¹ Bibliographical Note.—The ‘Simonides Question’ has not seriously advanced since Boas', M. elaborate torso, De epigrammatis Simonideis I, Groningen, 1905Google Scholar. See also Reitzenstein, R. (Epigramm und Skolion, 1893, pp. 104Google Scholar sqq.: Art. Epigramm in Pauly-Wissowa, 1909, col. 71 sqq.), Geffcken, J. (Griechische Epigramme [Text and Commentary], Heidelberg, 1916Google Scholar: Studien zum griechischen Epigramm, in Neue Jahrb. 1917, pp. 88 sqq.: Art. Simonides in Pauly-Wissowa, 1927, col. 192 sqq.), and of course Wilamowitz, (Sappho und Simonides, 1913, pp. 192Google Scholar sqq.).

It is well established that ascription to Simonides in the Palatine Anthology is no guarantee at all that Simonides wrote the poem. Boas believes that a collection was made at the close of the fourth century B.C. of famous epigrams, which ascribed them all to Simonides. He further holds that certain poems of Mnasalkas, a writer of the third century B.C. who modelled his style on Simonides, were later ascribed to Simonides: the only one of this class I have occasion to quote is Epigram [O]. For an example of third-century epigrams on classical subjects which were actually inscribed, see Epigrams [W] and [X].

On the question of inscribed epigrams, see especially Wilhelm, A., Simonideische Gedichte (ÖJh. II, 1899, pp. 221Google Scholar sqq.), Bormann, E., Zu Denkmälerepigrammen des V Jahrh. v. Chr. VI, 1903, pp. 241Google Scholar sqq.), Wenz, S., Studien zu attischen Kriegergräbern (Dissert. Erfurt, 1913)Google Scholar, von Domaszewski, A., Die Hermen der Agora zu Athen (SB. Heid. 1914, Abh. 10Google Scholar), Weber, L., Steinepigramm und Buchepigramm (Hermes, LII, 1917, pp. 536Google Scholar sqq.), ζΥΚ ΕΦ ΕΡΜΗΙ (Philologus, LXXIV, 1917, pp. 248Google Scholar sqq., LXXVI, 1920, pp. 60 sqq., LXXVII, 1921, pp. 77 sqq.), Buschor, E., Maiandrios (Philologus, LXXXVI, 1931, pp. 424Google Scholar sqq.).—For the Attic Public Cemetery, see the references in IG. I². p. 242.

The collections of metrical inscriptions by Kaibel and Preger are now rather obsolete: a new edition is projected by Peek, W. (Gnomon, VIII, 1932, 559Google Scholar sq.). New discoveries, restorations, etc. are reported in due course in the volumes of S[upplementum] E[pigraphicum] G[raecum]. Good selections have been made by Geffcken, J., Griechische Epigramme, Heidelberg, 1916Google Scholar, and F. Hiller von Gaertringen, Historische griechische Epigramme, Bonn, [Kleine Texte, 156], 1926Google Scholar. I give references to the last wherever possible. The ‘Simonidean’ epigrams are given in Diehl's Anthologia Lyrica under Simònides, and other classical epigrams and epitaphs in his first fascicule (Poetae Elegiaci) and under Anacreon in the Poetae Melici, Monodia. Stadtmüller's Teubner edition of the Palatine Anthology is unfinished but includes Book VII (Sepula/cralia), and it is to be hoped that Waltz's Budé edition soon will: Paton's Loeb edition is complete.

² Dr. Macan has claimed Herodotos as an Irishman: surely we need not insert τούσ before ὑπὸ?

³ See Boas, M., De epigrammatis Simonideis, § 14Google Scholar (pp. 16 sqq.).

⁴ See Boas, , De epigr. Simon. pp. 3Google Scholar sqq., esp. p. 13.

⁵ The ‘historical’ implication of this ποτέ is developed by the τοτέ in line 3: cf. the dedication of Alexander's Lion-Hunt at Delphi, (BCH. XXI. 598Google Scholar: Gaertringen, H. v., Epigr. 82Google Scholar): ον ποτε, Αλεξανδροσ τοτε οθ ειπετο, etc.

⁶ Δαμάσαντεσ—ἄθεσαν: it is a dedication by the victors.

⁷ A similar ποτέ occurs in the lines quoted by Plutarch, (Arist. 19Google Scholar, De malig. 42) from the altar of Zeus Eleutherios at Plataiai. We must infer, after these instances, that it was inscribed later than 479.

