HESPERIA 75 (2006) 16 I3 1055 B AND THE Pages 83?119 BOUNDARY OF MELITE AND KOLLYTOS

ABSTRACT

on Two rupestral horoi found the Hill of the Nymphs inAthens, IG I31055 A : are a (?opo? ?io? [retrograde with reversed sigmas]) and B (#opo?), not as single boustrophedon text usually edited. Investigation of the possibility a a that B marked deme boundary, prefaced by discussion of deme formation and territoriality, yields evidence that the ancient street that passed south of on horos B its route from the Agora to the saddle between the Hill of the Nymphs and the Pnyx divided the urban demes ofMelite and Kollytos. This was argument challenges the traditional view that the Pnyx inMelite. The an study concludes with approximation of the full extent ofMelite.

IGV 1055 B

In 1835 Kyriakos S. Pittakis published the first edition of two Archaic rupes : tral horoi (IGV 1055 A: ?opo? ?io? [retrograde with reversed sigmas] and near IG I3 1055 B:h?poc?) that he had found four years earlier the midpoint western of the south edge of the northeast spur of the Hill of the Nymphs in Athens (Figs. 1-3).l On the plateau of this bedrock spur are rock-cuttings as noted by topographers and archaeologists of the 19th century remnants of

at am 1. Pittakis 1835, pp. 460-461; Marie Mauzy for scanning and editing all stages of the work. I also the two are to to inscriptions referred the illustrations; Anne Hooton for the indebted Stephen V. Tracy, Director in as A the present study horos and drawing of Fig. 1; Craig Mauzy for of the American School of Classical at to horos B. preparation of Fig. 5; Judith Binder, Studies Athens, and his staff for I would like to Kevin the resources express my gratitude Nancy Bookidis, Glowacki, the and services of the Biegen to Michael to the following institutions and per late Jameson, Mathilda N. and Gennadius Libraries, and the sons to trustees who contributed this study: the Liberman, Molly Richardson, and other benefactors of Grin Greek Ministry of Culture and A. I. Ronald S. Stroud, and Charles K. nell College for financial support of former in the Greek and to the to Trianti, ephor Williams, anonymous Hespe this study.Finally, Iwould like thank for ria all of whom this Editor o? and Archaeological Service, permission referees, improved Tracey Cullen, Hesperia, to for the work or read her for care study publication rupestral by generously discussing colleagues their and patience horoi on the northeast of the Hill and or all of it. I am in the spur ing evaluating part preparation of this publication. to are of the Nymphs; Munindra Khaund for especially grateful Dorothea Lalonde All translations those of the help with computing technology; for her generous assistance and support author, unless otherwise noted.

? The American School of Classical Studies at Athens 84 GERALD V. LALONDE

100 200 300 AH 1999

1. Figure Western Athens. A. Hooton, Agora Excavations THE BOUNDARY OF MELITE AND KOLLYTOS 85

"SE

Figure 2. Athens (1931) from the the to shrine of Zeus referred by horos A.2 In antiquity visitors approached northeast of the Hill of west; spur an this shrine from adjacent road, dubbed here "Nymphs/Pnyx Street" the Nymphs in the foreground. 1), via a crude rock-cut 3:C) up to the of the Photo courtesy Agora Excavations (Fig. stairway (Fig. plateau a cut spur and then path (Fig. 3:D) with treads in its surface leading from east to west two up the south edge of the spur. The inscriptions are located on m the north side of this path about 12 west of stairway C; horos B lies m 2.0 north of the rock-cut path D, and horos A, directly behind and par to m to were allel it, another 1.39 the north. Both inscriptions cut facing south with the obvious intention that they be read from path D. Beginning two a with Pittakis's editio princeps, the inscriptions have had substantial one editorial tradition, but that has left many questions unanswered and was a problems unsolved. As I preparing complete epigraphical study of horoi A and B for publication elsewhere, the question of the relationship of two the inscriptions loomed distinctly enough towarrant separate treatment a here. Could they be explained in common context, or did horos B have a was to separate purpose, and, finally, that purpose mark the boundary between the demes Melite and Kollytos? to Relevant the relationship of horoi A and B is the question of their dating. The dating of horos A in prior publications has ranged widely in the 6th and 5th centuries B.c.3 Its general appearance (Fig. 4) puts it among the in we can out oldest horoi published /GI3, but rule C. E. Ritchie's suggested a as upper and lower bars of closed heta (B)4 natural fissures parallel with

2. E.g., Pittakis 1852, p. 683; Milch p. 188; Ritchie 1984, p. 540; Lazaridou see h?fer 1885, p. 153;Wachsmuth 1890, 2002, p. 40. For 6th/5th century, pp. 255-256. Judeich (1931, p. 398) IG I31055 A and B. For the early 5th see andWycherley (1978, p. 188) note the century, Meritt, Lethen, and Sta to no no. shrine but refer evidence beyond mires 1957, p. 91, 37. the horoi. in term 4.1 follow LSAG using the 3. For a see Harri to 6th-century date, "heta" differentiate the early Attic son letter H from 1890, p. 108; Judeich 1931, p. 398; for the Spiritus asper Ionic Ervin H for 1959, p. 156;Wycherley 1978, (eta) the long-^ vowel. 86 GERALD V. LALONDE

N

t

s/ lg<^Aw?i4i??^*

we are 3. Northeast of the Hill others above and below the letter, and thus not obliged to date horos Figure spur of the After J. Schmidt, in A before the early third quarter of the 6th century, when Attic inscriptions Nymphs. Curtius 1868, 7, with additions seem to have H = pi. by fully adopted the open h.5 The evidence that tailed rho G. V. Lalonde first appears in the last years of the 6th century or, at least, in the very early an part of the 5th century is helpful for approximate anterior date,6 but setting the lower chronological limits of both tailed rho and three-barred amatter sigma is of continuing, intense debate.7 Nor does the right-to-left direction of horos A indicate an earlier date than B, as the old idea that to to was Attic writing evolved from retrograde boustrophedon orthograde refuted long ago by L. H. Jeffery.8 are a means While letter forms by themselves very fallible of dating on a inscriptions, except broad scale, reliably dated texts with lettering to an comparable that of horos A may provide approximate date. Particu are larly noteworthy parallels the Archaic horoi of the Athenian Agora a came (Fig. 5).When second in situ horos of the Agora to light in 1967 = (Agora XIX, H26 [I 7039] IG I3 1088), Homer Thompson noted that ca. b.c. was the date of 500 that proposed for the horos found in situ in = 1938 (AgoraXIX,H25 [I 5510] IGl31087) on the basis of pottery in its

c or 5. Ritchie 1984, p. 540, TA 112, (Agora I 2470, fir. a, [= IGV 231, the middle of the 5th century in 13 he drew fir. recent suppl. fig. (here correctly a, c]), pi. 29\a, c; p. 102; Vanderpool its last quarter. The articles of as the supposed horizontal top bar 1942, pp. 329-333, fig. 3 (= IGV 2 Henry (2001) andMattingly (1999) and for enclosed 500 oblique rounded); [ca. b.c.]). support those respective positions and see Attic 7. to the is can serve to heta, LSAG, p. 66, heta, type Central polemic whether recapitulate the debate and 66 for the Attic of the and its most 1, p. adoption Athenian Standards Decree (IG I3 bibliography. See, recently, = the H h. 1453; see, e.g., GXLVIII 58) and Matthaiou 2004. 6. see also the LSAG, p. 67; Jeffery Egesta Decree (IGV 11; see, e.g., 8. LSAG, pp. 43-50. on 1948, p. 88, commentary no. 66 ffiGXLVIII 55) are to be dated about AND KOLLYTOS THE BOUNDARY OF MELITE 87

Figure 4. Horos A (IG I31055 A). Photo and drawing G. V. Lalonde

not setting trench could establish the dating for the whole series, that is, long before the end of the 6th century.9 John Papadopoulos has recently a noted that reliable dating of relevant pottery deposits is still in question a as as and that the letter forms of the Agora horoi could allow date late was shortly after 480 b.c., when he has argued that the Classical Agora as founded.10 The letter forms could conceivably be that late, but T. L. Shear noted in 1939, the next layer of road metal above the level through which was the horos set contained ostraka of Themistokles and Hippokrates that to ostra may have belonged, according to Thompson andWycherley, the cism of 482 b.c.11 an treatment are as I suggested in earlier of these horoi that if they to enter sociated with a law of Kleisthenes forbidding atimoi the Agora more attrac (Dem. 24.60; Andoc. 1.71.76), then the earlier dating maybe are tive.12 To be sure, the letters of the Agora horoi somewhat smaller and

9. Shear 1939, pp. 205-206 trench); the datings by Shear (1978) 1976, pp. 78-80 and n. 27, for the metal (mentions only the layer of road and Camp (1994) closely approximate suggestion that Classical atimia in the ostraka above sense containing the ground that of Thompson. legal of the deprival of rights with the 289 level contemporary planting 10. Papadopoulos 2003, pp. rather than the older outlawry and of horos I 5510; see below and n. 11); 291 and nn. 112-113. exile may be associated with the 11. 117. abolition of and the reforms of Thompson 1968, p. 63;^r

r "T '-4r;

-> ^r I 5675 I 7039

^U*W*^14^V^^VAW^VV^^^^

I 55I0

W.B.D.,JR. -I967

,50 l,00

5. Horoi of the Athenian doser together than those of horos A, but this may stem in part from the Figure W. B. Dinsmoor of the texts and their schematic limitation to the Agora. Jr., Agora greater length Agora upper Excavations one margin and part of lateral margin of the front face. More importantly, texts common the Agora horoi have in with horos A three-barred sigma, rho with angular loop and tail, and omicrons smaller than the other let one ters; Agora I 7039 (Fig. 5, right) also has the retrograde direction and on reversed sigma. In sum, the basis of horos As letter forms and their to seems to similarity those of the horoi of the Agora, it reasonable date horos A to the very late 6th or very early 5th century b.c. With the exception of just two scholars, all who have treated both a horoi A and B?including, of course, all who have considered them text as must boustrophedon (see below)?date them one.13 It be noted, however, that the two texts differ considerably in style and script. Horos B not a (Fig. 6) only has rougher, larger lettering and left-to-right direction,

in 13. Exceptions: Threatte (1980, Ritchie (1984, p. 166) dated horos B to used also the late 6th century and p. 53), without giving dates, noted that the "first half of the fifth century B.c." possibly in the late 5th century (see a on nn. the style of the lettering indicates the basis of tailed rho and three above, 6-7). two were different dating for the texts; barred sigma, but these forms THE MELITE AND BOUNDARY OF KOLLYTOS 89

'' ' ' " ' ' 6. Horos B Figure {IGV 1055 B). j >"\^^ '^ / Photo and drawing G. V. Lalonde ;J

but also differs from A in the rounded loop of rho, the closer junction of to the loop and tail of rho the vertical, the proportionally larger size of the omicrons, the less-extended sigma, and the closer spacing of letters. These was to a differences indicate that horos B carved according different design, at a a and very probably different time, but they do not indicate significant chronological distinction. The brevity and eroded state of horos B make comparison with other inscriptions difficult, and, given the uncertain lower no date of tailed rho and three-barred sigma, it seems wise to put narrower on chronological limits horos B than the late 6th to the late 5th century b.c. even a Thus, though the similarity of the Agora horoi suggests date early in that period for horos A, and it is quite possible that B is considerably to sure a later, in the end it is impossible be of chronological difference. a Despite the enduring notion that A and B together constitute bous are reasons trophedon text,14 there ample to conclude that A and B are not boustrophedon: (1) it is inherently difficult to classify two inscribed lines in

14. In 1852 Pittakis the two term its two gave "boustrophedon," employed tinguished the horoi by the letters horoi but standard a a separate numbers wrote, indicators: single inscription A and B but still treated them under "eioiv aDTai tod tod arrows for two a at rpcmoD BoDOTpo number, opposing the single number, placed period the (pn?ov KaXoDja?voD" (Pittakis 1852, texts, and a reference to them in the end of B as if the two horoi were a no. Hiller von Gaertrin as "v. 1" and "v. 2." Most two p. 683, 1134). commentary single inscription, and printed the in IG V while not the the editors of 1055 dis texts with arrows. gen, 863, using recently, IGV opposed 9o GERALD V. LALONDE

on a different styles and letter forms two ridges of rock nearly meter and a as sense half apart boustrophedon, at least in the usual of the term, and even were same if the two inscriptions cut in the style, that would not make a use was conser them boustrophedon text; (2) albeit the of boustrophedon some rare vatively prolonged in religious texts, it is nonetheless after 540, as no and, noted above, tailed rho definitely puts A and B earlier than the as a text late years of the 6th century;15 (3) A and B boustrophedon would use on a be similar to the redundant of the word ?opo? stele, the rare occur rences of which are anomalous and not associated with shrines;16 (4) ifA were on and B in the same locations separate stelai, they would hardly be a considered boustrophedon inscription; (5) A and B lack the syntactic a a continuity typical of boustrophedon text; (6) boustrophedon horos in a scription would be unparalleled inAttic epigraphy;17 and finally (7), if bou strophedon inscription had been wanted, it could have been cut, with some on compromise in letter size, the single superior surface where A was is located. For these reasons, I suggest that Adolf Kirchhoff uniquely correct two among all Corpus editors in publishing the horoi under separate as numbers (IG I 504 and 505) and not boustrophedon. no There is cogent dissent from the view of the distinguished topog a raphers and archaeologists cited above that horos A marked shrine of on not Zeus the plateau of this bedrock spur.18The single-word horos B is so easily explained. Epigraphists have occasionally treated the horos of Zeus (A) alone, but nearly all who deal with both A and B take them both or as context without commentary question having the single of the shrine are a of Zeus.19 Given, however, that horoi A and B not boustrophedon next can a inscription, the logical question iswhether B be considered in or as no single context with A something quite separate. To this there is easy or definitive answer, in part because the inscribed word "horos" by itself or one in the Late Archaic Early Classical period could have of several mean meanings and refer to any number of things. It could "boundary," or a "boundary marker," "domain," and horos, with any number of words, a a could identify domain and simply its general proximity rather than precise boundary.20 Still, considering the proximity of the two texts, their lack of clear chronological distinction, their nearly parallel axes, and their similar a on sense to orientation to reader the adjacent path (Fig. 3:D), it makes consider first whether the purpose of horos B was related to that of A. Both inscriptions, for example, may have marked the shrine of Zeus, but was cut more the simply worded horos B conceivably first, and the fully

n. a to mark a struck 15.Jefferyl948,p.86. (above, 13) precludes boustrophe conjectured place by 16. IG II2 2569; Agora XIX, p. 32, don text. lightning. 4 twice on the same 18. a 19. Ritchie's H49, pi. (opo? Nonetheless, popular opinion (1984, pp. 167,542) in that the statement that both horoi stone); IG II2 2693 (use of stone for the modern neighborhood probably of are remnants of identified the same shrine of Zeus is consecutive security horoi). cuttings the spur to a that 17.Threatte (1980, p. 55) shows ancient houses is echoed occasionally closest consideration they may in in Lauter-Bufe have had that the horos IG V 862 (=IG V 1068) scholarship: e.g., separate purposes. is not as for horoi A and Lauter where no notice is 20. See 5 and nn. boustrophedon; 1971, Agora XIX, p. 1, to as taken of the horos of and in 2. and B, he refers them (p. 53) "a Zeus, Goette where the horoi are rock-cut horos," but his dating of them 2001, p. 57, THE BOUNDARY OF MELITE AND KOLLYTOS QI

a worded A added later for specificity. Marking shrine with the single word uncommon not "horos" would be but unique.21 A second possibility is that B was cut in order to give visitors coming up the South Path (Fig. 3:D) a more general notice of the shrine before the specific ?opo? ?io?. The are come two horoi close enough together, however, that they into view at same virtually the time. In another single-context scenario, horos B could more have had the general purpose of marking the outer limits of the entire to sanctum shrine, while horos A designated the approach its inner farther on north the plateau of the spur. none can con Although of these possibilities be proven, each of them forms to the rule that horos texts usually face the reader outward from the not domain towhich they refer.This would be true of another possible expla a nation of horoi A and B in single context, namely, that A marked the shrine in all its sacred elements to the north, while B differentiated those things to immediately the south?the adjacent path up the spur (Fig. 3:D) and the rooms rock-cut below to the south along Nymphs/Pnyx Street (Fig. 3:E) ?as the common of profane.22 Nevertheless, arrangement outward-facing horoi could logically have been suspended in this case, for if horos B in its present location faced to the north then one would have intruded on what was likely the sacred ground in order to read it right side up. Because of the obviously different styling and lettering of horoi A and one we B and their possible redundancy in such proximity to another, must a consider also the possibility that B had purpose independent of A and the shrine of Zeus. Athenians sometimes used such simple horoi to mark or or or the boundary location of public private land structures. If horos B a had been simple property marker unrelated to the shrine of Zeus, itmight rooms have delimited the rock-cut (Fig. 3:E) immediately to the south case more below the edge of the spur. In that it might have been logical on to inscribe it the spur just above those rooms, with the letters facing so as to north be read from path D, although the surface of the rock is very rugged in that location. Such single-word horoi have also been found at or a or no around tomb burial precinct,23 but there is evidence remotely close to horos B of Archaic rock-cut burials of the sort discovered at the east at western foot of the Pnyx and many other places in these Athenian hills. An anonymous Hesperia reviewer suggested that I consider whether B might have marked property encumbered as security for debt. This a more a seems at least possibility, but its consideration requires than brief explanation. Even though the earliest literary reference to security horoi

