The Victory of Kallimachos Harrison, Evelyn B Greek, Roman and Byzantine Studies; Spring 1971; 12, 1; ProQuest pg. 5 The Victory of Kallimachos Evelyn B. Harrison THE FRAGMENTARY EPIGRAM from the Athenian Acropolis! record­ ing a dedication to Athena by Kallimachos of Aphidna and apparently referring to his service at Marathon has been dis­ cussed and tentatively restored in various ways, but no complete restoration has found general agreement. Meiggs and Lewis, in A Selection of Greek Historical Inscriptions nO.18, print a conservative text, restoring only the first two verses completely and quoting in foot­ notes the supplements of Shefton and Ed. Fraenkel for the last three verses.2 Raubitschek, in a general discussion of Greek inscriptions as historical documents, also prints a text without the most controver­ sial restorations.3 Both still accept, however, the restoration which refers to the offering as "messenger of the immortals" and believe that the column on which the inscription was carved supported one of the winged female figures of which several examples have been lIG 12 609. Most recently, R. Meiggs and D. M. Lewis, A Selection of Greek Historical In­ scriptions (Oxford 1969). Add to their bibliography A. E. Raubitschek, Gymnasium 72 (1965) 512. The long article by B. B. Shefton, BSA 45 (1950) 14D--64, is the most informative about the stone, since it includes photographs and full discussion of difficult letters as well as a great deal of background material. 1 am grateful to Ronald Stroud for making me a good squeeze of the inscription and to Helen Besi for the drawings, unusually difficult to make because of the concave surface in which the letters are engraved. I have profited much from discussions with W. M. Calder Ill, Michael Jameson, Benjamin D. Meritt and Ronald Stroud. None of these bears any responsibility for the opinions expressed in this article. 2 Meiggs and Lewis, op.cit. (supra n.l) 33. [Ka,\tfLax6c fL' av]'8€K€V 'A.pt8vaio[c] ra8€Va[at: a!'[y€AOV a8]ava7"ov ho, 'O[AVfLlTta 86fLaT'] €X0Ctv,1 [ .... 8 .... 7TOA']fLapxo[ c] 'A8€Va[ov T6V ayova: T6V Ma[pa8ov . .•. h]€A£VOVO[ . .... II .....• :] 7Tatdv 'A8€Vatov fL!' ••••••... . 21 •••••.••.•• :] II Raubitschek, op.cit. (supra n.l) 512: [KaAlfLaXOC fL' av]'8£K€V 'A.pt8varo[c] Ta8€Va{a,': aV[Y£Aov a8]avaTov ho, 'O[A6fL7T,a 86fLaTa] €XOC'V I [- u u - 7ToM]fLapxoc 'A8€Va[ov T6V ayova': 7"6V Ma[pa8ovt ••• h]€Mvov o!'[U - U U - U:] 7Ta,dv 'A8€Va[ov fL!'[- - -]. 5 6 THE VICTORY OF KALLIMACHOS found among the archaic marble sculptures from the Acropolis.~ These figures have generally been called Nikai, but Meiggs-Lewis and Raubitschek now incline to identify the offering of Kallimachos as Iris rather than Nike. The purpose of this note is to suggest that not only did the figure represent Nike but the word 'victory' may have appeared prominently in the epigram. 4 Acropolis 690. H. Schrader, Die archaischen Marmorbildwerke der Akropolis (Frankfurt/M. 1939) 122-24, no.77. As Dr George Dontas has pointed out to me, Acropolis 690 must have held a fold of her skirt with her right hand, so that the reconstruction of Werner Fuchs, Die Skulptur der Griechen (Munich 1969) 178 fig.183, is wrong in showing her with a kerykeion in the right hand. She may have held a kerykeion in the left hand, but it seems unlikely to have been the beautiful bronze one with the head of Pan that Hampe wished to assign to it (Die Antike 15 [1939] 168-74). Besides the fact that the beauty and significance of the Pan-head would have been lost to the spectator because of the high setting of the statue, there is the difficulty that our literary tradition connects the Pan-cult with Miltiades rather than with Kallimachos (Simonides fr.143 Diehl; Anth.Plan. 232). Raubitschek (op.cit. [infra n.8]) makes the connection of the Nike with an Ionic votive capital (Acropolis 3776, 3820, 3830, 8312 and a fragment without a number) which seems to have the correct diameter for the column. Both these connections seem extremelyprobable, without being actually capable of proof. The cutting in the top of the capital is a rectangle which preserves its full length of 0.36 m. at one edge (Raubitschek gives 0.35 m. as the length; perhaps he measured at a different point). The width is not fully preserved but can be calculated from the