⁸ This results, I think, from in line 1. The figure of approximately 100 survivors of the siege results from Aeschines III. 187, ibid.

⁹ Cf. ἀρξάμενοι πρῶτοι in the Herm-inscription quoted by Harpokration (πρὸσ τῇ πυλίδι Ἑρμῆσ: = H. v., Gaertringen, , Epigr. 32Google Scholar).

¹⁰ Pindar's words, quoted by Plutarch, , Them. VIII. 3Google Scholar.

¹¹ Ηὖρον is, of course, a word proper to first inventors.

¹² Κρατερόν Aesch. κρυερόν Plut.

¹³ The ‘process’ in question is the breaking of Persian morale:

¹⁴ III. 183–4, Robert, , Pausanias als Schriftsteller, p. 327Google Scholar, speaks of this as a ‘decree of the Pentekontaetia quoted by Aeschines’! Weber, , Philologus, LXXIV, devotes 26 pages (257–282Google Scholar) to the ‘three Eion epigrams’: but though he appears to interpret πρῶτοι rightly (p. 263), he yet assumes that the poem was written before there were any δεύτεροι, and (pp. 278–9) speaks of the ‘decree’ as if it was real. So does Judeich, , in the welcome and excellent new edition of his Topographie von Athen (1931), p. 73Google Scholar. Hiller v., Gaertringen, dates the Hermai to 476/5, both in IG. I ², p. 277Google Scholar, and Epigr. 34. Meyer of course ascribes the ποτέ (as usual) to the poet's lack of skill: Forsch. II. 12 sq. 20 sq.

¹⁵ Plut, . De malig. 870Google Scholar e, f, ἀκμᾶσ ἑστακυῖαν ἐπὶ ξυροῦ Ἑλλάδα πᾶσαν ταīσ αὐτῶν ψυχαῖσ κείμεθα ρυσάμενοι: = H. v., Gaertringen, , Epigr. 22Google Scholar. Ibid. 41, the men who died at Byzantion ἰχθυόεσσανρυόμενοιχώρην hardly the Athenian conquerors: the Ionic Revolt? or some yet earlier resistance to Persia?

¹⁶ I do not wish here to broach the problem of how the canon of patriots (τοιδε τον πολεμον επολεμεον, Tod, , Gr. H. Ins. 19Google Scholar, Ditt., Syll. ³31Google Scholar) was determined. There were certainly isolated exploits not recorded in Herodotos, and not gaining for the doers admission to the lists: cf. Nachmannson, , Hist. gr. Inschr. 11Google Scholar = H. v., Gaertringen, Epigr. 24Google Scholar, the capture of two Karian ships by the Peparethians.—There is perhaps some controversial defiance in Herodotos' language, VII. 174.

¹⁷ Weber, , Hermes, LII (1917), 551Google Scholar, says, ‘tatsächlich ist ποτέ für das Steinepigramm charakteristisch.’ In the case he is considering (a Hellenistic monument to Solon), ποτέ is, of course, in place. That ποτέ is in fact rare, and never meaningless or stylised, will be seen in such collections of epitaphs and epigrams as IG. I². 927–1086 (‘Monumenta sepulchralia’), or H. v. Gaertringen's Hist. gr. Epigramme, or Geffcken, Griech. Epigramme, esp. pp. 1–64Google Scholar (archaic and classical), or the large collection given under Simonides in Diehl's Anthologia Lyrica. In this paper I have, of course, been at pains to assemble those poems which do contain ποτέ: need I add to them, e.g. IG. XII. 7. 106, Φαι|σατελησ ποτε καλοσ εβα<σ>κηνε<ν> μιν ο μαν[τισ]?

¹⁸ Where we may translate it ‘formerly’: it usually qualifies (in these cases) an imperfect. [H] is a good exceptio probans.

¹⁹ IG. I². 927, = H. v., Gaertringen, , Epigr. 20Google Scholar.

²⁰ Anth. Pal. VII. 256 (Plato: = No. 10 in Diehl's Anthologia Lyrica).

²¹ St. Byz. θέσπεια, = H. v., Gaertringen, , Epigr. 19Google Scholar.

²² Anth. Pal. XIII. 26 (Simonides: = No. 89 in Diehl's Anthologia Lyrica).