21. that a stelai in situ at the See Agora XIX, p. 22, H2, pi. 1, p. 101, for speculation dislocated (OPF) sanctuary one two of Demeter and Kore. these for the rediscovered of stelai [?]opo? from the Agora excavations Although two not (the second is lost) inscribed simply (I 2618; Agora XIX, H43) might have last markers do identify the a ?opo?,which D?rpfeld (1892, p. 91) come from hero shrine beneath the sanctuary by name, they apparently a on terrace of delimited the sacral area of its found by shrine the southwest side the Agora's Middle Stoa dining the and dated to the re rooms from the that ran be of Areiopagos (SEGXXX 37 and IG I3, p. 729, stairway b.c. on as side them. 6th century the basis of letter port the association stronger than see 22. Cf. Lazaridou 40. forms and masonry (IG II2 2507). The speculation); Goette 1995, pp. 235 2002, p. 23. Charitonides dating is controversial (cf. Meritt 1967, 237, for the abbreviated horoi (HO) E.g., 1961, no. an at pp. 98-99, 30, pi. 28; IG II2 2507; marking apparent shrine Panagia pp. 31-32, fig. 53; pp. 67-70, figs. 116, 117. IG I3, p. 729 and the note on "Succincti Thiti in rural Attica; also Corinth Termini"). See also Lalonde 1980, XVIII.3, pp. 200-201, for two inscribed 92 GERALD V. LALONDE

an is in Isaios (6.36) in 364 b.c., and the earliest extant, dated example of inscribed security horos (/Gil2 2654) is from 363/2 b.c., scholars have long debated whether property was alienable from a genos and thus usable as was security before the 4th century. The early favored view essentially that use of Gustave Glotz, who argued that the of real security went back at a least to the early 6th century, view that trusts that the Aristotelian school see was {Ath. Pol. 11-12; also Plut. Sol. 15.3-5) correct in interpreting So s as a lon claim to have freed the land from slavery by uprooting horoi state ment that he gave the "Sixth-parters" (hektemoroi) relief from their debts (seisachtheia) and removed the security markers from their land: rfj (j?aocivoc, xfj? ?yco tcote I?pouc ?ve?Aov KoXXa%r\\ 7T?7rnyOTo:?,I TipooGev Se once dovXzvovaa, v?v ?XevQ?pr\: "Black Earth, from which I pulled up now horoi planted everwhere, before enslaved, free."24 The opponents of this view see it as anachronistic. Prominent among these was J. V. A. Fine, who associated the first use of genos land as security with the debt and dispossession attendant upon the Peloponnesian War.25 was If real property in fact used for debt security in the Archaic and must Early Classical periods, then that property have been marked either or stone by perishable horoi, perhaps made of wood, by yet unrecognized markers of simpler text than the later opo? o?k?ccc 7C?7tpa|uivr|?ml Xvgei and as a the like.We cannot utterly preclude the possibility of horos B security on marker the grounds that to us it is too uninformative, for itmight have a as was implied detailed written agreement between debtor and creditor, case a explicitly the in number of the later multiverbal security horoi.26 It is was an to one or conceivable, then, that B early security marker pertaining more rooms case of the rock-cut shoplike to the south (Fig. 3:E), but the is tenuous. In fact, to return to the question of Solon's horoi, the latest round of as scholarship not only interprets them boundary stones rather than security sees as markers but also Solon's claim of their removal poetic metaphor, in one view standing for the removal of stasis (discord) between the disparate socioeconomic classes, and in another for the freeing of the subjugated hektemoroi, who, like boundary stones, had been bound to the land.27 a was Finally, possible independent purpose of horos B to mark the as ancient road (Nymphs/Pnyx Street) that skirted the south side of the spur a deme boundary, which in this region of the city could only be the shared 24. IE2 vol. fr. lines to 2, Solon, 36, border ofMelite and Kollytos. An immediate objection this proposition 5-6. See Glotz 1904, pp. 325-349; be that the horos is rather far removed from the street, but a may possible Finley 1951, p. 28; Guarducci 1974, rooms counterclaim is that the intervening rock-cut (Fig. 3:E) may have p. 233; Rhodes 1981, pp. 88-97,125 a we prevented closer juxtaposition. Then again, might expect the marker 130,169-179. an even to 25. Fine 1951, 167-208. of urban deme boundary, if it referred the nearby Nymphs/Pnyx pp. 26. XIX, 18 and n. 109. Street, to us more information than "horos," or at least to have Agora p. give just 1997 27. See, respectively, Harris some among the urban demes. rural Attica recognizable parallels Although and Gallo 1999, sources that may serve has several series of horoi from the 4th or later that to the yielded simple century also recapitulate history and are a as viewed by number of scholars deme boundary markers,28 within bibliography of this debate. not a 28. SeeTraill 116-122. the innermost territory of the ancient asty there is single inscription 1986, pp. a 29. Shrine southwest of the Areio that is demonstrably deme boundary stone, and, in fact, horos B is the /GIF Meritt area to pagos: 2507; 1967, only known inscription within that that is rupestral and limited the pp. 98-99; XIX, p. 22, H2. word "horos." the inner urban area has a fair number of Agora in the area Although yielded Tombs of Syntagma Square: one-word horoi inscribed on four were in one a stelai, only situ, marking Charitonides 1961, pp. 31-32, fig. 53; shrine and three marking graves.29 pp. 67-70, figs. 116,117. THE BOUNDARY OF MELITE AND KOLLYTOS 93

as a In the end, the case for IG I31055 B deme boundary marker is, at least from the evidence of its immediate context, inconclusive. Nevertheless, me the investigation of this possibility led to considerable other evidence course that the road in question, throughout its southwestward from the to some Agora point beyond the saddle of the Hill of the Nymphs and to the Pnyx, did in fact divide Melite to the north and west from Kollytos the south and east. Later in this study, I will raise the subject of horos B was a cut again, with the suggestion that it special marker by theMeliteans over to emphasize their authority the precinct of Zeus and other shrines within their territory.

THE TERRITORIALLY OF URBAN DEMES

even common No discussion of Attic deme boundaries, the single boundary as of two inner urban demes, can ignore the ongoing debate to whether the fundamental political units of the Kleisthenic constitution were, simply one put, territories separated from another by boundaries (the "artificial" or model), only borderless administrative units (the "natural" model).30 This too no dichotomy is, of course, neat, because proponent of the artificial model has been able to prove that the Kleisthenic demes were from the beginning precisely defined territories, and most who favor the natural re model allow that the demarch occasionally had responsibilities that was not quired his knowing that landed property in his deme and in that case a of another demarch. The for close and stable boundary of Melite some and Kollytos therefore requires prefatory argument that demes in and the central urban demes in were at some general, particular, time, pos sibly from the beginning, units of territory surrounded by boundaries that or were contiguous with those of other demes of exclusive polis domains, as such the Agora.31 no My research finds claimant to the proposition that the Kleisthenic no to demes received their original membership with regard territorial ity, that is, that anyone qualified for citizenship could enroll in any deme he wished. As C. W. J. Eliot noted, Herodotos usually used the word a a Sfjuo? to refer to village and not political division, and many Kleisthenic

The work of in the land demes into 30 10 30. fundamental modern Haussoullier's interpretation by parts, on of the demes as the 10 of the the organization Attic scribed trittys markers parallels for throughout asty, shore, a and 10 of the In recent is Traill 1975, esp. pp. 73-103; later deme boundary markers (Haussoullier midland."). treatment current and on the basis of accumulat comprehensive of the demes cited IG I 517, 518; for the decades, isWhitehead 1986a. array and interpretation of these ing archaeological evidence, significant see of some extent 31. The thorniness of the question markers, Siewert 1982, pp. 10-16; expressions support for of the initial and continued form of the Traill 1986, pp. 93-113; Agora XIX, of the "territorial model" of demes have can measured the emi four in 3-4 demes be by pp. 14-16,29-31). Just years after appeared Eliot 1962, pp. and nence n. of the scholars who have argued Haussoullier's treatise, his view seemed passim; Traill 1975, pp. 73-75 and 6; views from the same a in the opposing virtually corroborated by passage newly Siewert 1982; Langdon 1985;Traill body of evidence. Typical of the 19th discovered Athenaion Politeia (21.4.22 1986, pp. 116-122; Lohmann 1993, centuries was Bernard text vol. 57-59. Prominent advocates and early 20th 24; of F G. Kenyon, Oxford 1920): 1, pp. Haussoullier's (1884, p. 2 and nn. 3-4) ?i?veiue ?e Kai rr?v %copav KaT?c 8t|uod? of the "administrative model" have been view ??Ka tcov to E. of territorial demes thoroughly TpiaKovTa ji?pri, u?v 7iepi W. Thompson (1971), Andrewes and marked ??Ka surveyed, recorded, by aoTD, ?? Tri? 7iapaXia?, ??Ka ?? (1977),Whitehead (1986a, pp. 27-30), n. horoi. This view stemmed in part from ttj? laeaoyeioD ("[Kleisthenes] divided and Lambert (1998, pp. 7-8 and 24). 94 GERALD V. LALONDE

names demes took their from villages.32 It is fairly inferred from this that, some with variation for the divided demes and probably the inner urban were or demes, the administrative nuclei of the demes individual villages clusters of villages. Moreover, intended mixing of the citizens among the new trittyes and tribes would only have succeeded if the demes drew their membership, for the most part, from the territory around these nuclei. Even David Whitehead, recently the foremost spokesman for the natural model, has noted that "in and immediately after 508/7, naturally, all Eleu sinioi lived in Eleusis, all Skambonidai in (and as) Skambonidai, and so as a on,"33 and has written of the creation of the demes kind of centripetal at or were process in which all individuals registered villages towns that the traditional centers of their areas.34

Since deme membership descended unchanged in the male line, it a is most economical to define deme "in the political and constitutional sense ... a by single technical and non-topographical criterion: the sharing a a common by group of people of demotic."35 There is, however, ample reason for also believing, with Eliot and Traill, that the demes had territo some rial boundaries early, and that the territory of Attica, with exceptions as (probably the Acropolis, Agora, and Pnyx, and possibly, will be suggested below, the public streets and roads), "must have been associated theoreti not one or cally, if actually, with deme another."36 are Among the strongest evidence that demes had territorial limits the epigraphical testimonia of the functions of the demes and of their single chief magistrates, the demarchs, both in their own service and in that of no the larger polis.37 There is direct evidence of these specific duties before the second quarter of the 5th century B.c., but, ifwe can credit the claim of even Ath. Pol. 21.5, that Kleisthenes created the demarchs, it is plausible, not some if provable, that of the responsibilities cited here went back to the formation of the constitution.38 Especially pertinent to the question of are territoriality the demarchic duties that concerned residency and landed were to property within the deme, matters that subject change and therefore could hardly have been carried out without official, recorded reference to can a on delimited territory. We begin with ?yKiniiK?v, tax land within a the deme owned by nonmember or nonresident of the deme. Since Pi raeus's grant of exemption from eyKiniiK?v to Kallidamas of Cholleidai (IG II2 1214, lines 26-28 [ca. 300-250 b.c.]) is the sole extant reference to tax a this by name, Whitehead rightly rejected traditional assumption

in we will 32. Eliot 1962, p. 79. effectively all the subsidiary business mainly finance and law, but 33. 75. of state. Fundamental of this treat in Whitehead 1986a, p. proof religious duties below relation 34. Whitehead 1986a, pp. 23-30. division of labor is that the polis did to the shrines in the deme Melite. For not even a a account 35. Traill 1975, p. 73. keep comprehensive list full of the demarch's duties, n. of at least as as the see 36. Traill 1975, pp. 73-74, 6; citizens, but, early Whitehead 1986a, pp. 121-139 and Eliot 3-4. 1962, pp. third quarter of the 5th century (see Faraguna 1997. Faraguna, however, 37. needs of Athenian Many IGV 138, line 6), and probably from departs radically from Whitehead, administration were initiation re are polis necessarily the of the Kleisthenic arguing that cadastral registries to the as on to in of delegated demarchs, because, forms, relied the demes provide implicit the office the demarch in Pol. it a Aristotle implies 1264a6-ll, their respective XrfcxapxxKa ypaji|LicxTe?a and presuppose high degree of initial would have been for see impossible central when needed; Traill 1975, p. 73. territorial organization. to we note polis offices and magistrates manage 38. Here demarchic duties THE BOUNDARY OF MELITE AND KOLLYTOS 95

that all demes levied it, but he also allowed that the 4th-century grants of al?jela by the Eleusinians to resident Theban benefactors {IG II2 1185, lines 4-5; 1186, lines 25-26) and to Athenian nondemesmen (/Gil21187, lines 16-17; 1188, lines 28-30), and by Coastal Lamptrai (IG IT 1204, or lines 11-12) to its honorand conceivably imply ?YKrnTiK?v something no was like it.39Although there is evidence that eyKinxiKOv universal, it strains the argument from silence in the other direction to suggest that a were this tax and consequent specificity about boundaries unique to the deme of Piraeus.40

