51
Pittakys nihil
Ross nihil
Koehler A
Lolling-Kirchner 2A
Pritchett(1952) A
The sigma in the firstletterspace was printedby Koehler with
brokenlines in his majuscule text to indicate that the traces were
22 Op. cit. (above, note 9) 24-25.
23 The surface in this letter space is deeply exfoliated, and the indentation is in
the form of a square, which might have led some scholars to consider the letter an
epsilon. Kirchner in IG II2 2363 read 'E&o -- -. He offered no commentary, but
apparently assumed that this was a deme otherwise unattested. It may be mentioned
here that this list does contain in line 30 the name of one deme which appears in neither
the Lexicographers nor any other inscription: rpaAs of Pandionis. I see on the stone
no trace of the broken kappa which I had read in 1942 on a paper squeeze.
24 This solution has been suggested by Koehler, Schebelew (above, note 2) 30,
F. 0. Bates, The Five Post-Kleisthenean Tribes (Ithaca 1898) 44, and Sch6ffer,RE s.v.
Ai/Aot,73-74.

This content downloaded from 143.107.252.127 on Thu, 2 Jan 2014 18:37:40 PM

All use subject to JSTOR Terms and Conditions
166 W. KendrickPritchett [1954

faint. The letterwas enclosed in brackets in his minuscule text.

The only restorationwhichhas been offeredthat has taken account
of this sigma has been that of Schofferin RE s.v. A-,4ot,99, where
Sa[lamis] was suggested. Afterexaminingthe stone, I reject the
sigma. I believe the earlier editors were misled by the deep,
triangular-shaped scar in the lowerpart of the letterspace. Above
this scar, thereis a strokewhich mightbe mistakenforthe upper
of the fourhastas of a sigma. But it is long and curving,unlike
the upper strokeof other sigmas, and, I believe, fortuitous. The
second letteris an alpha or a lambda; it cannot be a delta because
the stone is ratherwell preservedat the base of the letterwherethe
horizontalstrokeshould fall. Accordingly,I would restorethe line
as VEls].25
Line 52
Pittakys nihil
Ross HP
Koehler TIE
Lolling HE
Pritchett(1952) HE ---
The firsttwo lettersare quite clear. Ross's readingof a rho in
the second letterspace is incorrect. Throughoutthe historyof the
tribe Ptolemais there were three demes whose names began with
IIE -: Pentele, Perrhidai,and Petalidai. As Schebelew long ago
pointedout, all are possiblerestorationsforthisline,althoughother
editors have consistentlyrestoredPerrhidai. Petalidai is not at-
tested as a deme until Roman times.26 Perrhidai is otherwise
knownonlyfromthe Lexicographers. On the evidenceof the stone
alone no decisionabout the restorationcan be made.
Line 53
Pittakys nihil
Ross II . . .221
Koehler Traces in fifthletterspace
Lolling .... P
Pritchett(1942; froma squeeze) ... Pi/
Pritchett(1952) ... P
This line has never been correctlyread. The most probable
letterhas seemed to me and to those who examinedthe stone with
25 This restoration has already been proposed by Schebelew (above, note 2) 31.
26 Pritchett (above, note 9) 26-27, 31.
27 Renorted by Rangab6 (Antiquites Helleniques No. 1258) as II .... 1.

This content downloaded from 143.107.252.127 on Thu, 2 Jan 2014 18:37:40 PM

All use subject to JSTOR Terms and Conditions
Vol. lxxxv] An UnfinishedInscription,IG IJ2 2362 167

me to be rho in the fourthletter space above and slightlyto the

leftof the rho of the line below.28 It can be seen in our photograph
of a translucentlatex squeeze (Plate II, fig.C). In the next letter
space there is the upper part of a vertical stroke which may be
original. No other traces in this line can be identifiedas original.
The only deme now assigned to Ptolemais which may be restored
in the line is ['JKa]p[a].29
Line 56
Pittakys . XX
Ross .AA.N. .2
Koehler . I . .. N . .2
Lolling AI 1i<
Pritchett nihil
The conditionof the surfaceof the stone at this line gives little
clue as to the original text. The only stroke which is possibly
originalis a verticalin the eighthletterspace. But one cannot be
sure whetherit is an iota or part of anotherletter. The transcrip-
tions of the foureditors who have examined the stone are not in
complete accord concerningany letter. Kirchner,accepting Lol-
ling's traces, restoresthe name as A[7yXt]E-s.The presentwriter
believes thereis not sufficientevidence to give any readingforthe
line.30
In summary,I offerbelow a text of lines 49 ff. The fourdemes
the location of which is known come fromthe inland trittysof
Ptolemais.
E)v[p,ywP15at]
50 [K]Xc[rlbat]
[J?]X[uEZs]
I1E[ - - - ]

['JKa]pl[a]
T7rc.wp[a]
55 EhvoarlbaL

28 Lolling, in
my opinion the most trustworthy editor of those who have reported
on the stone, read the rho, but in the fifthletter space.
29 It
must be noted, however, that aside from this inscription, the association of
the deme Ikaria with the phyle Ptolemais rests on the evidence of IG II2 2442, line 5.
M. Mitsos (BCH 73 [1949] 356, line 23) has recently corrected Kirchner's reading
'IKa(pLe6s) to 'EKa(XjOeV) in IG II2 2107. This reading had hitherto been regarded
as the chief evidence for the affiliation of Ikaria with Ptolemais.
30 Below line 56 Pittakys reported another line which began with the letters ipo;

this could be completed only as lpO['lavara]. No other editor has seen these letters.

This content downloaded from 143.107.252.127 on Thu, 2 Jan 2014 18:37:40 PM

All use subject to JSTOR Terms and Conditions