²³ Phrases such as οὐδένπω τοιοῦτον (H. v., Gaertringen, , Epigr. 49Google Scholar, line 3: vide not. crit.), πολισ ηεδε ποθει (ib. 53, line 9; cf. 18, line 1), εσ τοδε (ib. 36, line 6), show the poet's mind fixed on his own moment, not on futurity. When I read this paper to the Oxford Philological Society, my attention was called (by ProfMurray, , MrSyme, , ProfBeazley, ) to such passages as Aesch. Ag. 575 sqq.Google Scholar, Homer, , Iliad, 6Google Scholar. 459 sqq. and especially 7. 87 sqq. [not an imagined epitaph, but an imagined traveller's comment], ἀνδρὸς μὲν τόδε σῆμα πάλαι κατατεθνηῶτος ὄνποτ' ἀριστεύοντα κατέκτανε φαίδιμος Ἔκτωρ. In the passages from the Iliad, the speaker imagines a form of words to be used by posterity looking back on the present as a story: this is not, I submit, what the epitaph-writer is doing; though the passages show that the Greek imagination is agile enough to take that leap if it chooses. And the Iliad after all is a story: the poet's imagination is not fixed on his heroes' present as actually as the epitaph-writer's is; and this is especially evident in the Agamemnon passage, where τῶδε—ἐν ἡλίου φάει (575) is followed by δήποτε (577) and ἐρχαῖον (579).—-The long futurity foreseen in the epitaph on Midas (Plato, , Phaedr. 264Google Scholar c, Anth. Pal. VII. 53, etc.; cf. Simonides, ap. Diog. Laert. I. 6) is no more than a rather shocking elaboration of ἀθάνατον in [K]¹ [N]⁴ [W]¹ (or κάλλιστον in [M]¹ or 3ωόν in [F]³): it does not, any more than these passages, thrust the poet's present into the past.

²⁴ I alter the end of line 3 to avoid repeating αρετε: but to say ‘valour’ in half a hexameter leaves a paralysingly wide choice.—Επαρ has perhaps better fifthcentury authority than ετορ.—Powell's suggested [Φ]θ[ιμενοι] for line 4 would perhaps require ελαβομ, which it cannot have: cf. lines 6, 10, 11: in line 12 there is a strong pause before φσυχασ.—I feel certain Kaibel is right in insisting that και in line 3 must add a second object to σεμαινεν: if it introduced a new clause it would be quite intolerably flat. [Perhaps this slightly spoils the και in line 2, so possibly write γενεσομενοισ for και εσσομενοισ: but I think not.]— I am unable to trace Fauvel's sketches of the stone in its earlier stage: they are neither in the British Museum nor in the Archives of Inscriptiones Graecae in Berlin.

²⁵ If there are others, they will equally emand explanation. I do not count Anth. Pal. VII. 270, τούσδε ποτ' ἐκ Σπάρτας ἀκροθίνια φοίβῳ ἄγοντας ἐν πέλαγος μία νὺξ ἐν σκάφος (vel εἶς τάφος) ἐκτέρισεν: which is merely a variant on the original τούσδ' ἀπὸ Τυρρηνῶν, etc. The MSS. evidence is given by Stadtmüller ad. loc., ‘Entscheidet, wenn nicht schon das unbequeme ποτέ, so doch das bekanntere Sparta,’ Wilamowitz, , Sappho und Simonides, 213Google Scholar. Besides, ἀπὸ Τυρρηνῶν ἀκροθίνια is an integral phrase, like [αποΜεδ]ον ακ[ροθ]ινια in the inscription found on the side of the Athenian Treasury at Delphi, [Tod, Sel. Gk. Hist. Inscr. 14Google Scholar = Dittenb., Syll. ³23Google Scholar; see now too BCH. 54 (1930), p. 315]. Pomtow indeed suggests that they were Hieron's spoils from Kyme (Dittenb., Syll. ³35Google Scholar, B b).

For Anth. Pal. VI. 50 (Simonides 107, Diehl) see note 86 infra.

²⁶ ÖJh. II. (1899) 221 sqq.

²⁷ Especially when the Boeotian cavalry whom they were fighting were supported by the (whole?) allied infantry (Thuc. II. 22. 2). Cf. Xen., Hipparch. VII. 4Google Scholar, etc.— The careful stoichedon lettering also possibly favours a date close to No. 945: contrast 943 and 763. Yet this might be due to its not being a ‘routine’ epitaph: see note 31 below.

²⁸ And regularly from 450 onwards on this same stone, lines 24 sqq.

²⁹ The exact circumstances can only be guessed: unless Wilhelm's search for the stone should succeed at last. We might then date it more narrowly. I am not sure whether anything at all can be inferred from the Doric vocalisation (ηιπποσυναι etc.): possibly it points to the Kimonian panhellenic milieu.