Further evidence of deme territoriality is the official formula for iden name + oikcov + name a tifying metics (metic's ?v of deme) used for the a collection of poll tax (|I?to?kiov), the granting of concessions, and various other official accounts involving resident aliens.41 The verb o?K?co clearly denotes residence and not just political enrollment. The works of Clerc and Hommel exemplify the long-held opinion that the oikcov ?v designa tion for metics went back to the time of Kleisthenes.42 Whitehead, while on allowing that demes officially recognized, registered, and kept records as a metics separate class with legal locus standi dating from the reforms of as a Kleisthenes or, corollary of those reforms, from the early 5th century, has nevertheless argued that the relatively late appearance of the o?kcov ?v formula in its "embryonic form" in 414/3 (IGV 421, col. 1, fr. b, line 33) is an not just evidentiary accident.43 Since documentary evidence of Athenian politics is relatively plentiful for about four decades before that time, he may be right, but the practice may have predated the formula. to an The only alternative initial but changeable enrollment of metics seem to a by residence would be that for about century metics were, like citizens, fixed to the deme of their original ancestral registration, but that was as as this system given up late 414/3 in favor of reckoning by current a ex more domicile. The only argument for such hypothesis is si/entio, but, not as a than that, metics did figure in the constitutional objectives, such permanent broad scattering of demes and demesmen among the trittyes and tribes and the maintenance of a stable quota of representatives of the demes to the Athenian Boule, objectives on which the inherited enroll ment was sum no of demesmen predicated. To up, there is certainty about more the metic's early relation to the deme, but the equally plausible, and

n. as or 39. Whitehead 1986a, p. 76, 38; tively [oiK?Ja?^ Pfl, and concludes pp. 75-77; Henry 1983, pp. 244-245; p. 150. that "this inscription, therefore, would Heisserer and Moysey 1986, line 7 to to 40. Jones (1999, p. 59) argued from appear have nothing do with and commentary; Whitehead (1977, a or or that examination of squeeze of/Gil2 2623 boundaries territories, agoras p. 7) may be right "resident alien" (opo? n[ei]lpaecov [%colp]a?. rasura) land' of demes." Cf. SEGXUX 175, is "tired translationese" for jj?toikoc, as but his is con that its third line should be restored XLVIII164,XLVLI193. preferred "immigrant" was set 41. oikcov ev: IG as not all metics were immi yac and that it up specifically The formula V fusing, to an not were enforce the collection of eyKTnri 475-476; /Gil21553-1578,1654, grants and all immigrants to k?v unique Piraeus. Recently Lam 1672-1673,1951; Lewis 1959,1968; metics. see bert (2004) has rebutted Jones's epigra Whitehead 1977, pp. 31-32, 42. Clerc 1893, pp. 339-340; is in and RE cols. s.v. phy, noting that ya unparalleled supplemented by 1986a, pp. 81-83, XV, 1932, 1413-1458, an Attic prose and that autopsy of the 1986b, pp. 148-154; u?to?kiov: IG V Metoikoi (H. Hommel). stone text: shows the following opo? 106, lines 6-7; 107, line 3; /Gil2 61, 43.Whitehead 1977, pp. 31-32, nieipa?cov I [- up to 3-4 -]AI[. I-?]. lines 10-11; 211, line 5; 237, lines 25 145. Lambert restores the third line tenta 26; 545, lines 11-13;Whitehead 1977, 96 GERALD V. LALONDE

a economical, hypothesis is that metic's business with the polis and the was deme from the beginning of the Kleisthenic constitution administered by, and according to, the deme in which he resided, and that the formula oikcov was as an ?v later adopted official expression of that traditional

practice. Also apropos of metics and early deme territoriality is the provision in a lex sacra of Skambonidai that its metics are to have a part in the local sacrifices (IGV 244, Face C, lines 8-9: xo? |I?toik[o? ?,oc%]l?v).As White as head has noted in this instance, "aswell the obvious intrinsic importance a a of this early [ca. 460 b.c.] example of deme's having decided upon such concession, the provision demonstrates not merely that the deme recognizes a means the presence of metics but also that it commands of determining were not which of them living in Skambonidai and elsewhere [my emphasis] ."44 Surely the demarch's precise knowledge of the deme's territorial limits would have been necessary, albeit not sufficient, for that means of determination. This responsibility of the demarch would have been particularly critical in as the urban demes?such Melite and Kollytos?where the metic popula tion was concentrated.45 Further evidence of the territorial distinction of demes can be drawn

from the demarch's collection for the polis of ??o(pop?, the tax usually war on levied for all property-owning free residents, citizen and metic (Dem. 22.61). Presumably the citizen could have registered all his property, regardless of its location, in his deme of ancestral membership, but again on was to the levy metics likely assigned the deme inwhich they resided. The few extant and rather late decrees honoring metics who paid ??o(pop? were {IG II2 421, 554, 715) probably do not imply that metics normally exempt from this tax, although the paucity of evidence may indicate that were there not many metic landholders.46 The demarch also imposed estates ??iTODpyia (required public service financed by individuals with a of superior valuation), and in decree of the deme of Ikarion from the third quarter of the 5th century b.c. the allusion to residency, if correctly a restored in stoichedon, in provision for choosing tragic choregoi from de motai and from persons residing in the deme {IGV 254, lines 3-4: [..5.. .]i tov ??|iot?v Kai x?v 'lKoc[pio? o?k?vtIov 8t)o] t?v ?%op?y?xov), implies awareness of the deme's territorial extent. The same reasoning applies to the demarch's of of or rio supervision ocvt??ooi? (legal challenge A^ixoDpyioc 44. Whitehead 1986a, p. 81. and issuance of assessment of as a (pop?) ??roypa(pri (declared property 45. InWhitehead's survey (1986a, or number basis for payment of taxes, delinquent debts, legal confiscations), the p. 83)Melite has the highest as as (75) of attested metics; in latter attested at least early the confiscations from the profaners of the general, seeWilamowitz-M?llendorff 1887, Hermai and (415-413 b.c.) and the (402/1 b.c.).47 Mysteries Thirty Tyrants 116-124. a of the demarch that has for the pp. Finally, duty particular implications 46. SoWhitehead (1977, p. 78), of demes is found in the Athenian law in Pseudo territoriality quoted who adds that the earliest testimony Demosthenes which the demarch is under of of metics is the state 43.57-58, by required pain paying eiocpopa ment a 1,000-drachma fine?and in default of relatives or, in the case of slaves, of Lysias (12.20) that he and his brother contributed masters?to pick up and bury any corpse in the deme within the day of the greatly. 47.Whitehead 1986a, p. 131 and death.48 The wording of this polis-wide law, "in the deme," and the stated n. 64; Walbank 1982, pp. 95-96. noted of the land corroborates the sugges 48.1 am to objective purifying previously grateful the anonymous tion of Eliot and Traill that the of Attica was for the most reviewer who territory part Hesperia suggested pur one or a this associated with deme another. Lambert, generally proponent of the suing point. THE BOUNDARY OF MELITE AND KOLLYTOS 97

a "natural" formation of demes, agrees, but with chronological qualification, was noting that the community responsibility for burial carried out by the deme rather than the phratry because by the time this law was passed (he or dated it to the late 5th early 4th century),49 "the demarch, unlike the was a as as a phratriarch, responsible for tract of territory well group of per sons."50Whitehead, while giving considerable attention to this demarchic was an responsibility and noting that it anti-pollution measure, is silent about its territorial implications for the demes. One might infer, however, at seems that itwould least fall under those practical purposes for which he some a to allow late calculation of deme's geographical extent.51 are or Granted that the testimonia cited above sporadic unique, still, a case were con in the aggregate, they make strong that the Attic demes as as even tiguous territories well administrative units. Moreover, though none of these cases is dated before the second quarter of the 5 th century b.c., it must be kept in mind that for the first four to five decades under we the Kleisthenic constitution have very little extant detailed epigraphical evidence of Athenian politics in general, let alone of the demes. Therefore, must although claims about the territoriality of demes in these early decades be tentative, it seems very hard, as even the strong proponents of "natural" demes show, to infer from this lack of direct evidence that demes were even no at the outset purely administrative units with need for territorial boundaries. In any formulation of the natural or artificial models of deme to formation it is difficult conceive of distinctions of residency among demes without territorial boundaries. more answer It is probable that the to the question of deme formation lies somewhere between the theoretical poles of purely administrative units and of distinct and contiguous territorial units (the "divided" demes not withstanding) with all of their internal real estate surveyed and recorded. seem a It would reasonable compromise of the general territorial model an to postulate initial determination of at least the outer territorial limits of demes and their citizen memberships, with the data concerning other as inhabitants and interior domains being gathered and recorded piecemeal more the duties of demarchs required them.52 In this evolutionary model, to as the polis would have assigned political business the demes it recog nized that their territoriality allowed them to conduct this business more precisely and efficiently than the traditional gentilitial and political groups. The preceding conclusion leaves unanswered the question o? how the or demarchs marked otherwise kept track of these territorial distinctions, especially from the time of Kleisthenes to well into the 4th century, when

49. Humphreys (1980, p. 98), with 30 drachmas from the estate of the but less fairly Eliot (1962, pp. 146 reference to the Great condemned Theosebes for the some Plague, suggests expense 147), who, while showing that a date ca. 430 b.c. of the latter's in re burying parents. key elements the Kleisthenic 50. Lambert 1998, p. 227 and 51.Whitehead 1986a, pp. 137-138, forms were as late as 501 b.c., noted n. elsewhere in the same work n. that the 121; 258-259 (burial); p. 29 and Ill duration and chronology of Lambert notes (pp. 314-319) that in (calculations of deme boundaries). the survey is hard to calculate, and that a account in use Poletai found the Agora 52.Whitehead (1986a, p. 27 and the organizers may have made of n. excavations (SEGXII100, lines 25 102) fairly criticized Lauter (1982, pre-Kleisthenic censual and cadastral an 29) Isarchos may have been acting esp. p. 305) for positing initial surveys. as demarch of Xypete in recouping cartographical survey of several years, 98 GERALD V. LALONDE

or the earliest claimed deme horoi have been dated, later, with the possible creation of horismoi, putative archival records of deme limits.53 The Attic numerous or countryside has yielded series of abbreviated single-word rupestral horoi that, in part because of their distance from population on or over centers and their alignment ridges saddles extensive terrain, have won as considerable scholarly acceptance deme boundary markers.54 Many use of these inscriptions the lunate sigma, which first appears in the 4th as century b.c., and Traill has speculated that deme boundary markers they may have been occasioned by theMacedonian reorganization of Attica in 307/6 B.c.55 Lunate sigma is most common, however, in the Roman era, and there continues to be shifting opinion and debate about the age and function of the growing number of discovered rural Attic one-word horoi.56 The question of how the outer limits of rural demes were marked or recorded and which of the series of rural horoi might have served this pur as pose is necessarily complex and, the discovery of archaeological evidence a continues, will warrant investigative and theoretical study for long time come. now we as a to Here and will put aside that question and, prelude to our the discussion ofMelite and Kollytos, turn attention to the delimiting of urban demes, and specifically those in the interior of the city. as If, proponents of the artificial model believe, geographical bound aries were necessary for the function of the demarchs in the rural and a coastal demes of Attica, fortiori the greater density of population in the inner asty made official deme boundaries essential. Whitehead, while even eschewing any compromise of the "natural" formation of the demes in the city,57 made the important point that the demes of the asty do call as for consideration three different types: those in the inner populated as center of Athens, such Melite and Kollytos, those in the immediate as "suburbs," such Keiriadai and Kerameis, and those outer-city demes a centered in villages in open country, the last having character and mode of formation no different from the demes of the shore and interior.58 The

rest of this article is concerned mainly with the innermost of the demes in that tripartite scheme, and only incidentally with those in the suburbs. As Whitehead has noted, the archaeological evidence that 8th-century Athens was a cluster of distinct villages says little about urban Athens of

to cadastre I once 53. A possible reference polis of private property; p. 141; Eliot 1962, p. 3; Lewis 1963, of deme comes in a considered records territory (AgoraXJX, pp. 20-21) p. 27; Andrewes 1977, pp. 243-244; scholion to Ar. Av. 997 (Me?itri y?p that the demise of security horoi in Murray 1980, p. 255; Rhodes 1981, 0C7iaV ?V b.c. ?K??V0, C?? T01? X_)plG|U0?? the early 2nd century may have p. 254). y?ypa7rcou xfjc Tz?Xzwq, "for that whole followed such a cadastre. 58.Whitehead 1986a, pp. 25-26; it as is written in the 54. Eliot is in this that even thing isMelite, See, e.g., 1962, esp. noteworthy regard Traill n. the outer area of the has boundary records of the polis"), but pp. 1-2; 1975, p. 73, 6; Lang asty yielded were are of there is doubt that these records don 1985; Traill 1986, pp. 116-122; horoi that possibly markers later see earlier than the 3rd century (Lewis Lohmann 1993, vol. 1, pp. 57-59. deme boundaries; Traill 1986, no. 1955, p. 17, n. 48); Jacoby (FGrH 55. Traill 1986, p. 118. pp. 117,119-120, 4, for the sug that the Illb II, p. 402) suggested that the 56. See, e.g., Goette 1994; also gestion pair of single-word a on horismoi may have been book written Langdon 1999 the 19 horoi found horoi (IG IV 2521) that Christopher as as b.c. on first on late the 1st century about Alepovouni. Wordsworth noted Lykabettos not sections of the city and about 57. SeeWhitehead's (1986a, pp. 24 and ?%iOTf|n?Tpa (the latter probably seen lost to a demes. As will be below, the full 25) criticism of the majority, includ quarrying) may have marked text is some of "natural" rural of land to the tribe of this scholion quite unreliable. ing proponents boundary assigned nn. see creation of the the Finley (1951, pp. 14 and 19,27) demes, who the Antigonis from Kydathenaion, saw no for a urban demes as artificial most extensive urban deme. virtually evidence public (Young 1951, THE BOUNDARY OF MELITE AND KOLLYTOS 99

was the late 6th century,59 and perhaps he right that Andrewes begged the question when he wrote that at the time of Kleisthenes, "the area within was not an the city wall agglomeration of distinguishable villages," thus were an area implying that inner urban demes formed in that had become a fair continuum of domestic, industrial, political, and religious domains streets intersected by and roads and extending outward from the focal as as precinct of the Agora far the Archaic city wall and probably beyond.60 more Nevertheless, it is noteworthy that little than two decades after Kleis thenes' reforms, the placement of the Themistoklean city wall suggests an Athenian urban population considerably larger than that served by any estimated extent of the Archaic wall.61

To this day many factors complicate the question of the demographic shape of Archaic Athens: the literary evidence is later and sketchy; much of not no the ancient city has been excavated; there is comprehensive survey of findings at the Archaic level; the evidence of the 8th and 7th centuries consists mainly of graves and wells; and the 6th and early 5th centuries, while showing significant sacred and civil architecture of the Peisistratids and the early democracy, have not been closely studied for their residen tial patterns.62 Nevertheless, if Papadopoulos is right that the creation or was of the Classical Agora, whether in the late 6th early 5th century, area was a possible because the before that loose conglomeration of Iron Age cemeteries and ceramic industries, then it is quite possible that at the were time of Kleisthenes there other unsettled spaces among the populated areas case west of the Archaic city. This may particularly have been the was and northwest of the Acropolis when the main focus of city life still to the southeast (Thuc. 2.15.3-6) and before Themistokles as archon in to move 493/2 began maritime activity from Phaleron Bay to the ports of were Piraeus (Thuc. 1.93.3-7).63 Even if there closely settled expanses of inner Athens at the time of Kleisthenes, the new urban demes could have names were drawn their and locations from what earlier separate population or we no centers, villages, komai?"proto-demes."64 Although have direct evidence of the basis on which boundaries of inner urban demes were cre to some ated and stabilized, they would have had be devised along logical to lines analogous the marked ridgelines and saddles where horoi have been identified as markers of the late rural and coastal demes. Some scholars have inferred that the traditional streets and roads of the

urban region would have best served this purpose.65 Even W. E.Thompson,

59.Whitehead 1986a, p. 6 and Travlos (1960, pp. 33-34,40-42; 62. For summaries of Archaic Athe n. 26-27. in the between see 11, pp. Athens, p. 8, fig. 5) valley nian topography and monuments, 60.Whitehead those hills and the More 1986a, p. 26; Areiopagos. Judeich 1931, pp. 60-70; Camp 2001, Andrewes 1977, pp. 243-244. persuasive isWinter's (1982) argument pp. 22-47. 61. For a of the Archaic from and the location of 63. dating city defensibility Papadopoulos 2003, especially wall to ca. 560 see Archaic burials that it skirted the west his b.c., Vanderpool topographical summary, pp. 271 for that wall in and a later ends of the and 1974; general Acropolis Areiopagos, 316. to see dating the Peisistratid period, but it is hardly plausible that the inner 64. SeeWhitehead 1986a, p. 27, Weir 1995. In either case, the structure urban demes could have been confined n. 98 and the reference to Isocrates' of the Archaic wall could to so little In the end it must remark that city hypo territory. (7.46) the Archaic lawgiv have been the outer limit of be said that we should not ers thetically expect polit of Athens divided "the city by komai the inner urban but the location ical divisions to with and the demes, correspond city country by demes." of the wall is much the of which is a 65. disputed. Judeich walls, placement E.g., Langdon 1985, pp. 12-13; and it matter of in terms of ter (1931, pp. 62,120, plan IV) put defensibility Agora XIX, pp. 13-14,29. at the crest of the western and rain and hills, population. IOO GERALD V. LALONDE

in his seminal argument for purely administrative demes, admitted that the as demes of the asty initially needed particular lines such rivers and roads to determine the first registry of urban demesmen, although he added that the boundaries could then have been dispensed with and forgotten.66 Even were seen as if the roadways first simply the most efficient route between significant points, since they avoided natural and man-made obstacles in not the intervening terrain, this does defeat the hypothesis that Kleisthenes adopted existing streets or,where vacant space allowed, the addition of new streets as a urban deme boundaries. In fact, such hypothesis is bolstered by both logic and inference from indirect evidence. Since the courses of streets not to as tend change without deliberate initiative, those serving deme borders would have been easily retained in the collective memory of even the community and thus would have had stability without the ben or as efit of horoi archival records.67 Moreover, existing streets and roads, public property, would have been the logical choice for deme boundaries use or because they could have been put to that without the expropriation or disturbance of other domains, public private. streets Of 21 certain Attic horoi of and roads, mainly from the Athens are or region, nine inscribed simply o?ou opo? o?ou, thus not identifying name the roadways by but maintaining their official limits.68 The strict de streets limitation of from other domains is explicit in several horoi found in one to Piraeus, of them in situ, which refer both the street and the adjacent as as property.69 At least early the 5th century, the Athenian Poletai officially recorded the location of confiscated properties with reference to streets.70 One of the duties of the ten Athenian Astynomoi, five in the upper city and was streets five in Piraeus, to keep free of encroachment ([Arist.] Ath. Pol. to as as 50.2), probably in part maintain their function boundaries.71 If, seems streets were likely, this jurisdiction of the Astynomoi implies that the not deme territory, but were, like the Agora, in the direct and sole purview of the polis, then Andrewes's point that streets united rather than divided people loses cogency.72