center of the pour-hole in the middle of the bedding. Raubitschek gives 0.195 m.; my own somewhat hasty measurement 0.20 m. These dimensions, whether 0.36 m.x 0.20 m. or 0.35 m.x 0.195 m., are well suited to the plinth of the Nike, which E. Langlotz (Schrader, op.cit. 122) gives as 33x 17 em., "probably meant to be rectangular 34x 14 em." The depth of the cutting (0.045 m.) is exactly right for the height of the Nike's plinth, given as 0.04-0.05 m. The present museum setting makes it impossible to remeasure the plinth. A Nike is the most likely form of statue to fit such a long narrow plinth-cutting. Korai would have a quite different shape of plinth, and a sphinx in the late archaic period would have a shorter plinth in proportion to its width. Acropolis 690 is the largest Nike preserved. As the column bearing the Kallimachos dedication is larger than the usual inscribed columns, the probability of connecting capital and column is great, even though here again exact measurement is not possible. The top of the column is not preserved and we have no positive evidence for its height; the diameter where it can be measured is 0.32 m. The full circumference of the column is nowhere preserved, and the fluted portion has flattened the arc of its section, but this measurement seems to be approximately correct. The outside edge of the resting surface of the capital is chipped or eroded all around, so that the diameter can be calculated only by guessing at the amount of recession of the lost cyma reversa below its top edge which is preserved. Raubitschek's diameter of 0.32 m. is taken with reference not to an "incised line on bottom," as he says, but to a pencil line on the broken surface marking someone's calculation of this recession. The recession might have been a little greater, but the top diameter of the column might also have been a little less, so that they would still fit. Besides these coincidences, we have the fact that the style of this Nike seems exactly right for the date 490 B.C. All in all, the burden of proof would seem to be on anyone who wishes to question the connection. I have gone into the matter in some detail, since it would be easy for a skeptical archaeologist, noting that the correspondence is less than absolute, to cast doubt on the whole identification. Such doubts seem to me to be unreasonable. EVELYN B. HARRISON 7 I have tried to show elsewhere that Kallimachos is depicted as the most prominent figure in the south frieze of the Temple of Athena Nike and that this figure of Kallimachos is derived from the Marathon painting in the Stoa Poikile.5 A careful reading of the literary texts and the reflections in art that can be associated with the painting indicates that Kallimachos appeared in the center of the battle, mortally wounded but victorious over his immediate opponent. Polemon, in his pair of declamationes comparing the ap€'-7} of Kallimachos with that ofKynegeiros, states that Kallimachos died in the middle of the battle but that his brave stand turned the Persians to flight.6 His figure, both for the painting and for Polemon, seems to have marked the moment of victory for the Athenians. The three phases of the battle-attack. victory and pursuit-seem to have been high-lighted in the painting by the three famous figures ofMiltiades, Kallimachos and Kynegeiros. The pictorial narrative was progressive, moving from the beginning of the fight at the left with Miltiades urging the troops into battle and ending with the Persian ships and the death of Kynegeiros on the right.7 The representation of the battle on the Nike Temple shows only the middle phase. Neither Miltiades nor the ships are there. There thus appears to have been in the minds of the Athenians a specific identification of the heroic figure of Kallimachos with the idea of victory. Since the archaic winged female figures, of which several have been found on the Acropolis, correspond in form and dress to the archaic figure from Delos that is plausibly identified as the Nike of Archermos, it is natural to regard them all as Nikai.8 The cult of 5 E. B. Harrison, "The South Frieze of the Nike Temple and the Marathon Painting in the Painted Stoa," to appear in AJA 76 (1972). 6 Polemon, Declamationes quae exstant duae, ed. Hinck (BT, Leipzig 1873) 1.21-22; 2.25, 28, 31. 