³⁰ Unless the occurrence of ποτέ in [O] be taken as external evidence that the Eurymedon cenotaph was not erected until the battles had receded, in some special sense, into the past. See below, where [O] is discussed.

³¹ This case is very clearly distinct from [L]. The tone of this poem is flatter, it is a routine poem: the Tanagra verses have a certain restrained pity, almost indignation. The Eurymedon poem, though less ‘special,’ is well written in a noble tradition: I am most unready to impute to its poet the false caesura in line 1.

³² The date hardly matters here, so long as it is later than the Eurymedon: so I need not discuss the suggested alternatives (c. 408, etc.). Epigraphically, it cannot possibly be earlier than 450.

³³ Bergk's emendation κατ' ἀγλαὸν ὤλεσαν ἤβην restores the caesura, but the extremely rare κατόλλυμι with its notion of ‘utter destruction’ could never have stood here. We want ἀπὸ, and we have it.— I once thought of ποταίνιον ὤλεσαν ἤβην (cf. ηεβεν ολεσαντα in IG. I ². 976Google Scholar): but I am sure this is too individual for a public epitaph.

³⁴ Which are, of course, sensibly stricter for elegiacs than for hexameter verse. A false caesura such as the MSS. impute to our poet occurs, e.g., in the Delian Hymn (Hymn. Hom. III), line 53.

³⁵ He may have been predisposed that way by IG. I². 943 = [N]: unless indeed we are to think that the epigram was current in (incorrect) book-form already by 440 [or 408] and is responsible for αγλοον ηεβεν in the later poem. [For the early currency of books of epigrams (in Plato's early life), see, e.g., Geffcken, -Herbig, in Glotta, IX. 1918, p. 100Google Scholar, Weber, in Hermes, LII. 1917, p. 540Google Scholar.] Ἀγλαόσ is not normally feminine, but the phrase ἀγλαὸσ ἤβη occurs in Theognis 985 in a memorable poem. Τίμιοσ ἤβη (feminine, e.g. Soph., Ant. 948Google Scholar) recalls, of course, Mimnermos' ῆβη τιμήεσσα. The two words τίμιοσ and τιμήεισ have the same range of meaning; but Mimnermos probably means ‘honoured by others,’ our poet ‘precious to the owner.’ Cf. further Kallinos, 1. 6, and Mimnermos, 3. 2.

³⁹ Boas, p. 219, believes this to be a mere error for which stands in the title of the preceding epigram where we might expect the two words, he thinks, have been simply exchanged. This is one of many reasons for supposing that these two poems are from the same source.

⁴⁰ I think no one who knows the Attic style will dispute that the burden of proof lies with anyone who suggests this is fifth-century Attic. I am confident that the frigid and sentimental second couplet was not written, in Athens or elsewhere, within a century of the battle. is a special mark of the ‘refined rhetoric’ of Alexandrian poetry: contrast the last couplet of Simonides' (Anth. Pal. VII. 496 = Simonides 80 in Diehl's Anth. Lyr.) with Kallimachos', imitation (Anth. Pal. VII. 271Google Scholar).]

For those who like something more specific than the test of style, the use of for blood may serve. Its usual sense is light rain, drizzle, e.g. Hdt. III. 10, Xen., Cyneg. V. 4Google Scholar: in comedy of a sputtering talker or a spot of money (Arist., Ach. 1150Google Scholar, Peace 121). In the Agamemnon Klytaimestra thus describes Agamemnon's death:

There is little question but that one passage is copied from the other, and even less question which is the original. The brutal phrase grows out of its context in Aeschylus: in the epigram it (like the other borrowings ) only adds sonority and circumstance to the statement, These men were once killed by arrows [cf. in the original epitaph]. The Eurymedon battles were fought not later than 466, the Agamemnon was produced in 458. The rain of blood comes in a later passage of the Agamemnon (1534) with a clear echo of Klytaimestra's words: I think nowhere else in literature.

⁴¹ See Section IV below, ‘A Hellenistic Monument of the Eurymedon.’

⁴² The absence of proper names is striking; but they are equally absent from Mnasalkas' verses on Thermopylai, (Anth. Pal. VII. 242Google Scholar: I do not know why Mackail, , Select Epigrams, III. 5Google Scholar, refuses the reference to Thermopylai: the poem is certainly pastiche).

⁴³ Anth. Pal. VII. 442: i.e. the poem next before [O]. It is clearly similar in nature to [O] (see note 39 above) and is likewise ascribed to Mnasalkas by Boas.