66. 4th XXXIX Thompson 1971, p. 75; Langdon 69. IG V 1109-1113;^ XIX, Mygdonia, century b.c.; 1985 is a to nn. 1116 from 2nd strong rejoinder Thomp p. 13 and 66-71; JGI3 1123, Mylasa, century b.c.). was son's article. Eleusis probably another example, 71. See /Gil2 380, lines 16-23; 67. The of the course of but a of the text is see n. 6. Cf. in the wider Greek longevity key part lost; Sylt.3 313, streets ca. and roads, and therefore their Agora XIX, pp. 10-11,27, H25, H26, world SEGXI?I 785,Thasos, 470 as and for the two in situ reliability boundaries, is well illus above, Fig. 1, 460 b.c.; OGIS 483; SEG XIII 521, in of horoi of the that delimited the b.c. trated the archaeology Britain, Agora Pergamon, 2nd century civil the street n. where great numbers of routes, both precinct from that passed 72. Andrewes 1977, p. 244, 1; urban not since also the linear function of roads unites and rural, have changed through it; Agora XIX, pp. 11-13, for the six horos stelai Andrewes was Roman rule. 28, H30, H31, people, but clearly of the Kerameikos that either delimited to lateral function. An 68. Athens: JGI31093,1094, referring their 1094bis, 1095,1096; IG II2 2624, the district of Kerameikos from the amplification of Langdon's (1985, street that ran it or the 2626; SEG XU 13;AgoraXJX, H32 through designated pp. 12-13) analogy of delimiting H35; Piraeus: IGV 1109-1114; Eleusis: the street as Kerameikos. of modern French political districts is 70. an instructive to Athenian IG V1116; IG II2 2625; Panormos See, e.g., Agora XJX, pp. 57-143, parallel a see streets as boundaries (Laureion region): SEG U 157; horoi and, for specific instance, below, polis-administered ooou or o?o?: 106. roads were used that divided urban demes. American inscribed simply opo? p. Similarly, as state codes are also and IG V 1093,1094,1094bis, 1114; IG1V widely in the Greek world points analogous, SEG U even more than Athenian law 2625,2626; 157;Agora XIX, of reference in boundary regulations precise SEGXL in the midline of streets and H32, H35. and agreements (e.g., 542, setting THE BOUNDARY OF MELITE AND KOLLYTOS IOI

THE BOUNDARY OF MELITE AND KOLLYTOS

we can now return to that the street With this background the proposition to crest saddle was the from the Agora the of the Nymphs/Pnyx boundary a we deal with a controversial of Kollytos and Melite. As first step, should the observation of Eratosthenes (quoted in Strabo 1.4.7) about boundary an to of these two demes?an observation made as analogy the boundaries criteria. of continents, which geographers argue about without agreed-upon no to we assume that Strabo's Here, with particular evidence the contrary, is understanding and communication of Eratosthenes accurate, although was not the case: that always

nil ovxcov y?p ocKpi?cov opcov KaGarcep Ko??dto? Kai Me?rcn?, oiov orn?Av n, 7??pi?OAxov, xomo U?v 8%8iv (pavai fijaac, on tout! ji?v ?axi KOAA/UTO?,TOUT? ?? MeX?XT\, TO?? OpOU? (??) UT| ?'%?lV81718?V.73

no as case If there are precise horoi, in the of Kollytos and Melite, we such as stelai or enclosure walls, can say this: "This is Kollytos, that isMelite." We cannot, however, point to the horoi. as as a The beginning of this passage is usually translated, here, present an actual instance. In this particular condition, with KaB?rcep introducing translation, the ambiguity of the word opoi is deliberately maintained. as as must be Translating it "boundaries," is sometimes done,74 wrong, or Eratos because boundaries cannot be equated with stelai enclosure walls. or thenes must have meant by ?poi "boundary markers," simply "markers." a not Thus, if this is present general condition, it does imply that Kollytos and Melite do not have boundaries, but that their boundaries do not have to precise markers. It is reasonable conclude then that, lacking such precise one two markers as stelai and enclosure walls, could distinguish the demes some less exact means. by was In conceiving of ameans of delimiting Melite from Kollytos that sense to less exact than stelai or enclosure walls, it makes good fall back on our for the use of and and arguments above streets, roads, prominent permanent features of the cityscape for the determination of urban deme boundaries. These features would allow one to say, "This isKollytos, that is as Melite," without being able to cite markers as precise stelai and enclosure runs counter to walls. This interpretation of the passage, however, that of three distinguished scholars of Greek, W. M. Leake, Rodney S. Young, andW. Kendrick Pritchett, who inferred from this passage that there were horoi inscribed "This is Kollytos" and "This isMelite." It is grammati a cally conceivable that Eratosthenes intended here mixed condition with were no a a present contrary-to-fact protasis ("if there precise horoi,")

as who roads the official boundaries of on either side of the street. Here the of G. Aujac, Paris 1969, accepts or wards. This is in the last clause the ?? ofMSS a townships precision, position of the residents somewhat is for administrative to that of the demesmen and B2. however, merely comparable as 74. in Loeb edition of purposes, for the American citizen is of ancient Athens, state and county E.g., the not allowed to reside or build are to the trans. H. London along officers Astynomoi, and their Strabo, L.Jones, streets out to to 1917. these the legal boundary record offices the putative horismoi. or even on a of margin right-of-way 73. The text is the Bud? edition I02 GERALD V. LALONDE

are use present general apodosis ("we able to say"), and the of Koc0?7t?p to as a introduce Kollytos and Melite hypothetical example ("for instance, case Kollytos and Melite").75 In that the inference of Leake, Young, and a Pritchett, that Kollytos andMelite shared boundary marked by inscribed horoi, would be correct. Either of the preceding interpretations would fit the wider context of case Eratosthenes' macrogeography. He presents the of Kollytos andMelite as an to analogy to his immediately preceding point that it is futile quibble a about the division of the continents. All geographers have rough idea an of the continents, but without agreed mode of marking their bound aries?as it is, some use rivers and call the continents islands, others use are as can isthmuses and call them peninsulas?they the Athenians. They over say "This isAsia" and "This is Europe," but they argue the fine points of delimiting them. After the analogy of Kollytos andMelite, Eratosthenes, to according Strabo, then reemphasizes his point with historical examples over of the perennial disputes of the Argives and the Lakedaimonians over no Thyrea, and the Athenians and the Boiotians Oropos.76 There is we tradition about such contention between Kollytos and Melite, but do not have all the evidence, and I suggest below that Melite may have used to to on horos B and another rupestral marker emphasize its claim shrines own its side of the border with Kollytos. However Strabo's quotation of the geographical views of Eratosthenes no were is construed, it leaves doubt that Melite and Kollytos contiguous. common The location of their boundary, however, is quite another mat are or even a ter. Few exact boundaries of urban demes known argued, but number of modern maps and commentaries suggest the general location sources was west of demes.77 Virtually all of these agree that Melite and some southwest of the Agora, but they vary somewhat about Kollytos: limit area it to the south of the Areiopagos and the Acropolis, while others add to that the area directly south of the Agora. Before David M. Lewis refuted was a the long-held belief that Kolonos Agoraios deme, inclusion inMelite to even was of the Pnyx and, according some, the Mouseion considered as one most essential to accommodate its status of the populated of the no urban demes.78 Lewis's finding, however, has apparently effected shift of opinion about the southern extent of Melite.

n. as the to inset. 75. Leake 1841, vol. 2, p. 442, 3; missing superfluous attempts pp. 123-144, map (end page), n. the boundaries of the con 78. Lewis 12-17. For the Young 1951, p. 140, 12; Pritchett determine 1955, pp. in tinents. Eratosthenes' how 4th and the 1953, p. 276; opposing these schol point was, century following, substantial of bouleutic ars,Thompson (1971, p. 74) rightly ever, that determining the division of data quotas out that is the continents would not be a matter of 125-140 pointed KaOarcep normally (Traill 1975, p. 67; 1986, pp. and rosters used to introduce actual instances, but useless disputation if the geographers map), ephebic (Reinmuth use here is had an set of no. Gomme the possibly irregular. agreed aKpi?oi ?poi 1971, pp. 5-10, 2; 1933, over rivers and Hansen 76. Strabo himself in the following instead of quibbling p. 67; 1986, pp. 77-79), intimates that he is isthmuses. et al. chapter (1.4.8) funerary inscriptions (Hansen table and metic resi uncertain of Eratosthenes' meaning 77. Judeich 1931, pp. 165-175, 1990, p. 33, 9), him 14 and Kirsten and Krai indi (r\ TotJto Aiyei), for he castigates fig. plan I; dences (Whitehead 1986a, p. 83) cate that Melite was at least second to for inconsistency in stressing the prac ker 1967, p. 97, fig. 21; Traill 1975, in the cases of Siewert in ticality of horoi Kollytos, pp. 37-55, map 1, inset; 1982; Kydathenaion population. dis Traill Melite, Thyrea, and Oropos, but Langdon 1985, pp. 11-15; 1986, THE BOUNDARY OF MELITE AND KOLLYTOS 103

We will return to the specific question of the Pnyx below, but the first our step in the argument of hypothesis of the boundary ofMelite and Kol some lytos must be to establish with certainty the full extent and course of as the road proposed that boundary. For this, there is good archaeological, not epigraphical, and literary evidence that has previously been synthesized in its entirety. Young, excavating the so-called Industrial District southwest of the to Agora intermittently from 1939 1949, uncovered much of the initial course of this road farther down from the Hill of the Nymphs, where it west street branched off from the Agora's (hereafter, "AgoraWest Street"; see m Fig. 1) about 20 southeast of theTholos?at the horos of the Agora found in situ in 1938 (Agora I 5510).79 From that point it followed the line of the Great Drain southwestward for about 170 m, in a stretch that Young called the "Street of theMarble Workers" (Fig. I).80 Just after passing on the northwest side of the poros complex identified by Vanderpool and m the Agora plan as the State Prison,81 the road jogged about 60 directly west street before it crossed the main from the northwest (Young's "Melite Street") and turned southwestward up the valley between the Hill of the Nymphs and the Pnyx. In this last stretch (Nymphs/Pnyx Street) it passed the south flank of the northeast spur of the Hill of the Nymphs, with its over horos B, before continuing the saddle in the Nymphs-Pnyx ridge and exiting the city through the gate toward the north end of the Diateichisma, at least when that wall existed.82

In his excavation report of 1951 Young carefully pondered whether his Industrial District and the land farther south and southwest were in or area was the deme Melite Kollytos. He conjectured that the probably in was not so as Melite, but he specific to question the judgment of K?hler own and Judeich that the merger and continuation of his Areiopagos and Melite Streets southward into the valley between the Areiopagos and the were narrows Pnyx the cnxvam??... Koaa/uto? ("a called Kollytos"), said by = the 4th-century rhetorician Himerios (Or. 31.63-65 Phot. Bibl. cod. 243) to name be in the very middle of the city, to give its to the deme through which it and to serve as an ran, agora.83

tant of 5th cen ferred from the in 79.AgoraXD?,H25. historical figures the gate the Themis been derived from ear toklean wall to the 80. Young 1951, pp. 137-147 and tury, may have city corresponding the in no fig. 3. lier testimonia that Melitides Gate gate the Diateichisma, there is was in evidence of it. 81. Vanderpool 1980, but cf. Crosby the original Themistoklean city 1951; Koumanoudis 1984; Hunter wall, which in the 5th century looped 83. Young 1951, p. 140; K?hler out 1998, pp. 319-323. considerably southwest of the line 1872, p. 112; Judeich 1931, p. 169, n. n. 82. The traces of this gate have been of the later Diateichisma (see Judeich 1. Stroud (1998, p. 89, 10) has us obliterated by quarrying. See Fig. 1, 1931, p. 140 and plans LB-D/5-7, IV; recently reminded of Judeich's is the "Melitides Gate," which desig Wycherley 1978, p. 12). Judeich (1931, (1931, p. 169) observation that the nation of Travlos (Athens, pp. 161,168 p. 140 and plan I:C 7), for example, aphorism of Plutarch, Mor. 601B, that not all Athenians were 169, fig. 219, Gate XV). The Melitides tentatively located theMelitides Gate wealthy was KaxoiKouoi Gate said by rather late ancient farther south along this Themistoklean (ou?? y?p AOrjva?oi 7iavxe? sources (Marcellin. Vita Thucydidis 17 loop toward theHill of theMuses, but Ko?Xdtov: "For certainly all Athenians and Paus. to be near his was doubtless influenced not in 55; 1.23.9) Koile, placement do live Kollytos") iswell reflected the ex in on where tombs of Kimon and Thucy by his belief that the deme Melite the grand houses excavated the dides the historian could be seen. These tended as far south as to include the western and southwestern slopes of their mention of If the name Melitides was trans sources, with impor Pnyx. the Areiopagos. I04 GERALD V. LALONDE