7 Pausanias 1.15. a Schrader-Langlotz, op.cit. (supra n.4) nos. 67-84. The indication that these figures were called Nike comes first from the scholiast to Ar. Aves 574, who says that Eros and Nike acquired wings at a relatively late date, that Archermos the father of Boupalos and Athenis was the first to make a Nike with wings but that some say rather it was the painter Aglao­ phon of Thasos. Since a base with the names of Mikkiades and Archermos found in the sanctuary of Artemis at Delos would fit in size and date a winged figure also found there. this statement of the scholiast supports the attribution of the winged figure to Archermos. For the best discussion of the inscription, see L. H. Jeffery, The Local Scripts of Archaic Greece (Oxford 1961) 294-95. For the date of the 'Nike' see E. Harrison, The Athenian Agora XI: Archaic and Archaistic Sculpture (princeton 1965) 5-6. So long as the inscription belonging to the winged figure in Delos is not certainly restored, we can not be sure whether this figure was actually called 'Nike' in Delos or Archermos, when he came to Athens to work (cf. A. E. Raubitschek. Dedications from the Athenian Akropolis [Cambridge (Mass.) 1949] 8 THE VICTORY OF KALLIMACHOS Athena Nike already existed on the Acropolis in the archaic period.9 There is no comparable tradition which would link Iris with the Acrop­ olis. It is therefore reasonable to suppose that all the winged female figures of the archaic period in the type resembling the Nike of Archermos that have been found on the Acropolis do in fact represent Nike and that figure of Nike was a common dedication to Athena.10 484-87), first used such a figure for what the Athenians called 'Nike'. Athena Nike, with a pomegranate in one hand and a helmet in the other, well illustrates the complex nature of the Nike of the Athenian Acropolis. The vegetation-fertility aspect embodied in the pomegranate shows her akin to the Aglaurids and the Graces; the helmet belongs to the city-protecting goddess. The winged female figure in Athenian art is often a counterpart of Eros. Many figures resembling closely the marble Nikai of the Archermos type occur among the Acropolis bronzes (A. de Ridder, Catalogue des bronzes trouves sur I'Acropole d'Athenes [BIFAR 74, Paris 1896] nos. 799-801, 805-14) as attachments for vessels and stands, and among them, in the same scheme and used in the same way, are three male figures, two of which are nude (ibid. nos. 802-04). These figures are not unknown elsewhere, but their relative frequency among the Acropolis bronzes is striking. Two similar figures in an antithetical composition crown the handle of a mirror in the British Museum (A. Greifenhagen, Grie­ chische Broten [Berlin 1957] fig.28). As it is uncertain whether this beautiful object comes from the Mainland or the West, Greifenhagen names the figures either Nike and Eros or Iris and Zephyros, but says that if it is from the Mainland they must have been called Nike and Eros. Hedwig Kenner, who investigated the representations of the female winged figures on Attic wedding vases (JOAI 31 [1939J 81-95), affirmed that they were probably most often called Nike. She rightly accepts the possibility of a certain overlapping of Iris and Nike in the archaic representations, as these winged figures, like the mistress of animals, belong to the world of orientalizing images which were variously attached to variously named Greek gods and daimones. In broad terms, Iris seems to belong more to the marine world, being allied with the sea-storm spirits like the Harpies and Gorgons, while Nike lives with Eros in the land-world of fresh water and flowers, but as the images become crystallized in the more conSciously abstracting art of the classical period, a new kind of overlapping occurs. Nike brings victory at sea as well as on land, and her iconography changes accordingly. It is not possible always to name a given image from its attributes alone, but it seems clear that Iris and Hermes are not subject to the kind of multiplication that we find with Nike and Eros. Iris and Hermes are the messengers of the gods par excellence, but Erotes and Nikai also appear with the kerykeion, the messenger's staff. Kenner and Greifenhagen both give a number of examples. In fact, these winged deities represent in a certain sense conveyors of power rather than the actual source of power. Thus the phrase "messenger of the immortals" in the Kallimachos epigram does not in itself argue for identifying the offering as one or another of these deities. It is the fact of its being an Acropolis dedication to Athena that makes Nike by all odds the most likely name. S G. Welter, AA 1939, 12-13. Raubitschek, op.cit. (supra n.8) 359. Both accept a date around the middle of the sixth century for the inscribed altar of Athena Nike that seems to have been the first archaic altar of this goddess. 