⁴⁸ Also by Plutarch, , Kimon, VII. 4Google Scholar, whose text is better, and so is presumably drawn from some source independent of Aeschines: see note 76.

⁴⁹ Also by Aelius Aristeides twice (Dindorf, Vol. II, pp. 209 and 512) and in the Palatine Anthology, VII. 296. The variants are given in full by Boas, (De ep. Simon. 104Google Scholar sqq.) and Weber, (Philologus, N.F. 28Google Scholar [1917], 249). Diodoros seems to be the least corrupted: but the others are not derived from him and we can consequently use them as independent evidence. I desert Diodoros in [Q] line 1; is certainly wrong, Aristeides gives In [R] line 1 I have preferred (Aristeides and Anth. Pal.) to Diodoros' the partitive genitive is common in fifth-century Greek and I think improves the verse. Aristeides' in the same line (vice ) is quite worthless as evidence: it attempts the impossible task of harmonising [Q] line 4 and [R] line 1. See, however, note 65 below.

⁵⁰ I cannot indeed agree to all: e.g. pp. 17 and 18, or note 74, or his general topography.

⁵¹ It is quite certain that the sea-battle was fought close to the Eurymedon River: see below, Epigram [W].

⁵² Or which accepts the late variant for See, however, note 65.

⁵³ The repetition of the double event is striking and would no doubt be rejected as a doublet by many scholars, were the evidence less good.—The actions were not in fact very similar: at the Eurymedon the driving ashore of the fleet was preliminary to the great land-fight [see, however, note 65]: in Cyprus, the defeat of the land-force was preliminary to the sinking of the reinforcing fleet. But the poem stresses precisely the thing which they have in common.— That the sea-fight came firs, at the Eurymedon, is commonly accepted (since Meyer) from Plutarch's account (Kimon, XII–XIII), and is perhaps confirmed by [W] (see below) if I am right in believing Maiandrios was killed before the land-fight began: nevertheless the land-fight is almost always mentioned first, e.g. in line 4 of [Q], in line 3 of, [M], in Thuc. I. 100. 1, and Lycurg, . Leocr. 72Google Scholar: presumably because it was the more important.

⁵⁴ The Battle of Mykale had been fought on sea and shore in one day, yet had been on nothing like the scale of the Eurymedon. I imagine the poem de liberately invites comparison with Mykale.

⁵⁵ Ephoros apparently understood it so. But against him I must appeal to the principle stated at the opening of this paper.

⁵⁶ As opposed to Epeiros proper (Od. 14. 97, etc.), the Troad, (Il. 1. 485Google Scholar), Boeotia (Hesiod, , Erga, 624Google Scholar), etc.

⁵⁷ I imply, of course, no etymological relation between ἤπειροσ and ἀπειρέσιοσ.

⁵⁸ Except in the latter books of the Odyssey, where it means definitely Epeiros.

⁵⁹ Just as, in Od. 14. 100, ἐν ἠπείρῳ means ‘in Epeiros.’

⁶⁰ Liddell and Scott, Lexicon ⁹, quotes this (wrongly) as an instance of ἐν ἠπείρῳ = on shore.

⁶¹ When Thucydides calls the Sicilian cities τῆ παρασκευῆ ἠπειρώτιδασ, ‘their resources are on a continental scale,’ he is precisely not calling Sicily an ἤπείροσ.

⁶² Cf. Meyer, , Forsch. II. p. 11Google Scholar; Weber, , Philologus, LXXIV, pp. 248Google Scholar, 257 (note).

⁶³ The poem has commonly been treated with contempt: or damned (as by Meyer, p. 13, Weber, p. 251) with faint praise.

⁶⁴ A double action, on sea and land: unexampled: on the Continent of Asia.

⁶⁵ Uxkull-Gyllenband (a pupil of von Domaszewski, ), in his Plutarch und die Griechische Biographie (Stuttgart, 1927), p. 55Google Scholar, denies that the narrative demands ἐν Κύπρῳ, insists that Aristeides is as good evidence as Diodoros for Ephoros' text, and therefore retains ἐν γαίῃ, and refers the whole eight lines to the Eurymedon. His whole treatment of the Eurymedon narratives (pp. 45–59) is stimulating and important, and contrary to the view taken in this paper: he thinks the land battle preceded the sea battle [see, however, my note 53]. I think he is wrong; the land battle was first not in time but in importance: and ‘Ephoros'’ [Diodoros'] narrative, though not true to the poems [Q] and [R], has yet been warped by them, and why should anyone change γαίῃ to Κύπρῳ? But I am so uncertain about [R] that I willingly leave the matter in suspense.