Two years after Young's article appeared, Pritchett published among a = fragments of the Attic Stelai notice (stele VI, lines 13-15 IG I3 426, one a lines 5-8) of the confiscation from of the Hermokopidai of house as on one described being in Kollytos and bordered side by the Agora s and on the other, according to Pritchett restoration, by the Aifonceiov], the shrine of Aiakos. Joining this evidence with the testimony of Hime rios about Kollytos and that of Herodotos (5.89.3) that shortly after the reforms of Kleisthenes the Athenians marked out a shrine of Aiakos eki

xfj? ?yopa?, Pritchett concluded that the Aiakeion should be sought at corner the southwest of the Agora.84 Against Pritchett's Aiakeion, both on Lewis and Eliot suggested the restoration of Aijavxeiov] the grounds was a son that there probably cult-place of Aias together with that of his on Eurysakes the Kolonos Agoraios.85 Ronald Stroud, however, in his recent B.c. editio princeps of the inscribed Athenian grain-tax law of 374/3 (Agora I 7557), confirmed Wycherley's earlier consideration that the two an only mentions of Aianteion inAthenian inscriptions (/Gil21008, line 87; Agora I 286, decree V, lines 140-141) may well have referred to the shrine of Aias in Salamis and thus leave us with no clear evidence of a or as name such shrine inAthens, either by itself another for the Eury more our new sakeion.86 Moreover, and to point, Stroud presents significant s as as an evidence supporting Pritchett restoration well identification of was the Aiakeion (where the law says the grain to be brought, stored, and as sold) the large, nearly square structure by the southwestern boundary of as or the Agora, previously identified the Heliaia simply the Rectangular as Per?bolos.87 Textual details in the grain-tax law, Stroud shows, are suited to the chronology, architectural form, size, and location of this building. In short, it is hard to disagree with Stroud's summation that his evidence, while circumstantial, "is, however, better evidence than that used in support of all previous identifications of the building." His Aiakeion does lie south more in of the southerly Agora horos (Agora XIX, H26 [I 7039]) found to situ in 1967 (see Fig. 1), and, with reference that marker, he proposes that among the houses excavated a little south of the east-west line of the horos and northwest of the proposed Aiakeion was the confiscated house as referred to in the Attic Stelai being in Kollytos between the Aiakeion and the Agora.88

con was 84. Pritchett 1953, p. 276; his nia that it inMelite and by the dis Pritchett 1953, p. 276; cf. J.McK. of with on or near the southern of in junction Kollytos the southwest covery part Camp II Agora XXVIII, pp. 99-103, ad Kolonos of of for the Per?bolos as a part of the Agora has found other the Agoraios fragments Rectangular stelai inscribed with of the court of herents: Traill 1986, p. 126 and map; documents probable law; earlier, Thomp W. E. son Stroud 1998, pp. 86-90. Thomp genos Salaminioi and the phyle Aiantis 1954, pp. 33-39; Agora XIV, counter to son's (1970, p. 67) Pritchett, that include provisions for the erection pp. 62-65; Camp 1986, pp. 46-47; in see that Kollytos bordered the north side of of those stelai the Eurysakeion; Camp 1990, pp. 180-181. Melite near the northwest corner of the Ferguson 1938;Agora III, pp. 90-93; 88. Stroud 1998, pp. 94 (summa is and Pritchett's SEG excavated houses: Agora, unique ignores AgoraXlV, pp. 40-41,171,228; tion), 95-96, fig. 5; attention to the Ai[aK8tov] and Hime XLVII 146bis;Malouchou-Da?liana AgoraXIV, pp. 173-177; Camp 1990, . .. rioSS GT?VC?71?C Ko??dto?. 1998, p. 75, no. 324;ffiGXLVIII 264; pp. 56-58; it had not been certain n. on see that 85. Lewis 1955, p. 16, 40; for the genos Salaminioi, Parker heretofore the southerly (Agora see 49. No structure 308-316. I the two Eliot, Agora III, p. 1996, pp. 7039) of standing boundary or as stones of the marked its south has been found identified the 86. Stroud 1998, pp. 89-90; Wy Agora its loca 91. ern but that is a clear of Eurysakeion, but approximate cherley, Agora 111, p. limit, corollary seems 87. tion assured by ancient testimo Stroud 1998, esp. pp. 85-108; Stroud's conclusion. THE BOUNDARY OF MELITE AND KOLLYTOS 105

can now seen It be how the preceding scholarship, culminating with Stroud's argument, fits my proposition about the street boundary of as corner Kollytos and Melite it proceeds from the southwest of the Agora (see Fig. I).89 There Agora West Street at its southern end divides into two at branches the point of the northerly Agora horos (Agora I 5510), about 20 m southeast of the Tholos. The eastern branch continues south

by southwest, passing the southerly horos of the Agora (Agora I 7039) and becoming Young's Areiopagos Street, which continues around the west end of its namesake hill. The western of the two streets branching from the Agora diverges gradually from its eastern counterpart, heading as directly southwest Young's Street of the Marble Workers and passing Vanderpool's State Prison before turning westward up to the Hill of the as Nymphs Nymphs/Pnyx Street. If Stroud is right that the confiscated was one house in Kollytos of those excavated in the tongue of land formed two by the divergence of these streets, then the easterly of the two routes, as Young's Areiopagos Street, is ruled out the deme boundary, for it lies as a entirely within Kollytos.90 Tentatively then, the sole remaining road candidate for the deme boundary isYoung's Street of theMarble Workers, to which radiates from the Agora, neatly dividing Kollytos the east (with Stroud's confiscated house and Aiakeion) and Melite to the west with its Eurysakeion. seems The attentive fieldwork of Young to support the choice of one boundaries presented above. He suggested that of the streets of his was as a excavation chosen by Kleisthenes deme boundary?obviously that not between Kollytos and Melite. Although Young did commit himself to a a specific street, his later description of the stratigraphy at point in the Street of theMarble Workers is telling: "The westward branch of the Street use of the Marble Workers continued in from Archaic through late Ro man times. A cut made in its filling at the bend in its course [about 180 m southwest of the Agora] revealed six successive layers of road metal, of which the uppermost produced fourth century b.c. [sic] sherds, the lowest or a sherds of the late sixth early fifth century. Beneath lay layer of soft red use area as a earth which evidently predated the of this street and which contained sherds of the sixth century."91 This chronology strongly suggests was or that the Street of theMarble Workers laid, at least regraded, about to connect the time of Kleisthenes' reform the Agora with Nymphs/Pnyx Street.92 Furthermore, since the horoi of the Agora, as noted above, are themselves arguably Kleisthenic in date and purpose, it does not seem far to fetched suggest that the delimiting of the Agora with horoi, and of the

89. For a more detailed of tional Kleisthenic or the or cen plan dating post by the very late 6th early 5th the see Stroud was a area, 1998, p. 96, Salamis dating of Papadopoulos (2003, tury B.c., when there great in 5. use a crease on fig. p. 285), supplanted the earlier of in traffic roads to and from 90. Stroud 95 and much tract of land. Whether the west 1998, p. fig. 5. larger the and northwest region of 91. 162. route over as Young 1951, p. Nymphs/Pnyx the saddle of the city the Athenians created the 92. If the Street of theMarble these hills was a already full-fledged Classical Agora and shifted their Workers was first created at this time, road before this time is uncertain. At chief harbor from Phaleron to Piraeus the area must not have been well western pop this gap through the hills there (Papadopoulos [2003, p. 285] would ulated. This is not since was at a events implausible, probably always least footpath, date both after the Battle of the creation of the it a Classical Agora but certainly would have become Salamis). whether one nearby, accepts its tradi road for pedestrians and pack animals io6 GERALD V. LALONDE

demes Kollytos and Melite with the Street of theMarble Workers, which one were a exited the Agora at of its horoi, of piece in the Kleisthenic pro gram of political boundaries.93 was Further indirect evidence that this street the deme boundary lies in another entry in the confiscation records of the Poletai. The relevant pas . . .voto: sage from Agora I 1749 is ?pyaoxf|pia 8vo ?[jnME?irni] r\ ?[8oc f) ??r? too ?HpocK]I???oToi) A?^iKaKOD d? ?yofp?cv cp?po'oaa] ("two work . . . on shops inMelite bounded the south by the road leading from the Shrine of Herakles Alexikakos to the Agora").94 Wycherley concluded that was the road described here the extension of Young's Street of the Marble as we Workers up the south flank of the Hill of the Nymphs what have was near as called Nymphs/Pnyx Street, and that the Herakleion this road it passed between the Hill of the Nymphs and the Pnyx. As I have argued an s in earlier publication,95 Wycherley general location of the Herakleion can more on be further refined to agree with Pittakis's exact placement of it the upper tier of the northeast spur of the Hill of the Nymphs.96 Wycherley was certainly correct that Nymphs/Pnyx Street and the Street of theMarble a Workers were continuous segments of single road, for, after exiting the was Agora, the line of this thoroughfare unbroken, and it alone ascended the draw between the Hill of the Nymphs and the Pnyx.97 Moreover, by describing the road with reference to two of itswidely separated points, the Herakleion inMelite and the Agora, this official text shows that, although on was this road crossed Young's Piraeus and Melite Streets its route, it as an thought of integral line radiating from the Agora, and thus would a be logical deme boundary.98 we sure Since cannot be that horos B marked Nymphs/Pnyx Street a was as deme boundary, or that the Herakleion in the shrine of Zeus, to the burden of proof for this boundary shifts significantly the question was raised briefly above: In what deme the Pnyx? The significance of this question lies in the fact that Nymphs/Pnyx Street is the only exit road area from the city in the between the Hill of the Nymphs and the Pnyx. was or In other words, if the Pnyx not inMelite, surrounded by Melite, then Nymphs/Pnyx Street is the only road that could have been the south boundary of Melite.

in and in as 93. On the dating and purpose of mentioned proximity; (3) boundary, for, Young noted (p. 136, see of was over the horoi of the Agora, above, confiscation documents, the mention fig. 1), it built with houses and 10-11. the location is formulaic after entries of in the 5th and 4th centuries b.c. pp. 86-88, AgoraXlX, pp. For see Meritt real 94. Agora 11749, property. 98. The Agora, with itsAltar of the no. 95. Lalonde 86-93. is in ancient 1936, pp. 393-413, 10, Face A, 2006, pp. Twelve Gods, recognized col. 4, lines 105-109; Agora XIX, 96.Wycherley 1959, pp. 67-68; sources (Hdt. 2.7; IG IV 2640) and Pittakis 461. modern p. 114, P26, lines 450-454; although 1978, p. 187; 1835, p. scholarship (Agora XIV, no trace of the name Melite survives on 97. Young (1951, p. 161) recorded pp. 192-193; Wycherley 1962, p. 9; the its restoration is not a that at the where the Street of the as fragment, point Siewert 1982, p. 55) the hub from case of Marble Workers turned west to mount which main roads as it explaining ignotum per ignotius, radiated; such, is certain for a number of it was exclusive of all demes and but virtually the slope of the Hill of the Nymphs probably reasons: is had another branch as old as the Geo the inner (1) it the only urban deme therefore formed boundary name that fits the stoichedon pattern, metric period that may have continued of the urban demes that surrounded it; rest is between so presuming the of the restoration southward up the valley the Siewert 1982, pp. 56-57; Langdon correct; (2) Herakles Alexikakos, whose Areiopagos and the Pnyx. This branch, 1985, p. 12. was accounts in is not been a deme shrine by all Melite, however, could have THE BOUNDARY OF MELITE AND KOLLYTOS IO7

a It is long-standing tradition of scholarship that the Pnyx was in Melite. A first step in the argument against that tradition is to allow that at least the east and north slopes of the Hill of the Nymphs were inMelite. even Most modern topographers agree with this, though the southernmost certain monument of the deme is the shrine of Artemis Aristoboule founded

by Themistokles (Plut. Them. 22; Mor. 869C-D) and unearthed in 1956 about 150 m north of the summit and northeast spur of the Hill of the was was Nymphs (see Fig. I).99 But the belief that the Pnyx inMelite largely rooted in the further belief, prominent up to the mid-1950s, that Kolonos were area as Agoraios and Kerameikos demes, and thus that the ofMelite, one of the most populous of the urban demes, must have extended quite as far south.100 But Lewis, noted above, showed conclusively in 1955 that was a an area Kolonos Agoraios not itself deme but within the deme of now Melite, and it is widely accepted that the Kerameikos horoi marked a or not deme but either the public district of the Kerameikos the street adjacent to the horoi.101 In short, Melite extended much farther to the east was once and north than thought, although its northern boundary is not known precisely. A reasonable guess is that in the north it bordered the on deme Skambonidai the line of the Panathenaic Way from the Dipylon to at the Agora.102 In any case, Lewis's finding least allows the possibility as that Melite could accommodate its population without extending far or south as the Pnyx the Mouseion. now Despite the much greater extent allowed to eastern and northern were Melite, and despite the abandonment of hypotheses that concomitant as with the Pnyx's being inMelite, such the location of Otto Walter's Hera kleion and Homer Thompson's Thesmophorion, the long-standing belief was that the Pnyx within Melite has continued unchallenged.103 If this belief a hangs by any ancient evidence, it is the thin thread of scholion to Aristo a phanes,^^ 997 (Philochoros of Athens, FGrH 328 F122), text inwhich the generally reliable Philochoros is introduced to test claims of earlier a writers, but text that is still difficult to read and suspect by any reading:

[10] Oncji ?? KaAAioxpaxo? ?v Koac?vcoi av?on(i? ti e?vai . a?xon IaaxpoAoyiK?v . . ? ?? OiAolxopo? ?v Koax?vc?i u?v a?x?v ou??v Oe?vai ?iyei, ?ni Axj/eu?ou? [15] ?? (xou) 7ipo IIuGo?copou f|?,ioTpo7ciov ?v xfji vuv ouani ?KKA,na?ai Ircpo? xcoi xei%ei xcoi ?v xfji rivuK?. uT|7tox? o?v x? %a>p?ov (cpao? xive?) I?ice?vo 7tav, an

99. the demes and of Melite in envisioned a of the deme See Threpsiades and Vanderpool particular, however, strip see n. 78. 1965, pp. 31-33; the shrine's control above, Kerameis extending inward from the rests on 101. to con by the deme the discovery of AgoraXlX, pp. 11-13; Dipylon the Agora, which is a decree ofMelite (SEGXXll 116) Brueckner 1914, p. 91; Ohly 1965, sistent with his later argument (1999, b.c. its most the of about 330 honoring promi pp. 327-328; see, recently, p. 3) that the district and its horoi nent of were to some extent identified with the demesman Neoptolemos (APF lengthy argument Papadopoulos deme Kerameis. 10652), apparently for his restoration (2003, especially chap. 5) that the was of the shrine; see also Frost 1980, Kerameikos (Potters' Quarter) 103. Walter 1915, pp. 97-98, fol of pp. 184-185. originally north the Areiopagos lowed by Judeich (1931, pp. 396-397); 100. in see but moved northwestward with the cf. On the Pnyx Melite, Thompson 1936, pp. 156-192; of the Classical Young 1951, p. 140; Judeich 1931, founding Agora. Thompson and Scranton 1943, p. 295; n. 102. So on the location of the of p. 169 and 1;Wilamowitz-M?llen Langdon 1985, p. 12 and Thesmophorion n. n. see dorff 1880, p. 148; on the population of 21; Siewert (1982, p. 29, 140), Melite, Clinton 1996, pp. 123-125. io8 GERALD V. LALONDE

7U?piAO:|j?(iv?Tai Kai f] IIvTl)^,Ko?XOVO? ?oTW, ? ?T?pO? ? (I?G01O? I ??yo|i?vo? oikco? pipo? ti v?v cjwn6?? y?yov? to Ko?cov?v kccA??v to ?7iia0?v Ixr\q MaKpa? Zto?c?. ?XX' otjk ?axi- MzXur] y?cp a7iav ?K?ivo, cb? ?v to?? 'Opia[20]|io?? y?ypocTiTai ttj? 7to??co?.104

Kallistratos says that a certain astronomical monument in Kolonos ... is his [Meton's] but Philochoros says that he [Meton] set up nothing in Kolonos, but that in the archonship of Apseudes one a [433/2 b.c.], the before Pythodoros, he set up heliotropion in so some the present ekklesia by the wall in the Pnyx. And perhaps, say, that whole place, inwhich is included also the Pnyx, is Kolonos, now the other [Kolonos], the so-called hiring place. Thus it has a become customary to call certain part the Kolonon, namely the part behind the Makra Stoa. But it is not; for that whole thing is as Melite, iswritten in the boundary records of the polis. a This text, like many scholia, is cobbled together with little coherence from number of earlier commentaries, and modern attempts at emendation seem to an only complicate its problems. Foremost among the difficulties here is apparent confusion of the deme Kolonos, the Kolonos Agoraios, and per haps other Kolonoi.105 The scholiast allows the possibility of Kallistratos's claim that Meton set up an astronomical apparatus in Kolonos, but then he (the scholiast) indicates that he finds nothing in Philochoros to the or effect that Meton set up anything there, and he, perhaps Philochoros, a asserts that Meton actually set up heliotropion in the contemporary ekklesia by the wall in the Pnyx. The disagreement may have derived from a confusion of the Kolonos Agoraios with Kolonos Hippios, the deme a on of Sophocles, and tradition that Meton installed devices both the Pnyx and the Kolonos (perhaps Agoraios).106 In any case, this confusion a seems to have given rise to popular claim (line 16: cpoco?tw??) that as as Kolonos extended west and southwest of the Agora far the Pnyx. The