10 In the classical period the golden Nikai dedicated to Athena by the state are the most famous examples; see D. B. Thompson, Hesperia 13 (1944) 173-209. The Nike Temple Para­ pet makes it absolutely clear that Athena Nike, though a wingless goddess, was served by winged Nikai. EVELYN B. HARRISON 9 It is generally believed that the cult of Athena Nike in Athens in the sixth century B.C. was associated with the Panathenaic Games, so that these dedications, if they were thank-offerings for victories, will have referred to victories in games rather than in war.ll In the Kallimachos dedication we have the word aywva. Shefton has pointed out that at the time of Marathon the meaning of this word had not yet been extended to include J..'I]vWV rT€, quoted by Eur. Hecuba 929-30, W 7Tatlle-c 'E>J..&vwv. By the fourth century B.C., when the language of the Persian War epigrams was taken over for civilian honors (cf Harrison, op.cit. [supra n.8] 117), 7Tatlle-c' A8T]valwv seems to have become simply a poetic equivalent for &fjp.oc 'A8TJvalwv, but that is a late development. 14 Eugene Vanderpool, Ostracism at Athens (Semple Lectures SER. 2, preliminary publica­ tion 1970) 14-15. EVELYN B. HARRISON 11 In the second verse &r[yeAov &OJav&Tov is universally accepted, and this restoration fixes the length of the lacuna in the line below, which is the first lacuna in the third verse of the inscription. The plaster re­ construction of the column in the Epigraphical Museum at Athens is based on this restoration, and it can be considered as approximately correct, allowing only for the minor variations in the width and spacing of letters which obviously exist in this inscription.15 For the relative clause qualifying the 'immortals' whose 'messen­ ger' the offering is, Lolling restored: hot o[p£1C71I6vwv a &£8Aot( (F. Hiller von Gaertringen, Historische gnechische Epigramme [Bonn 1926] no.2). EVELYN B. HARRISON 13 supposed that at this date the Polemarch was in charge of the arrange­ ments for the festival. We have absolutely no evidence, however. that the Polemarch ever had such duties. It appears most likely that the conduct of the festival was already in the hands of a board of L€PO- 7TOL01.23 If we restore IlLK'CaC and assume that the occasion of the dedication was a Panathenaic victory, the sense of the first three verses is more or less complete in itself, and it is not necessary to assume that 761' at the beginning of the fourth verse depends on ayolla. This, however, is what nearly all previous restorers of the inscription have assumed. Hiller's restoration, based on Lolling's reading, was: 761' M€[oov 7€ Kat. h]€Mvov. This was generally accepted until Shefton pointed out that the fifth letter is unmistakably an alpha.24 After that, some form of ()~Tov aiEl:] would take the same space. Merely as a demonstration that this length is not impossible, we could sketch in hot ·O[AVP.7TO~ h'8pac E7T]'XOCLV for the end of the second line (PLATE 1, figure 2). That is not to imply that this rather clumsy (though not unmetrical) expression was actually on the stone, but simply to show that a space of this length would not be beyond the bounds of sense and metre. If a space as short as that occupied by 'O[Avp.ma 8&p.aT'] €XOCLV were called for, one might read ovoJ.L[a DE KaALcTov :] (for ovvop.a D~ KaAAtCTov) but this seems less probable in the light of our observations about the letter space before EXO~IN. In any case, the verse can very reasonably be terminated with some expression extolling the fame that Kallimachos won for Athenian men-at-arms. Now, however, the beginning of the line becomes difficult if we insist on restoring the word Marathon, since it is hard to find anything that will fill the metre and not overfill