⁶⁶ See Geffcken, and Herbig, in Glotta, IX. 1918, p. 100Google Scholar (note 35 supra).

⁶⁷ Meyer, p. 19, says the monument (at Delphi) referred to Cyprus, but being in effect Kimon's memorial, was later thought to refer to the Eurymedon. Weber, p. 256, says it was an epitaph, but lines 3–4 (of [Q]) sounded very like the Eurymedon.

⁶⁸ The fiction occurs in the spurious last couplet of Simonides 90 (in Diehl, Anth. Lyr. Gr.).

⁶⁹ Cf. the dedication of the Nike, Temple: IG. I ². 24Google Scholar [= Tod, , Sel. Gk. Hist. Inscr. 40Google Scholar (with improved text)]: Welter, , Vom Nikepyrgos in AM. XLVIII, 190Google Scholar sqq., makes it certain that the Nike Bastion was constructed in its present form before the foundations of the Propylaia were finished.

⁷⁰ See above, p. 74.

⁷¹ See notes 35 and 92.

⁷² See, e.g., Keramopoullos, in AM. 34 (1909Google Scholar), 40 sqq., Pareti, in Studi sicelioti ed italioti (Florence, 1920), 173Google Scholar sqq. The Stoa of Zeus is to be further excavated this year, AJA. 36 (1932), 383.

⁷³ III. 183 sq.: the Greek is quoted in note 14 above.

⁷⁴ See the text above, notes 11–14.

⁷⁵ Can we take Aeschines (III. 186) so literally as to assert that Miltiades' name was not written beside him?

⁷⁶ Hypereides is evidently using the same argument as Leptines and Aeschines. He may well be the source (similar to Aeschines but without his corruptions) from whom Plutarch, Kimon VII, derives.

⁷⁷ This is well known for the Basileios and the Poikile: it is made certain for the Herm-Stoa by Aesch. III. 186.

⁷⁸ It is just possible that the Herm-Stoa was destroyed and replaced by Attalos' Stoa: προέλθετε in Aesch. III. 186 (spoken from the Heliaia, south of Attalos' Stoa?) might be held to confirm this. If so, traces of the fifth-century building should be discoverable.—Hobein in Pauly-Wissowa, Art. Stoa, col. 19–20, suggests that the Herm-Stoa, is the same as the Poikile! He has evidently overlooked Aesch. III. 186Google Scholar, where the Dikasts are invited to walk in imagination from one to the other.

⁷⁹ Basileios I. 3. 1: Zeus' I. 3. 3: Poikile I. 15. 1.

⁸⁰ The three Stoai in Aristoph, . Ekkles. 684–6Google Scholar are evidently the Basileios and Zeus' (τὴν παρὰ ταύτην) and the Alphitopolis. [If the initials B Θ K conceal nicknames, we can hardly hope to guess them.] The last (if Judeich rightly identifies it with the Makra) is not in the Agora but westward of it, between the Dromos and the ‘Theseum’ hill: Judeich, , Topogr. ², 365Google Scholar. It is a group of three, on the west edge, or westward, of the Agora.

⁸¹ Codd. the Epitome has Suidas s.v. has

⁸² Neither Menander nor Didymos (as quoted) says explicitly that the title Eleutherios was added later. It is possible that from the first the cult title was Soter, the popular title Eleutherios. Yet I find this unlikely.

⁸³ The parallel citations have schol. Plat. 392A, Et. M.] Suidas. I suppose Harpokration to be nearest to Didymos' words: Suidas and Et. M. by their variations seem to be refer ring the phrase to 479 B.C.

⁸⁴ Sic: not του Ελευθεριου.

⁸⁵ Cf. Thuc. V. 111. 2, VIII. 53. 3, Lysias, XII. 69.

⁸⁶ Thuc. II. 71. 2 (cf. schol. Plat., Eryx. 392Google ScholarA, where for read , and for in schol. Paus. I. 3. 2, ). The ‘Simonidean’ epigram was, I imagine, inscribed later: (Simon. 107 Diehl).