104. text is on The printed here that (to Ko?xovov)?in the second instance, suggests the possibility the basis of for emen the n. 106: tco Ko?xov a in the oiFGrH, except Dobree's dative (below, ) passage Monotropos (414 b.c.) ou to dation of 8T8po? in line 17, where leaves the gender uncertain. Since of the comic poet Phrynichos (Kock I retain ? erepo? of codex V. See also Kotaovov seems otherwise unattested, 1880, p. 376, fr. 21, probably derived and notes the recourse in a text as 1 to Ar. Av. and Jacoby's commentary perhaps only from Hypothesis the as this one is to view it as an a (FGrHlllb 1, pp. 496-497; Hlb 2, prickly didaskaliai) thatMeton installed on error tov or statue or pp. 401-403) this fragment of for KoAxovov. fountain astronomical where the Philochoros discussed several on Philochoros, difficulties of instrument the Kolonos Agoraios text are in the two in to his on the scholiast's made clear Kolonoi (and perhaps Attic addition device the Pnyx: greater detail. demes called Kolonai) in his third seeD?bner [1877] 1969, scholion to see 105. Note the scholiast's alternating book (FGrH328 F26; also Illb 2, Av. 997, p. 233: ?gc?? ?? ?v KoXcovq. use and nonuse of the definite article p. 401); cf.FGrHlllb 2, p. 402, for Ja Kprjvr|v uva KaxeoKeuaGaTO. (prjaiv on with KoAxov-, both here and further coby's idea that the subject of the scho ? Opuvixo? MovoTp?7i(p "xi? ?' ?axiv in the same scholion the horismoi have been ? (as quoted below lion and may jiexa xa?na xau?ric cppovxicov; M?xcov n. are in 106). Even if these variants regions of the city rather than demes; ? AeuKovoe??, ? x?? Kpr|va? aycov." of a to inconsistent vestiges distinction this nearly led him (p. 403) antici KaGe?xai?? Kai ? Movoxporco? ?nl to? sources auTot) in the scholiast's between pate Lewis's finding about Kolonos Xa?piou, ob? eipTixai. 'AMxo?, Kolonos as he Aris tco Kaxe the Kolonos Agoraios and Agoraios, considered that ?ggd? ?v Ko?covco Kprivnv Tiv? the this seems to have have meant Kolonos (Hippios) deme, tophanes may by (xrixavnoaTO, r\ ayaA,|ia, r\ ?vaGrijia on a a amco. been lost the scholiast. It also does quarter of the town and not deme. ?oTpoA,oyiKov KaTeaKEDaGaTO not neuter in line 18 106. Further same explain the article on, the scholion THE BOUNDARY OF MELITE AND KOLLYTOS 109

scholion, however, perhaps again deriving from Philochoros, contradicts this, clarifying that the Kolonos meant is the other Kolonos,107 the hiring area place, the behind the Makra Stoa (that is, Kolonos Agoraios and not area as as Kolonos Hippios), and that not only that but also far the Pnyx isMelite?assuming that there is any grammatical coherence here, and that the final ocTcav ?ice?vo refers not to ouxco? pipo? xi kx?,., but to the seems earlier ?Kx?vo 7iocv.108Thus this crabbed text rightly to imply what now to us is clear from Lewis's scholarship, that the Kolonos Agoraios was inMelite, but it is also the sole ancient testimony invoked by scholars our own era to up to that Melite extended south far enough embrace the root Pnyx.109 Moreover, Jacoby may have been right that the discussion in the scholion was not about demes at all. one a In any case, may readily put up against that scholion passage on of the sober, early, and learned Tertullian (De anim. 20) the supposed effects of environmental factors on the inherited character of local

peoples:

Nam et hic etiam de locis interest. Thebis hebetes et brutos nasci

relatum est, Athenis sapiendi dicendique acutissimos; ubi penes mense Colyttum pueri citius eloquuntur praecoca lingua.110

For in this matter even locale makes a difference. At Thebes it are at is said that the people born obtuse and irrational, Athens, very keen at comprehension and expression; and there in the deme are so Kollytos the children precocious in language that they learn to talk a month sooner.

errone Here, for the purposes of argument, we must accept the logical but ous Lamarckianism and interpret this remark about children in Kollytos as an allusion to the rhetorical influence of the Ekklesia. So Judeich understood was to it, but, believing that the Pnyx inMelite, he strained the passage mean was a area that Kollytos in main of traffic in the vicinity of the Pnyx.111 was was But surely Tertullian's point that the Pnyx in Kollytos or, if the was assembly-place, like the Agora, separate from the demes (the Pnyx

107. This is the no. no apposition point also Agora III, p. 21, 3, forWycher The story about Kollytos survives note stoa was of keeping theMS reading ? erepo? ley's that this probably where else in the ancient tradition of instead of Dobree's ou in line on erepo? "one of those along the road from the environmental influences the soul 17. latter in to on mense The phraseology, making Dipylon the agora, the south, and mind. Some translators read as as mense as com ? uiaO?o? Kotaovo? well the Pnyx where [in accord with the scholion to of citius the ablative of of another seems a part Kolonos, unduly Ar. Av. 997] it would have the Kolonos parison ("sooner than month"), but complicating. Agoraios behind it." my translation of the ablative of degree 108. The distinction of this Kolonos If 109. E.g., Dunbar 1995, p. 554. of difference may be preferred for as the "the so to statement Agoraios by qualifier there is need emphasize further the making the less outlandish. called is confirmed the on I hiring-place" by unreliability of scholiasts demes, thank Ioannis Evrigenis for providing scholia and one not a me lexicographies apparently need look far, for scholion (to with Waszink's text, and Molly based on earlier commentators who is to Aeschin. 1.125) also the sole direct Richardson for calling my attention noted reference was some Hyperides' (Blass 1894, testimony that Kolonos Agoraios of the textual difficulties. fr. to as see a deme. 8) hirelings Kolonetai; Ill.judeichl931,p.l69,n.l. no. Agora III, pp. 91-92, 251; p. 99, 110. Edition of J. H. Waszink, Less likely isWaszink's notion (n. 110, no. Lambert n. 11. 1947. context as an 286; 1997, p. 295, Amsterdam Tertullian probably above) of Kollytos in this For location of the Makra Stoa "in borrowed the of Thebans to comparison allusion Plato's deme and birthplace see Kerameikos," IG IV 968, line 14; and Athenians from Cicero (Fat. 4.7). (Diog. Laert. 3.3). no GERALD V. LALONDE

did have its own horoi; e.g., IG V 1092),112 largely embraced by Kollytos. Otherwise, the saying should have been about children inMelite. we Before quitting the subject of Melite and Kollytos, return to the on simple horos B (?opo?) the northeast spur of the Hill of the Nymphs and one a consider other possible function, that is, that horos B may have been deme marker in a sense apart from the official polis-wide program of Kleis thenic divisions at the end of the 6 th century b.c. The demesmen ofMelite themselves may have cut horos B to emphasize to all visitors to the shrine of Zeus, and thus preclude any dispute, that this t?menos and all those northward to its other boundaries were within the control of the deme.113

Subsequent to, and independent of, this hypothesis, Eugenio Lanzil a lotta published remarkably parallel idea, namely, that the shrine at the summit of the Hill of the Nymphs (Fig. 1), and therefore also just north of not to the proposed deme boundary ofMelite, did belong the Nymphs and Demos, but that its rupestral horos (IGV 1065: hiepov IN\)|i(p[?]v A?|Lio), a roughly contemporary with horos B, meant "shrine of Nymphs of the a deme," with the intent of asserting territorial claim.114 Although horos B lacks the word ??jLio, the proximity of the horos of Zeus may have helped to clarify Melite's territorial purpose in the simple marker. near Melite's purpose in marking its territorial claims at key shrines was its boundary with Kollytos very likely to emphasize its possession of resources. these prestigious and valuable religious Unlike rural demes, which as as had land for plowing, pasture, and minerals well religious ternene, the resource inner urban demes had their greatest public in cults. At least four or other shrines of specific gods heroes besides those of the Zeus of horos are to A and the Nymphs at the summit of their namesake hill known have been inMelite, and three of these shrines (Fig. 1: the Eurysakeion, the Hephaisteion, and the temple of Artemis Aristoboule) have been discovered or are otherwise convincingly located north of the deme boundary proposed here. The fourth is that of Herakles Alexikakos, probably located, where on Pittakis first suggested, the northeast spur of the Hill of the Nymphs, case within, or adjacent to, the precinct of Zeus.115 In the of the shrine of a was Artemis, although its founder Themistokles, resident of Melite, not a member of the deme, there is direct epigraphical evidence that by about 330 b.c. the precinct was administered by theMelitean demesmen.116

ca. nosce e 112. Horos of the Pnyx, mid la Kr?n [1979], tuttavia appare should add that Parker earlier (1996, see noted that the combination of 5th century b.c.; Trzvlos, Athens, preferibile la seconda, per la consolidata p. 233) was unusual and pp. 466-467, fig. 588. attestazione dei culti del demo." Lanzil Nymphs and demos 113. that the shrine not have Cf. Traill's somewhat analogous lotta argues that the Nymphs and their suspected may concern been but owned the demos suggestion (1982, esp. p. 168), that the with fertility and childbirth "of," by, in common the of the precise division by inscribed horoi of have nothing with poli (i.e., Athenians). or Erechtheid andAntigonid Lamptrai tics of the Athenian demos the per 115. Pittakis 1835, p. 461; also was is no Lalonde 86-93. necessitated by the establishment sonified Demos, that there parallel 2006, pp. of theMacedonian phylai of Athens in for such a connection, and that Demos 116. See above, p. 107, n. 99. Since b.c. as a of Artemis was founded the late 4th century heroized cult figure, often and ap this shrine 499: "che with or the is after the institutions of 114. Lanzillotta 2000, p. propriately Grace Graces, Kleisthenes, the and Themistokles was a lo iep?v sia dedicate alle Ninfe del po attested only from 4th century on; strong sup o alle Ninfe del inteso come see in of demotic the shrine p?lo demo, also P. Gauthier BullEp 2001, porter politics, le no. 149.1 R. well have come under deme ripartizione territoriale. Ambedue p. 505, thank S. Stroud for may sono come to attention. control from its letture accettabili, pure rico calling this article my We beginning. THE BOUNDARY OF MELITE AND KOLLYTOS III

Two other shrines very likely inMelite that have been excavated but remain are on anonymous the small hieron discovered by Thompson and Scranton the northwest slope of the Hill of the Nymphs and, further to the north, on the apparent shrine the site of the old Church of Agios Athanasios on Kourkouris (Fig. 1) Heptakhalkou Street, just west of the Theseion Metro Station.117 as The dating of horos B and that of the Nymphs' shrine, inter some preted above, suggest that the deme Melite had early interest and authority regarding the cults and shrines within its boundaries. Some of this authority probably originated with Kleisthenes' reforms, but there is also evidence that other demotic control evolved later. The role of the

phratriai in the Apaturia and Thargelia festivals and in the cults of Zeus as as Phratrios and Athena Phratria, well details of the cultic activity of some phratries and their subgroups, give credence to the Aristotelian claim matters to (Ath. Pol. 21.6) that Kleisthenes left sacrificial the phratriai and gen? in accordance with ancestral customs,118 but, at least for the reason some that of the old gentilitial groups would have had territorial roots that were or outside overlapped the deme boundaries, the demarchy from its must some inception have had of the administrative role in religion that we see in later evidence, especially that of the 4th century.119 In state decrees as well as those of the demes there is evidence that

the demarchs religious duties gradually increased, often not replacing but we merging with those of older cult functionaries and laws. For example, a read of such merger in Demosthenes 57.46-48, 62, where the priest of Herakles in the deme Halimous is said to be appointed by lot from a list of "best-born" (enyev?oxaxoi) candidates nominated by the deme a assembly. IG II2 204, state decree of 352/1 B.c., orders (lines 16-23) that from then on, various polis officials "whom the law prescribes in each instance" and demarchs should see to the permanent care of all shrines. A decree of the demesmen of Piraeus (/Gil2 1177) stipulates that the in cumbent demarch along with the priestess is to see to the proper use of the to deme's Thesmophorion and impose fines for illegal wood-gathering

117. and Scranton see Lambert Thompson 1943, post-Kleisthenic survival, political coalitions by undermining the B. For the Church of here p. 381, app. Agios 1998, chap. 6; Lambert clearly lays religious institutions that held them see Mommsen to rest the a view that relied on Athanasios, [1868] idea ofWade-Gery (1933) together, heavily n. as 1977, pp. 49-50, 47; Biris 1940, that the phratry survived the reli Aristotle's theoretical rule (Pol. 1319b), no. Travlos X some p. 51, 90; 1960, pi. gious antipode of the secular deme. with reference to Kleisthenes, Some of the arched Note (following p. 192). particularly pp. 224-225, for that "private religious rites should be and niches cut into the Lambert's consideration of the to a rectangular possi reduced few public ones." facets of the rock on which stands the that the Pol. meant in bility Ath. part 119. On the subject of religion in old of Hosios Athanasios that the Kleisthenic reforms had the the I have chapel may demes, benefited especially be the work of devotees of the Chris indirect effect of some transferring from the following: Mikalson 1977; tian but I thank Kevin holy man, cults of the phratries and their sub Osborne 1985, pp. 154-189; White Glowacki for to see (pers. comm.) sharing groups the demes; also Kearns head 1986a, pp. 176-222 (in general), me most with his opinion that of the 1985, pp. 204-205; 1989, pp. 755-777; 163-169 (cult finance), 127-128,134 niches are ancient and that one of them andWhitehead 1986a, pp. 176-222. 137 (religious duties of the demarchs); has a in the bottom surface for a as setting This scholarship, along with Lambert Whitehead notes (1986a, p. 176, a common feature of Classical counters the earlier view n. the loci classici on deme tenon, 1998, chap. 8, 1), religion, votive niches. for Lewis are (see, example, 1963, p. 37) especially sacrifice, Arist. Eth. Nie. 118. On phratry religion and its that Kleisthenes broke up the old local 1160a.l4-25, and PI. Lach. 187d-e. 112 GERALD V. LALONDE

to on according "the ancient laws the subject."120 The demarchy, from the time of its institution, certainly would have been responsible for seeing that only those eligible partook of the hiera of deme festivals and sacri or fices those of the polis conducted by and within the deme. Keeping in mind that the evidence of deme inscriptions is mostly scattered and late, one and that every internal practice of deme may not be that of the others, we extant as may still cite these isolated documents suggestive of wider practice. over The suggested territorial imperative of the deme its local shrines, was not cults, and festivals motivated only by interest in divine favor and the political prestige of these religious institutions, but also by what these institutions did for the deme's economy. Whitehead neatly summarizes cost the deme's interest in religious finance thus: "The of cult?upkeep of temples and shrines, offerings of regular sacrifices, celebration of recurrent festivals?surely represented, for any deme, the major object of regular as a expenditure, and indeed the fundamental raison d'?tre of the budget extant sacrae whole."121 Evidence of this is the fact that the leges and fasti sacri of Attic demes are less the rules and schedules of ritual than accounts run of the personnel and finances of sacrifices and offerings, accounts that into the hundreds of drachmas annually for single demes.122 For the 59 sacrifices listed in the sacred calendar of the deme of Erchia (SEG XXI 541) the total cost is approximately 547 drachmas, and all but five of those sacrifices took place in Erchia itself. A decree of about 400 B.c. from the deme of Plotheia lists cult outlays, including the demarchs budget, totaling an amount so was not 22,100 drachmas, large that it certainly expenditure but rather the total capital investment in loans and the leasing of deme lands, the interest from which would have supported the deme's cult budget.123 a Melite, being heavily populated inner urban deme, would, unlike Plotheia, have had little real property to lease. Like Erchia, itwould have raised religious funds through leitourgiai and other forms of taxation, but, a being in the heavily populated inner city with large traffic in sacrifices, numerous some its shrines would themselves have generated of the funds for their own function and maintenance as well as material benefits for was a the deme. For example, if the demarch himself making sacrifice, at was duty attested in least six deme inscriptions, he often in charge of meat sacrifice was distributing the of and the recipient of perquisites (gera) such as the skins of the sacrificial animals.124 While the great part of these extant records concerns the deme's own on more expenditures shrines and rituals, beyond these, and to the point

120. For elaboration of this and n. 39, for a summary of the debate sacrifice, see Sokolowski 1954; Burkert other seeWhitehead in evidence, 1986a, about the meaning of the figures this 1985, pp. 95-97; Rosivach 1994, notes. pp. 127-128,134-137, with decree; Haussoullier (1884, p. 64) esti pp. 68-142. For the extent and direc 121.Whitehead 163-164. mated from these investments a tion of see 1986a, pp. yield deme cult expenditures, or 122. On the sacral calendars and of 12 percent 1,812 drachmas after Mikalson 1977, pp. 424-435. Dow laws as see practical accounts, Jameson discounting capital expenditure for the (1965, pp. 206-207) suggested that in 1965, pp. 155-156; for a summary shrine of Herakles. the Erchia calendar (SEGXXI 541) of the sacred calendars of the to study 124. SeeWhitehead 1986a, p. 128 the unusual granting of gera the Attic see Verbanck-Pi?rard and nn. for a of the lines demes, 40-49, summary k?ryx (col. E, 54-58) KaGarcep 1998. ritual duties of the demarch and their ? ?rijiapxo? meant that the demarch 123. On the various of was See Finley 1951, pp. 284-285, financing. expenses the normal recipient. THE BOUNDARY OF MELITE AND KOLLYTOS 113

or on of popular cults and shrines like those of Zeus the Nymphs this hill inMelite,125 individuals and small groups must have spent, and offered in their sacrifices, a great deal that has gone unrecorded but might be loosely extant sacrae projected from votives, analogous inventories, and leges of the case polis and demes. In the of these "private" sacrifices, the deme itself probably profited from sale of victims, bloodless offerings, and wood fuel, to the sale of the services of mageiroi slaughter and cook sacrificial animals, the collection of shares of the sacrificial meat and bloodless offerings, the gera, and, at least relatively late in the evolution of Greek cults, the receipt of monetary fees and other gifts. Obviously such expenditures by the in dividuals and small groups would have been more modest on the average than those of the deme and other large corporations, but in the aggregate they must have been considerable.126 was a The deme ofMelite always territory bordered by other territo to rial demes, but in the short period from the Kleisthenic reforms the mid-5th century it became also the neighbor of great polis institutions as such the Agora, the Pnyx, and the Themistoklean city wall. Although to it probably took pride in its proximity the central greatness of Athens, as a Robert Parker has said, "in many respects the religious life of deme can seen as a on a be that of mini n?Xiq, closely comparable small scale to own that of Athens itself."127 In this sense, Melite's communal highlight a as deme must have been its collective shrines and cults, and it is quite plausible, though admittedly uncertain, that the purpose of rupestral horos was B and that of the shrine of the Nymphs to emphasize its territoriality, resources thereby distinguishing its religious from those of Kollytos and as a marking its identity polis within the polis.