the space. Weare forced to ask whether Marathon is all that certain. It is by no means the rule that battles must be named in the epigrams that refer to them. The Ther­ mopylae epigrams are a striking example. In the present poem 7T(udv 'A()Evalov makes it clear that this part of the poem refers to a battle, and the mention of Kallimachos as Polemarch leaves no doubt as to which battle it is. If Marathon is not named, then, there is a fairly wide choice of words beginning p.a, and it no longer seems pos­ sible to suggest a restoration with much faith that it was really so on the stone. Working with what we have accepted as natural restora­ tions before and after this lacuna, we may find it easiest to take the first word as TOV, referring to the Athenians mentioned in the preced­ ing line. This would give a significant pause between the part of the poem concerned with the Panathenaic dedication and that directly extolling the bravery of Kallimachos at Marathon. Such a relative use of TOV at the beginning of the verse was surely known to the author of our poem from the same Acropolis epigram which must have in­ spired 7Ta~dv 'A()Evalov, the epigram on the base of the four-horse chariot dedicated as a tithe of the victory over the Boeotians and Chalcidians in 506 B.C.26 I think we would like our poem to say what the later tradition also se See n.13 supra. EVELYN B. HARRISON 15 said about Kallimachos, that he was the bravest of the Athenians, and it does not seem at all impossible that this could have been said metrically in the space available. Tov",aN aptcr€vov (PLATE 1, figure 3) would fit the space and give the sense that we want.27 If this will fit so easily, it seems obvious that many other restorations might do so too, and we can hardly feel sure enough of anyone of them to include it in a conservative text. At the same time it seems equally obvious that 'Marathon', for which no one has yet found a restoration free of ob­ jections, cannot belong in a conservative text either. The problems of this lacuna furnish a very instructive lesson in the inadequacy of standard epigraphical method for restoring archaic in­ scriptions. The usual texts indicate by dots or a combination of dots and a numeral an exact number of letter-spaces represented by the lacuna. It is easy to be misled by this into thinking that any restora­ tion is all right so long as it has the designated number of letters. So Fraenkel, in suggesting Mapa8ovo8€v, says, "In questions of spacing I have followed Shefton's indications." If one draws his restoration and Shefton's, however, it is instantly apparent that Shefton's fits the space determined within fairly narrow limits by a!'[y€Aov aOJava.rov in the preceding line, whereas Fraenkel's with one more round letter and no iota or rho, comes out definitely too long.28 Archaic letters inscribed on stone do not live in private spaces like houses in a row of city lots; they assemble amiably with a kind of flexible tact by which each adjusts to the space needs of its neighbors. So a tau between round or oblique letters can actually invade the air­ space of the letters next to it without making them seem crowded, whereas an omicron or theta between vertical neighbors needs its own 27 Forms of apLCT€VW are exceedingly common in this position in the hexameter, so much so that it would not be surprising to have a rather colorless intensifying word such as lJ.&Aa filling up the space before aptCT€VWV. The poet of the archaic Corcyrean epigram Peek 73 (Friedlander 25) wrote 1ToA>.dv aptCT€VOVTa, which recurs in the Thessalian epigram of around 500 B.C. Peek 69 (Friedlander 160). ·Ov 1TOT' aptCT€VOVTa, used in Iliad 7.90 with a specifically retrospective sense, is taken up in the epigram from Thisbe, Peek 321 oC 1TOT' aptCT€VWV. See A. E. Raubitschek, in Entretiens Hardt 14 (Vandcevres-Geneve 1967) 5-7. M&Aa with the superlative occurs in an epigram formerly in Salonika, Peek 876 (eIG II 1973), for which Peek suggests with a query a date in the second century of the Christian era, € ~V€K€v ~c ap€r1jc Ka~ cw.' aplcrT}c. 28 It is impossible to calculate the exact space required for any restoration, owing to small variations in the size, shape and spacing of the letters, all of which depend on the let­ ters preceding and following; but by simply