⁸⁷ I.e. in the year following the battle. The date 469 for the Eurymedon I take from Plut, . Kimon VIII. 7–9Google Scholar: it seems to me almost certain that the honour conferred on the Strategoi at the Dionysia of 468 was consequent upon that victory. See Weizsäcker, , Untersuch. über Plut, biogr. Technik (Heft 2 of Problemata, 1931), 61–64Google Scholar. I cannot follow Weizsäcker in details, but his main thesis, that Plutarch is by temperament a moralist and by purpose an historian, and consequently chooses his material ‘eidologisch’ and presents it (falsely) chronographisch,’ seems to me true and illuminating. [The Eurymedon date is discussed by Taeger, F., Ein Beitrag zur Gesch. d. Pent., Stuttgart, 1932Google Scholar.]

⁸⁸ Sparta had freed her thousands, but Athens her tens of thousands.

⁸⁹ I base this hypothetical judgment chiefly on the tone of Isokrates' Panegyrikos: esp. such passages as ch. 179, where he seems to have his eye on the Eurymedon-Herm, whilst contrasting the Kallias and Antalkidas treaties. It is typical too, I think, that the Decree of Aristoteles, (IG. II ². 43Google Scholar), which directly challenged the Antalkidas Treaty, was placed beside this Stoa. [Peace of Antalkidas 386, Panegyrikos 380, Decree of Aristoteles 379.] It was, in fact, designed as the Athenian counterpart to the Spartan memorial at Plataea, to symbolise Athenian hegemony in rivalry to Spartan. Note that in IG. II². 43, line 65, the title Eleutherios is used: contrast IG. II². 689 and 690 (note 84 above).

⁹⁰ For the appearance and functions of Herms, see Curtius, L., Die antike Herme (Munich Dissertation, Leipzig, 1903)Google Scholar, S. Eitrem, Art. Hermai in Pauly-Wissowa, , Lullies, R., Die Typen der griechischen Herme (Heft 3 in Königsberger Kunstgeschichtliche Forschungen, 1931)Google Scholar. From the use of Herma in Cic. De legibus 2. 26. 65, and ἑρμογλυφεῖον in Plat, . Symp. 215Google ScholarA, it is possible that the word had sometimes a more general sense than we give it (see Lullies, op. cit. 59 sq.), and the Herms in the Agora may have included various types of small sculpture.—Judeich rightly complains that there is no evidence at all for von Domaszewski's thesis of ten Tribal Herms in the Agora, (Topogr. ²336Google Scholar, note 2).

⁹¹ Cf. Solon, fr. 5, line 3 (Diehl), μηδἑν άεικὲσ: ‘no injury,’ sc. for those whom they lead. I think we should emend, Ἀθηναίουσ καλέεσθαι.

⁹² The evidence in Glotta, IX. 100. See note 35 above.

⁹³ His actor's training had no doubt given him a fluent and inexact memory. Χαλκοχιτώνων in [T] 3 is clearly an actor's extemporisation: so is the commonplace κρατερόν in [E] 3.

⁹⁴ See the first discussion of [E].

⁹⁵ The virtues of Menestheus were suited to a pitched battle rather than a siege.

⁹⁶ This was inscribed on some herm γράμμασιν ἀρχαίοισ, i.e. with three-stroke sigma, etc. It cannot therefore be a mere quotation from comedy?

⁹⁷ I see no reason for dropping the last couplet whether of [S] or [T], as von Domaszewski wishes. On the other hand, I find the first line of [S] very difficult: what is τάδ'? I could understand, e.g., Ηεγεμονι μνεμειον Αθεναιοπ τοδ εθεκαν, as an inscription of 449 dedicating the whole monument to Kimon's memory: but one cannot start rewriting the poems. Τάδ' may refer to something quite different, possibly in the Basileios or Poikile.

⁹⁸ Lycurg, . Leokr. 109Google Scholar: Simonides, 88 Diehl: H. v., Gaertringen, , Hist. gr. Epigr. 12Google Scholar. Perhaps too, if (as is possible) Polygnotos signed his Attic ‘Sack of Troy’ in the same way as his Delphian, we may add Simonides 112 Diehl (= Paus. X. 27. 4),

⁹⁹ Αστοι (as regularly, e.g., in Pindar) means the heroes' fellow-citizens; Μεγαρησ is redundant and clumsy. ξυνον, besides its obvious fitness, has in this context a further emotional fitness, ‘we are all Megarians together’: cf., e.g., Pind, . Pyth. IX. 93Google Scholar [Ol. VII. 90–3, Pyth. XI. 52–4].

¹⁰⁰ Klaffenbach dates the writing (by comparison with the Samian, Berenike inscription, AM. 44 [1919], 21Google Scholar) to about 250 B.C. The poems are simple, quite out of the Kallimachos or Anyte-Mnasalkas tradition. I imagine they are provincial.