THE EXTENT OF MELITE

Although the foregoing discussion has not involved all the boundaries of Melite and Kollytos, it has included enough information about the other boundaries of Melite that itmay be useful to present the evidence for the full configuration of that deme. Lewis's definitive striking of Kolonos Agoraios from the list of demes showed that the eastern border ofMelite was virtually coextensive with the west side of the Agora. Without any hard evidence, but again with the premise that major roads radiating from the as city center served boundaries of the inner urban demes, a good tentative placement of Melite's northern limit, perhaps bordering Skambonidai, is or the line of the Panathenaic Way from the Agora out to, through, the an Dipylon Gate.128 Langdon's preference for outer border of the central

the shrine of Zeus would victims for sacrifice is 125. Elsewhere (Lalonde 2006, purification, public applic I have have not to pp. 40-80), argued from dedica been frequented only by able sacrifices by small family groups tions to Zeus Meilichios found in the demesmen, but also resident and visit and individuals; for fees and contribu to see vicinity of theHill of the Nymphs and ing citizens of other demes, metics, tions paid cults, LSCG 69, lines even theAgora that this popular god of foreigners, and perhaps slaves. 20-24; Petropoulou 1981, pp. 53-54; purification and individual and familial 126. Much of the research of Ro Sourvinou-Inwood 1988, pp. 263-264. was on eco welfare the chief deity of the shrine sivach (1994, pp. 68-142) the 127. Parker 1987, p. 137. a marked by horos A. Having cult of nomics of supplying and acquiring 128. Siewert 1982, p. 42. 114 GERALD V. LALONDE

some urban demes at the Themistoklean city wall may correspond to fact in a to cases, including that of Melite,129 but general rule that effect raises questions of chronology and purpose. Even if the inner urban demes took their territories and names from what were once separate communities, the streets as topography of Athenian and natural features was, noted above, to complex enough at the end of the 6th century for these divisions have been made by Kleisthenes himself. were We have no evidence that the city deme boundaries reset in 479 new new not B.c. to match the city wall, and, to be sure, the wall could have was been made to fit political boundaries, for its purpose defense, and the a or main criteria in siting defensible city wall inArchaic Classical Greece were as manpower and terrain. Still, suggested above, Melite may have one was been of those urban demes whose outer boundary coincidentally on or close to the Themistoklean circuit. I hinted above, in the discussion outer extent of the Melitides Gate, at the difficulty of determining the of the proposed southern boundary of Melite southwest of the saddle as a between the Hill of the Nymphs and the Pnyx. The Diateichisma, construction of the 4th century or later, is out of the question, unless that on wall coincidentally sat the deme boundary. The limited evidence of the area to original Themistoklean city wall in the between the Long Walls out Piraeus suggests that it looped considerably farther southwest than the Diateichisma.130 If the southern street boundary of Melite went that far as as beyond the crest of the hills, it probably bordered Koile well Kollytos on the south. we to Although cannot be precise about how farMelite extended the a western was west, there is complex of evidence that its flank contiguous area with the deme Keiriadai in the shortly west and northwest of the at as as Hill of the Nymphs. One testimonium, from least early the 2nd A.D. century (L?xica Segueriana), associates Keiriadai with Athens's place as a of execution, describing the barathron pit in Keiriadai into which the Athenians cast those condemned to death.131 Some Classical writers bear was out the implication of this phraseology, that casting into the barathron was not always just to exhibit the corpse and deprive it of burial, but also was a sometimes a mode of execution, and that the site therefore rocky some sources precipice of kind.132 Two other ancient bring into closer focus the location of the barathron and thus the approximate boundary of Melite and Keiriadai. Plutarch wrote (Them. 22; Mor. 869C-D) that near Themistokles founded the temple of Artemis Aristoboule his house in Melite, "where now the ofijiioi throw out the bodies of those put to death." Since this Artemision has been excavated and securely identified (Fig. I),133

11. s.v. and For the Kai 129. Langdon 1985, p. ?apaOpov ?puyjia. (a7T?KX?ivav 7tavxa? ?? (p?payya in see RE the historian 130. See Judeich 1931, plan LB-C/ barathron general, II, 1896, EGE?a^ov)?unless s.v. Either 5-6, where, however, many of the lines col. 2853, ?apaOpov (T. Thalheim). intended hysteron-proteron. are Hell. ??v in tentative. 132. E.g,Xen. 1.7.20: sequence might be implied the s.v. common one over 131. Bekker 1814, p. 219, KaTayvcooBfj ??iKe?v, a7io6ave?v ei? t? idiom for handing xi to the xa> BapaGpov: A6rjvr|Gi ?? r\v ?puyjia ?apaOpov eu?^n?evTa. Thucydides executioner, rcapa?o?vai v term xot) ?v Keipia? ?r||iq) xr)? Oivni?o? (pvXri? (2.67), using the less parochial ?7ii ?p?yjaaxoc (e.g., Lycurg. 121; [Thalheim (see below): vielmehr xfj? cpapayya, indicates that the corpses of Din. 1.62). n. I7t7io6ocovxi?o?], ei? ? xo?? ?7il Oav?xco heinous criminals executed in other 133. See above, p. 107 and 99. were cast into the barathron KaxayvcoG0?vxa? ?v?ftaXXov. Cf. Harp., ways also THE BOUNDARY OF MELITE AND KOLLYTOS 115

we can be quite certain that Plutarch meant that they cast the bodies out ruts beyond the city gate that Judeich postulated from wheel running per m pendicular to the line of the Themistoklean city wall about 240 south of the Piraeus Gate and just northwest of the Hill of the Nymphs. Judeich as tentatively identified this gate (Fig. 1) the Anjaiai nvXax ("Executioner's Gate") mentioned by Hesychius.134This location fits neatly the other piece a scene in of evidence from Classical literature, paradigmatic Plato (Resp. on 439e), where outside the north Long Wall the road from Piraeus, Leontios feasts his eyes on the corpses of executed criminals.135 Plutarch's use of the word "now" (vuv), his mention of Melite rather than Keiriadai, and the different terms, ?puyjaa and ?apa?pov, should not, I think, lead some see us, as it has late antique and modern writers, to two successive an one a one execution places, early in Keiriadai and later inMelite.136 If use of the barathron survived to Plutarch's time?and there is little evidence use seems more now of that beyond his of vuv?it likely, especially with our knowledge of the exact location of the temple of Artemis Aristoboule, was that Plutarch simply being elliptical in his observation, meaning that was Themistokles' house inMelite by the temple of Artemis, and that the was not barathron much farther out, but beyond the "Demian Gate" and in Keiriadai.

If the public exposure of corpses was, in addition to sheer punishment, an expression of civil authority intended to deter crime, then the barathron would have been located, as the story of Leontios suggests, near the well a frequented road leading to city gate. As Stroud has shown with regard to the transportation of grain from Piraeus to the upper city, this road was or outside the north Long Wall the hamaxitos (Fig. 1), carriage road, was most which the level and certainly the most heavily trafficked route. It avoided the western hills of Athens and divided into several branches

that entered the city through various gates, including the Demian, in the northwest section of the city wall.137With all this evidence, it remains only to a look in the vicinity for suitable rocky precipice. The south to west sides are of the Hill of the Nymphs quite precipitous, but the construction of area the north Long Wall (ca. 460 B.c.) would have made that inaccessible to or anyone coming up the hamaxitos from Piraeus going down that road more through the Demian Gate. A larger but less steep rocky declivity than

134.Judeichl931,p. 140 and northern part of the Themistoklean the women in the rites of the eastern I:C 5 wall. or to plan ("Henker Thor?"); Travlos, city Mother prophesy that Demeter 136. for was to Athens, p. 159; Hesychius (s.v. Aru_iiaGi See, example, Thalheim coming look for Kore; and is the source for a of n. 7c6Xai?) only gate (above, 131); late ancient testimonia the absurdity that the Bouleuterion/ to are to was on this name, but he associates it only with this view the scholion Ar. Plut. Metroon erected the site of s.v. as prostitutes and denies that Ar|(iiaGi is 431 and Suda, jirjxpay?pxrj?. These this old barathron the shrine of the a mistake for The relation of sources in addition to the barathron Aio|ifJGi. see, Phrygian Mother (see also Poll. 3.11). a was the shrine of Artemis and, approxi of historical sources, primitive %aG|-ia If there any relocation of the bara 150 m a road to the ... was mately along cppeaxco?e? that the Athenians filled thron, it likely intended to keep its the Demian Gate can be in as atonement for a southwest, partial killing pollution outside the city, and therefore seen in Travlos, Athens, pp. 168-169, Phrygian priest of theMother goddess the move would have been in the oppo Gate I and reference 244. there. This has all of the site from an earlier out fig. 219, story trappings direction, place 135. Since Leontios's was to a journey of etiological folklore: the grisly details side the Archaic city wall farther from Plato's vnb xo that this was like a dark well Piraeus, phrase 6pi)y|ia place beyond the Themistoklean wall. to at ?opeiov xe?^o? ekxo? clearly refers and had hooks the top and bottom; 137. Stroud 1998, pp. 104-107, not come to the northern Long Wall, and the that the Phrygian had initiate fig. 7; Papadopoulos 2003, p. 287. ii6 GERALD V. LALONDE

m 100 west of the Hill of the Nymphs has been identified by several early as modern topographers the barathron, and it fits the evidence well enough, being quite visible from the hamaxitos from Piraeus and not far outside the Demian Gate.138 So, to relate this evidence to the deme boundaries, it indicates that the dividing line of Melite and Keiriadai would have been near, but not necessarily on, the line of the Themistoklean city wall. can a We close, then, with brief recapitulation of the approximate was an configuration and extent of Melite. Its shape roughly that of isos one celes trapezoid with its nonparallel sides diverging from the Agora, from its southwest corner over the the other from Nymphs/Pnyx saddle, 138. Curtius and Kaupert 1878, corner its northwest out the Panathenaic Way. The shorter parallel side p. 18;Wachsmuth 1874, pp. 349-350 was and contiguous and practically coextensive with the west side of the Agora, plan (end page); 1890, p. 265; was Judeich 1931, LB-C 5. The while the longer western side perhaps roughly parallel and proximate plan drop off of the terrain here is still visible to the Themistoklean wall but not on its line. All of this city necessarily even beneath the construction of the to area, with the of the road the Gate, would fit Dem not at exception Dipylon modern city, and it is odds with osthenes' observation that the drunken son of went (54.7) Ktesios, Konon, the term opuyuct, as it may have been "from the Agora up toward Melite." an old quarry.

REFERENCES

= The Athenian Results and trans. Agora Agora: of Archaic Classical, J. Raffan, Excavations Conducted Ameri by the Oxford. can Classical Studies at School of Camp, J.McK, II. 1986. The Athenian Princeton Excavations in Heart Athens, Agora: the of III = R. E. Wycherley, Literary Classical Athens, London. and 1957. -. 1990. The Athenian Epigraphical Testimonia, Agora: = H. and to XIV A. Thompson A Guide the Excavation and R. E. Wycherley, The Agora of Museum, 4th ed., Athens. Athens: The and Uses -. 1994. History, Shape, "Before Democracy: an Ancient 1972. of City Center, Alkmaionidai and Peisistratidai," XIX = G.V. M.K. in and Lalonde, The Archaeology ofAthens and M. B. Attica under Proceed Langdon, Walbank, the Democracy. Poletai an International Inscriptions: Horoi, Records, ings of Conference Leases Public 1991. Years since of Lands, Celebrating 2,500 the Birth XXVIII = A. L. et Held in Greece at the Boegehold al. ofDemocracy, The Lawcourts at Athens: Build American Classical Studies at Sites, School of and December ings, Equipment, Procedures, Athens, 4-6,1992 (Oxbow D. Coul Testimonia, 1995. Monograph 37), ?d.W. E. Andrewes, A. 1977. "Kleisthenes' son et al., Oxford, pp. 7-12. Reform 241-248. -. Bill," CQ 27, pp. 2001. The Archaeology of APF = K. ]. Davits, Athenian Proper Athens, New Haven. tied Families, 600-300 b.c., Oxford Charitonides, S. 1.1961. "AvaaKaipfj 1971. K?,aootKcov Tacpcov rcap? rrjv E. 1814. An?cdota Graeca 1: Bekker, 7i?,aTe?av luvTayuxxTOC," ArchEph L?xica Segueriana, Berlin. 1958 (1961), pp. 1-152. Ai tcov Biris, K. 1940. 'Eiac?ri