tracing off existing letters it is possible to get an idea whether a given restoration is definitely impossible or within the bounds of possi­ bility. The skeptical reader will have to try this for himself. There is no short-cut. 16 THE VICTORY OF KALLIMACHOS full width plus a breathing space on either side. Oblique strokes tend to become more or less oblique according to the shape of the adjacent letters, as we can see clearly by comparing the nu's in our inscription, no two of which have exactly the same shape. The result of all this is that TOY J.ufA' aptcT€voy with one more letter than Shefron's TOY MapaOoYt 7Tp6 takes the same space, whereas TOY MapaOov60€v, with the same number ofletters, takes about one ~.~~ ( N \ .... ~~~{t/I ~~> '~.~~N ----- --- --- ---- - - --- - -- -.-~ o Ot>/\i\,W ~JI_· Wi __ [ __________ • __________________ ~·:-:" _____ ~.~n o .10 .20 ~ .-==::.3 M. o .10 .20 H bd ~ M. Figure 3 Figure 2 DEDICATION OF KALLIMACHOS OF APHIDi'l'A FRO~l THE ATHENIAN ACROPOLIS (IG 12 609) Drawings by Helen Besi EVELYN B. HARRISON 17 break the 'rule', it seems worthwhile to weigh the irregularity of such a reconstruction against the various difficulties raised by the other re­ constructions that have been proposed. The suggestion of Hiller that the third verse began like the first with 'Kallimachos' would of course raise no spatial or historical prob­ lems, but it would make a rather awkward-looking monument with the parallel repetition of the name at the beginnings of the two lines. From the purely spatial point of view the best is Lolling's c'T'Y}ca­ J.L€VOC, which works out very well with a two-lambda 'Kallimachos', making the second line begin just a little (less than one letter-space) lower than the first. Such a relationship is quite common in column dedications with two vertical lines of text. We have mentioned the historical difficulty: there is no support in any of our sources for the idea that the Polemarch was responsible in any way for the conduct of the Panathenaea. There may also be a question whether the middle c'T'Y}caJ.L€voc is properly used of a single official putting on a festival for the whole people, though it is properly used in the Homeric Hymn of the plural Ionians holding their festivals (for themselves) at Delos.33 The restoration hoc (c)'TECaC proposed by Shefton34 comes out near­ ly two letter-spaces too short, so that it is not enough better than VLKEcac to be preferred on the grounds of spacing alone. In either case we should have to admit that the mason did not really intend to start the second line at the top of the flute. To write out the double sigma would hardly make things better, for it seems highly unlikely that a mason who used only one lambda in 'Kallimachos' would use two sigmas here. Since two sigmas can nestle closer together than two lambdas the inequality would be greater with double letters than with single. Shefton preferred the relative to a simple participle in this place because he wished to connect the whole phrase with a main verb in the fourth verse. He admitted, however, that it was the presumed connection of ayova with Marathon that made this necessary.35 He does not discuss the space problem here except in terms of "number of letters," though elsewhere he seems to have been very careful about spacing. In view of all this, we might well feel that the advantage of the easy 33 Hymn.Hom.Ap. 150,57' C2v Cn]CWVTaL aywva. 34 Shefton, op.cit. (supra n.l) 148. 35 Shefton, loc.cit. 2-G.R.B.S 18 THE VICTORY OF KALLIMACHOS sense and normal sound of v'Klcac outweigh the possible esthetic­ architectural disadvantage of having the second line begin noticeably lower than the first. When we come to consider the lower end of the second line, however, we are struck by the fact that, unlike any other column dedication that we know, it extends much lower than the first. To those who believed that the second fluting and the three verses carved in it were added later than the two hexameters of the first flut­ ing, this was not disturbing. Perhaps one might say the same about those who see the break between the dedication part and the Mara­ thon part of the poem as coming at the end of the first line, i.e. at the end of the second hexameter, even though they believe the whole poem was composed and carved at one