^100a For the blocks of figs. 2 and 3 the editors are indebted to the Dieterich'sche Verlagsbuchhandlung, Leipzig.

¹⁰¹ The list can hardly have been long enough to contain all who fought.—I note that among the patronymic genitives in the first column cannot be , since the is used for rather,

¹⁰² Conceivably we should restore, e.g., επι καλωι—Ευρυμεδ[οντοσ ελει].

¹⁰³ To judge by the Attic epitaph ([M] 3 supra, ), and the account in Plutarch, , Kimon, XIII. 2Google Scholar. Cf. Meyer, , Forsch. II. 6Google Scholar sq., and see note 65 above.

¹⁰⁴ If the original fifth-century monument contained a list of fallen (note 101, supra), Maiandrios' name may have been among them. Did a man-statue (of Maiandrios?) surmount the original monument (cf. Hdt. VIII. 121, and the Note on Akroteria)?

¹⁰⁵ I think Mr. Blakeway first suggested to me that Nike's akroteria were aphlasta; Mr. Lobel called my attention to the akroteria of Artemis in Plut. Phokion, XXVIII.

¹⁰⁶ Possibly some Aeginetan battle; possibly even the Euripos, where some Chalkidian ships may have been captured. Could τον πολε[μιο]ν refer to Persians?

¹⁰⁸ Fig. 4, for which ProfBeazley, has given me his drawing. Photograph in CVA. London, III. Ic. Pl. 52, 2Google Scholar.

¹⁰⁹ Prof. Beazley also tells me of a goddess with aphlaston on a weak-Brygan lekythos in New York, unpublished.

¹¹⁰ Ἀκρωτήριον can also be used of a ship's beak, and these beaks be used as trophies, Hdt. III. 59. [Such ‘prows like boars’ heads' can be seen, e.g., in Pfuhl, , Mal. und Zeichn. Nos. 231, 233Google Scholar, 259: 231 and 259 also good aphlasta.] But the akroteria carried in the hand are always stern-ornaments. [Cf. πρύμνης, ἀκρωτήριοα, Hymn. Hom. XXXIII, 10Google Scholar.]

¹¹¹ I note that in Ἀριθ. 1, line 39, and Ἀριθ. 2, line 19, there should be a comma between ακρωτηριον and οπισθιον: they are separate items, the latter is the һοπισθιον, χρυσιον of IG. I ². 329Google Scholar, line 12. This χρυσιον һοπισθιον I imagine included the wings.

¹¹² The plural might be used of the ornament of a single stern. Herodotos speaks of ἄφλαστα νεός (VI. 114).

¹¹³ He is, so far, like the statue of Hdt. VIII. 121.—There is some awkwardness in expression, if the shield had eight aphlasta and the hand only one; but I think this must be meant.

¹¹⁴ This shows little in the photograph. It is unmistakable on the stone or in a squeeze.

¹¹⁵ The fact that the later line 1 is inscribed more or less stoichedon with line 2, misled me (in Pythian Odes) into thinking they were inscribed together: it is, however, amply clear that line 2 is original, and the later line 1 has been adapted to it. The earlier line 1 has a four-horizontal E: we must expect its script to be identical with that of line 2.

¹¹⁶ Cf. supra, note 98, and in H. v. Gaertringen's Hist.gr. Epigr., No. 40, 38, 37, 33, lines 3, 6; 8, line 1; 5, line 1; 4, lines 3,4.

¹¹⁷ I need hardly perhaps discuss here Pareti's strange heresy that Gelon was never (or hardly ever) King of Gela. Though accepted by Hackforth in the CAH., it is impossible to reconcile it with Herodotos' account of Hippokrates' reign. That king spent his reign creating the Geloan Empire (Hdt. VII. 154): so that his appearance before Zankle as its suzerain must come towards the end, and yet not much later than the Battle of Lade.—Yet I would like to express once more the debt I owe to Pareti's Studi sicelioti ed italioti, a rich and stimulating book.

¹¹⁸ Almost certainly, since otherwise he would have succeeded to Hieron's throne. In any case, the tone of the inscription and style of the statue alike put 466 (or later) out of the question.

¹¹⁹ This reconstruction is given in its fuller context in my Appendix II (pp. 161–4., cf. p. 64) to the Pythian Odes of Pindar. It is much to ask scholars to look at a book whose edition was limited and where there is no documentation. I hope some day to remedy these things, but this is not the place.