on Ancient Greek -. 1995. "Der der Cult, Organized H?gel Panagia Sacrificial Calendar from Erchia," Institute at und by the Swedish Athens, Thiti bei Vari seine Inschrif BCH 89, pp. 154-172. AMI 16-18 October 1992 (SkrAth 8? 14), ten," 10, pp. 235-246. Jeffery, L. H. 1948. "The Boustrophe ed. R. Ill? -. 2001. and H?gg, Stockholm, pp. Athens, Attica, the don Sacral Inscriptions from the 125. 86-111. Megarid:An Archaeological Guide, Agora," Hesperia 17, pp. = rev. London. N. F 1999. Associations CorinthXVU1.3 N. Bookidis and English trans, ed., Jones, The of R. S. The A. W. 1933. The Classical Athens: to Stroud, Sanctuary of Gomme, Population The Response Demeter and Kore: and in and Oxford. Topography ofAthens theFifth Fourth Democracy, Princeton 1997. Centuries Oxford. W. 1931. von Architecture, b.c., Judeich, Topographie M. 1951. "The Poros M. 1974. 3: 2nd Munich. Crosby, Building," Guarducci, Epigraf?a greca Athen, ed, 168-187. di Rome. E. 1985. Hesperia 20, pp. Epigrafi carattereprivato, Kearns, "Change and Conti E. 1868. Sieben Karten zur M. H. in after Curtius, Hansen, 1976.Apagoge, Endeixis, nuity Religious Structures von Gotha. and in Crux: in Topographie Athen, Ephegesis against Kakourgoi, Cleisthenes," Essays and A. 1878. and A in Greek Presented to M. Curtius, E., J. Kaupert. Atimoi, Pheugontes: Study History G E. Atlas von Berlin. the Athenian Administration de Ste. Croix onHis 75th Athen, ofJustice Birthday, 1892. AM in the Fourth b.c. P. F D?rpfeld,W. "Funde," 17, Century (Odense ed. A. Cartledge and D. Har pp. 90-95. University Classical Studies 8), vey, London, pp. 189-207. S. 1965. "The Greater Demarkhia Odense. -. 1989. Heroes Dow, The ofAttica of BCH 180-213. -. 1986. and Democ Erkhia," 89, pp. Demography (BICS Suppl. 57), London. Number Citi Diibner, F. [1877] 1969. Scholia Graeca racy: The ofAthenian Kirsten, E, and W. Kraiker. 1967. in Hildesheim. zens in Fourth Ein zu Aristophanem, repr. the Century b.c., Her Griechischelandkunde: F?hrer N. Dunbar, 1995. Aristophanes, "Birds," ning, Denmark. klassischen St?tten, 5th ed, Heidel M. et al. Oxford. Hansen, H, 1990. "The De berg. 1962. Coastal Demes Eliot, C. W. J. of mography of theAttic Demes: The Kock,T. 1880. Comicorum Atticorum Attika:A in Kleis Evidence of the Comoediae Study thePolicy of Sepulchral Inscrip Fragmenta l'.Antiquae thenes Toronto. AnalRom (Phoenix Suppl. 5), tions," 19, pp. 25-44. Fragmenta, Leipzig. M. 1959. "Geraistai E. M. 1997. "A New Solution U. 1872. "Der in Ervin, Nymphai Harris, K?hler, Areopag Genethliai and theHill of the to the Riddle of the Seisachtheia," Athen," Hermes 6, pp. 92-112. UXaxcov 146 in The Polis in S. N. 1984. > Nymphs," 11, pp. Development of the Koumanoudis, "Perhaps > 159. Archaic Greece, ?d. L. G. Mitchell Horos 2, " Usually Certainly," Faraguna, M. 1997. Regis trazioni and P. J. Rhodes, London, pp. 103 pp. 71-81. catastali nel mondo Il caso 112. greco: Kr?n, U. 1979. "Demos, Pnyx, und di Athenaeum 7-33. E. 1890. and Atene," 85, pp. Harrison, J. Mythology Nymphenh?gel: Zu Demos W. 1938. Monuments Ancient zum Ferguson, S. "The Salaminioi of Athens, Darstellungen und ?ltesten of Heptaphylai and Sounion," Hes London. Kultort des Demos in Athen," AM 1-74. B. 1884. La vie 49-75. peria 7, pp. Haussoullier, municipale 94, pp. FGrH = F. Die der en Essai sur des V. Hero in Jacoby, Fragmente Attique: l'organisation Lalonde, G. 1980. "A Shrine Berlin 1923 d?mes au Paris. the Athenian griechischen Historiker, quatri?me si?cle, Agora," Hesperia 49, 1930; Leiden 1940-1958. Heisserer, A. J, and R. A. Moysey. pp. 97-105. Fine, J. V. A. 1951. Horoi: Studies in 1986. "An Athenian Decree Hon -. 2006. Horos Dios: An Athenian Real and Land and Cult Zeus Mortgage, Security, oring Foreigners," Hesperia 55, Shrine of (Monumenta Tenure inAncient Athens 177-182. Graeca et (Hesperia pp. Romana 11), Leiden. Suppl. 9), Princeton. Henry, A. S. 1983. Honours and Privi Lambert, S. D. 1997. Rationes Cente M. 1.1951. Studies in Land inAthenian Hildes simarum: Sales Public Land in Finley, leges Decrees, of and Credit inAncient 500? heim. Athens, Lykourgan Athens (Ap%aia EMla?: 200 b.c.: The -. 2001. "The on Horos-Inscriptions, Sigma Stigma," Monographs Ancient Greek New Brunswick. 93-105. and ZPE137, pp. History Archaeology 3), F. 1980. S. C. Frost, J. Plutarch's ((Themis Humphreys, 1980. "Family Tombs Amsterdam. tocles":A Historical and in -. 1998. Commentary, Tomb Cult Ancient Athens: The Phratries ofAttica, Princeton. Tradition or Traditionalism?" JHS 2nd ed, Ann Arbor. -. Gallo, L. 1999. "Solone, gli hektemoroi 100, pp. 96-126. 2004. "A House of the Pirae e n.s. gli horoi," AION, 6, pp. 59-71. Hunter, V. 1998. "The Prison of Ath ans," ZPE146, pp. 91-92. G. 1904. La solidarit? de la ens: A Glotz, fa Comparative Perspective," Langdon, M. K. 1985. "The Territorial mille dans le droit criminel en Phoenix 296-326. Basis of the Attic Gr?ce, 51, pp. Demes," SymbOslo Paris. IE2 = Iambi et elegi graeci anteAlexan 60, pp. 5-15. H. R. 1994. "Neue attische drum 2nd 2 -. 1999. Goette, cantati, ed., vols, "Hymettiana III: The Klio 120 ed. M. L. Oxford 1989-1992. Felsinschriften," 76, pp. West, Boundary Markers of Alepovouni," 134. M. H. 1965. "Notes on the Jameson, Hesperia 68, pp. 481-508. Il8 GERALD V. LALONDE

Lanzillotta, E. 2000. "Il culto del demo Mikalson, J. D. 1977. "Religion in the in Atene," in 'Emypa?a? Miscella Attic Demes," AJP 98, pp. 424 nea in onore di Lidio Gas 435. epigr?fica ed. G. 495 A. 1885. in Denk perini, Paci, Rome, pp. Milchh?fer, "Athen," 501. m?ler des klassischen Altertums zur H. 1982. "Zwei Horos-In des Lebens der Griechen Lauter, Erl?uterung Vari: und R?mer in Kunst und schriften bei Zu Grenzziehung Religion, und Demenlokalisierung in S?dost Sitte, ed. A. Baumeister, vol. 1, Attika,"^ 97, pp. 299-315. Munich, pp. 144-209. Lauter-Bufe, H., and H. Lauter. 1971. Mommsen, A. [1868] 1977. Athenae "Wohnh?user und Stadtviertel christianae, repr. Chicago. AM 0.1980. des klassischen Athen," 86, Murray, Early Greece, Brigh pp. 109-124. ton. K. evo 1965. Lazaridou, 2002. "Epyaoie? Ohly, D. "Kerameikos-Grabung 7ro?noT|c ap%aioXoyiKc?v %copa>v," T?tigkeitsbericht 1956-1961," AA ArchDelt 52, B'1 (1997), pp. 39-43. 80, pp. 277-376. 1841. R. 1985. Demos: The Leake,W. M. The Topography of Osborne, Discovery 2 Classical Athens, vols., 2nd ed., London. of Attica, Cambridge. on K. Lewis, D. M. 1955. "Notes Attic Papadopoulos, J. 2003. Cer?micas BSA 1-36. Redivivus: The Iron Pot Inscriptions (II)," 50, pp. Early Age -. ters'Field in 1959. "Attic Manumissions," the Area of the Classical 208-238. Athenian Hesperia 28, pp. Agora (Hesperia Suppl. 31), -. 1963. "Cleisthenes and Attica," Princeton. Historia 12, pp. 22-40. Parker, R. 1987. "Festivals of the Attic -. Phialai at in to Gods: Pro 1968. "Dedications of Demes," Gifts the 368-380. the Athens," Hesperia 37, pp. ceedings of Uppsala Symposium H. 1993. 1985 T. Lohmann, Atene: Forschungen (Boreas 15), ed. Linders zu und and G. Siedlungs- Wirtschaftsstruk Nordquist, Uppsala, tur des klassischen Attika, 2 vols., pp. 137-147. -. 1996. A His Cologne. Athenian Religion: = LSAG L. H. Jeffery, The Local Scripts tory, Oxford. rev. ed. with A. 1981. "The ofArchaic Greece, sup Petropoulou, Eparche and at plement by A. W. Johnston, Oxford Documents the Early Oracle 1990. Oropus," GRBS22, pp. 39-63. = LSCG F. Sokolowski, Lois sacr?es Pittakis, K. S. 1835. L'ancienne Ath?nes, des cit?s grecques (Ecole fran?aise Athens. et -. 665 d'Ath?nes, Travaux m?moires 18), 1852. ArchEph 1852, pp. Paris 1969. 688.

Malouchou-Da?liana, G. E. 1998. Pritchett, W. K. 1953. "The Attic Ste ou Part "L'ancienne Ath?nes la descrip lai, I,"Hesperia 22, pp. 225 et tion des antiquit?s d'Ath?nes 299. ses W. 1971. de environs, Ath?nes 1835," Reinmuth, O. TheEphebic Tcov rcov Fourth b.c. Apxe?ov /ivrijueicov ??nvcdv Inscriptions of the Century Leiden. Kai TTJ?ATTIK?J?3, pp. 15-244. (Mnemosyne Suppl. 14), on Matthaiou, A. 2004. "nepi xfjc; IG I311," Rhodes, P. J. 1981. A Commentary in ?rriKai emypacpai: ?paKTiKa the Aristotelian Athenaion Politeia," Oxford. IvfiJtooo?ov ei? ?VTJjunvAdolf Wilhelm, Athens, pp. 99-122. Ritchie, C. E. 1984. "The Athenian B. Are Mattingly, H. 1999. "What the Boundary Stones of Public Domain" Right Dating Criteria for Fifth (diss. Univ. of Colorado, Boulder). 1994. Century Attic Texts?" ZPE126, Rosivach, V.J. The System of Public in pp. 117-122. Sacrifice Fourth-Century Meritt, B. D. 1936. "Greek Inscrip Athens, Atlanta. 355-430. T. L. 1939. "The of tions," Hesperia 5, pp. Shear, Campaign -. 1967. "Greek 201-246. Inscriptions," 1938," Hesperia 8, pp. 57-101. T. 1978. and Hesperia 36, pp. Shear, L, Jr. "Tyrants Meritt, B. D., M. L. Lethen, and Buildings inArchaic Athens," in Comes From Solon G. Stamires. 1957. "Greek Inscrip Athens ofAge: 24-97. to Salamis. a tions," Hesperia 26, pp. Papers of Symposium THE BOUNDARY OF MELITE AND KOLLYTOS II?

Princeton Soci and and Their -. 1890. Die Stadt imAlter Sponsored by theAIA, Trittyes, Phylai, Rep Athen and Art and resentation in the Athenian Council ety, the Department of thum 2, Leipzig. Princeton Princeton. H. T. 1933. "Studies in Archaeology, University, (Hesperia Suppl. 14), Wade-Gery, -. ed.W. A. P. Childs, Princeton. 1982. "An Interpretation of Six the Structure of Attic Society: II. P. Die Attikas in the Attic The Laws of Siewert, 1982. Trittyen Rock-Cut Inscriptions Kleisthenes," CQ 27, und die des Kleis Demes of in Studies in 17-29. Heeresreform Lamptrai," pp. zur alten Attic and M. B. 1982. thenes (Vestigia: Beitr?ge Epigraphy, History, Topog Walbank, "The Confisca Geschichte Munich. Presented to in 33), raphy Eugene Vanderpool tion and Sale by the Poletai 402/1 -. 1999. "Literarische und b.c. of the of the epigra (Hesperia Suppl. 19), Princeton, Property Thirty 74-98. phische Testimonien ?ber 'Kera pp. 162-171. Tyrants," Hesperia 51, pp. meikos' -. 1986. Demos and 0.1915. "Zu attischen und 'Kerameis,"'yfM 114, Trittys: Epi Walter, Reliefs," pp. 1-8. graphical and Topographical Studies ?JhBeibl 18, pp. 87-98. in R. Sokolowski, F. 1954. "Fees and Taxes the Organization ofAttica, Weir, G. A. 1995. "The Lost Ar in the Greek Cults," HTR 47, Toronto. chaic Wall around Athens," Phoenix pp. 153-164. Travlos, J. 1960. noAeoSojuucrj ????i?i? 49, pp. 247-258. Sourvinou-Inwood, C. 1988. "Further tcov AOrjvmv, ano tcov npoioTopiKcov Whitehead, D. 1977. The Ideology of Polis Ann tcov rod 19ov Aspects o? Religion," Xp?vcov ju?xpi apxodv theAthenian Metic (PCPS Suppl. 4), ArchStorAnt 10, pp. 259-274. ai vo?, Athens. Cambridge. R. S. 1998. Grain Athens = Pictorial -. 1986a. Demes Stroud, The Athenian Travlos, J. Travlos, The ofAttica, Tax Law 374/3 b.c. Ancient Lon 508/7-ca. 250 b.c.: A Political and of (Hesperia Dictionary of Athens, Princeton. don 1971. Social Princeton. Suppl. 29), Study, H. 1936. and E. 1942. "An -. Thompson, A. "Pnyx Vanderpool, Archaic 1986b. "The Ideology of the Inscribed Stele from Athenian Metic: Some Pendants Thesmophorion," Hesperia 5, Marathon," a PCPS pp. 151-200. Hesperia 11, pp. 329-337. and Reappraisal," 32, -. -. 1954. "Excavations in the 1974. "The Date of the Pre pp. 145-158. in U. von. Athenian Agora: 1953," Hesperia 23, Persian City-Wall of Athens," Wilamowitz-M?llendorff, pp. 31-67. 0opoc: Tribute to Benjamin Dean 1880. Aus Kydathen (Philologische -. 1968. in the Athenian ed. D. W. Bradeen and "Activity Meritt, Untersuchungen 1), Berlin. M. F Locust -. Agora: 1966-1967," Hesperia 37, McGregor, Valley, 1887. "Demotika der attischen pp. 36-72. N.Y, pp. 156-160. Metoeken," Hermes 22, pp. 107 H. R. -. Thompson, A., and L. Scranton. 1980. "The State Prison of 128. 1943. "Stoas and Walls on the Ancient in From Athens F. E. 1982. intra City Athens," Winter, uSepulturae 269-383. to Gordion: aMemo Pnyx," Hesperia 12, pp. The Papers of urbem and the Pre-Persian Walls W. E. 1970. "Notes on rial S. of in Studies inAttic Thompson, Symposium for Rodney Athens," Epi Attic 64 of and Demes," Hesperia 39, pp. Young (University Pennsylvania graphy, History, Topography 67. Museum ed. K. Presented to Papers 1), DeVries, Eugene Vanderpool -. 1971. "The Deme in Kleis 17-31. Philadelphia, pp. (Hesperia Suppl. 19), Princeton, thenes' 1998. Reforms," SymbOslo 46, Verbanck-Pi?rard, A. "H?ros pp. 199-204. 72-79. au le R. E. pp. attiques jour jour: Les calen Wycherley, 1959. "Two Athenian 1980. Grammar in Les Threatte, L. The of driers des d?mes," Panth?ons Shrines," AJA 63, pp. 67-72. Attic 1: des cit?s: Des ? la -. How Inscriptions Phonology, origines P?ri?g?se 1962. the Greeks Built Berlin. de Pausanias. Actes du colloque orga Cities, 2nd ed., London. and E. nis? ? l'Universit? de du 15 au -. 1978. Stones Threpsiades, J., Vanderpool. Li?ge The ofAthens, 1965. "Themistokles' of 17 1997 Princeton. Sanctuary mai (2epartie) (Kernos Artemis Aristoboule," ArchDelt 19, Suppl. 8), ed. V. Pirenne-Delforge, Young, R. S. 1951. "An Industrial Dis A' 109-127. trict of Ancient (1964), pp. 26-36. Li?ge, pp. Athens," Hesperia Traill, J. S. 1975. The Political Organi Wachsmuth, C. 1874. Die Stadt Athen 20, pp. 135-288. zation Attica: A of Study of the Demes, imAlterthum 1, Leipzig.

Gerald V.Lalonde

Grinnell College

department of classics

grinnell, iowa 50ii2

[email protected]