time. In this case, however, it would seem less likely that the mason, knowing in advance that the second part of his inscription was to be the longest, would have wasted space at the beginning of the second line. In the case of our present suggested restoration, which places the break after the third hexameter, a different explanation suggests it­ self. It could well be that the decision to add to the Panathenaic dedi­ cation two verses in commemoration of the bravery of Kallimachos at Marathon was taken after the first three verses had been carved. This would have meant that only one verse (or possibly in the original plan one verse plus the sculptor's signature) was intended for the second line. At some time after the second line was begun, the deci­ sion was made to add the final couplet.36 If we look at the inscription with this idea in mind, we can see some possible indications that the verses beginning with TOV were added, if not in a different hand, at least at a different time. The three dots of the interpunct are not in a straight line, but seem to have been fitted in as an afterthought. The alphas in the last two verses are more slant­ ing in all their strokes, no longer so boldly upstanding and right­ angled as the alphas of the first three verses, and all the carving tends to be broader and a little less crisp, so that the cores of the closed let­ ters become smaller. It seems conceivable that the stone was in a different position when the last two verses were carved. 3S This need not mean that the first three verses were actually carved before the battle. It could well be that the work on the monument was at first continued after the battle ac­ cording to the original plan, with the text that had been requested by Kallimachos him­ self. The use of the rough column with only two flutings may imply haste in setting up the monument. Perhaps the Nike and the capital, which show no such signs of haste, were begun before the column shaft. EVELYN B. HARRISON 19 No kind of space requirement limits the restoration of the last verse. Shefton's suggestion, JLV[EJLEV 0' apE-TEC Ka'TEAEL7TEV], seems to give the required sense, whatever the actual wording may have been. Meiggs and Lewis note, «We regard v after JL as probable rather than certain." An alpha cannot be absolutely ruled out, but nu, suggesting the idea of memory or memorial, is attractive. If Shefton's objections to JLvEJLa instead of JLVEJLEV37 do not hold and if the final couplet was, as it possibly might have been, an elegiac instead of a pair of hexam­ eters, the last line might have ended: JLp[EJLa OE hEc apE'TEc.38 We have arrived, then, at a conservative text which is somewhat less full than those recently published, and at a very tentative restora­ tion exempli gratia that gives a slightly different emphasis to the Pole­ march's achievement. The conservative text would read: [KaAA{JLax6c JL' aV]EOEKEV 'Acf>LSVaLo[c] TaOEva{aL: ap[YEAoV &B]ava'Tov hot ' 0[- U U - U]EXOCLV, I [- uu - 7ToM]JLapxo[c] 'AOEva{ov 'TOV ayova: 'TOV JLa[U - uu - h]EAEv ovof.L[a - uu - -:] 7TaLcLv 'A8Eva{ov JLY[- UU - UU -] or [- - uu - uu - _]39 The restoration exempli gratia would be: [KaAA{JLax6c JL' aV]EOEKEV ' Ac/novaLu [c] Ta8Eva{aL : aY[YEAoV aO]ava'Tov hot ' o [AVJL7Tot hESpac €7T ]EXOCLV, I [vLKlcac 7ToM]fLapxo[ c] , AO€va{ov 'TOV ayova : 'TOV fLa[A' apK'T€VOV h ]EA€V OVOf.L[ a fLEV KaAALC'TOv:] 7TaLcLv 'A8Eva{ov JLY[ Ep.a OE hEc apE'T€c.] or JLY[EJLEV 0' apE'TEC Ka'TEAEL7TEV.] 37 Shefton, op.cit. (supra n.1) 157-58. Actually Shefton was restoring "he left behind" rather than "he won," which may have made some difference in his thinking on this ques- tion. 38 Or p."ip.a T€ hie ap€Tie, according to the restoration of the preceding line. The fact that Kallimachos died is not expressly mentioned, since this is not a funerary monument. 39 The sigma of 7ToMp.apxoe was seen and drawn by Lolling (op.cit. [supra n.16]) hut has since been broken away. The stroke which I have given as a dotted mu and Raubitschek and Shefton as a dotted nu is not actually preserved to the extent that any chiselled facet sur­ vives, but the line of the break seems clearly to follow a stroke here, becoming irregular where the stroke terminated above and below. 20 THE VICTORY OF KALLIMACHOS In both restorations it is understood that