¹

² Rose, , Ath. Pol. p. 423Google Scholar, accepted by Sandys, , Ath. Pol. ² p. 107Google Scholar.

³ Kenyon, ad loc.

⁴ Thuc. i. 135–8; Plut., Them. 22 f.Google Scholar; Diod. xi. 54–59 (Ephorus); Corn. Nep., Them. 8–10Google Scholar (mainly Thucydides).

⁵ E.g. Holm, Hist. Gr. ii. p. 94Google Scholar; Meyer, E., Ges. d. Alt. III. i. p. 519Google Scholar.

⁶ See below, p. 171, n. 27.

⁷ Mitchell and Caspari in their edition of Grote, p. 283, n. 1; Cauer, F., Deutsch. Literaturzeit. 1894, p. 942Google Scholar. Cp. also (inter alios) Bérard, , Rev. Hist. xlix. (1892) p. 296Google Scholar; Botsford, , Cornell Stud. Class. Phil. 1893, p. 220Google Scholar, n. 2; Busolt, , Gr. G². III. i. p. 29Google Scholar; Costanzi, , Rivista di Fil. 1892, pp. 353–5Google Scholar; Dufour, , Const. ďAth. p. 113Google Scholar; Giles, , Eng. Hist. Rev. 1892, pp. 332–3Google Scholar; Meyer, E., Oes. d. Alt. III. i. p. 519Google Scholar; Niese, in Sybels Hist. Zeits. xxxiii. (1892), p. 43Google Scholar; Reinaeh, Th., Rev. Et. Gr. 1891, pp. 149–151Google Scholar; Ruehl, Rhein. Mus. 1891, p. 431Google Scholar; De Sanctis, , Stud. Costituz. ď Atene, pp. 4–6Google Scholar, and Rivista di Fil. 1892, pp. 108 f.; Sandys, Const. of Ath ². p. lxix.; Walker, , C.R. vi. pp. 95–99Google Scholar; Wilamowitz, , Aristot. u. Athen, i. pp. 140 fGoogle Scholar.

⁸ Reinach, Th., C.R. Acad. Inscr. June 1891, and Rev. Et. Gr. 1891, pp. 149–151Google Scholar, cp. Répub. Athénienne, p. 46; Buseskul, , Aristot. Ath. Pol. Their arguments are answered by Politis, Parnassos, 1893, pp. 95–6Google Scholar, and Schoeffer, , Jahresber. Fortschr. cl. Alt. lxxxiii. (1895), pp. 220–1Google Scholar. The substance of this paper will make it unnecessary to revert to them in detail.

⁹ See the list given by Th. Reinach, Rep. Ath. pp. xxvi–xxvii (Cimon ‘youngish’ in 462 B.C., c 26; Spartan peace proposals put after Arginusae instead of Cyzicus, c 34, cp. Philoc., F.H.G. i. fr. 117–8Google Scholar; all the generals put to death after Arginusae instead of all who were put on trial and appeared before the court, c 34; confusions or contradictions in the accounts of the c. 6, and c. 54, cp. c. 48). For the Ath. Pol. drawing inferences, sometimes wrong, as to early constitution-usages, see Swoboda, , Arch. Epig. Mitt. xvi. pp. 57Google Scholar f. on Ath Pol. 16. 10. It is, of course, easy to find in the Constitution's account of the fifth century much that is ‘palpably legendary,’ De Sanctis, , Studi Cost. Aten. p. 11Google Scholar, if we regard as such any new information that disagrees with our preconceived conceptions of the period. For a good protest against this attitude, see Politis, , Aristot. Ath. Pol. in Parnassos, 1893, p. 13Google Scholar.

¹⁰ Cp. Plut., Cim. 5Google Scholar, where Cimon, the protégé of Aristides, is described as ‘inconceivably the superior” of Themistocles as a statesman.

¹¹ Forsch. Ath. Pol. p. 171 f.

¹² C.R. vi. pp. 95–99. Against Bauer's chronology, see also Ruehl, , Jahrb. Cl. Phil. Suppl. xviii. (1892), p. 695Google Scholar.

¹³ Jahresber. Fortschr. cl. Alt. lxxxiii. (1895), p. 251, cp. Nordin, , Stud. i. d. Themistoklesfrage, p. 61Google Scholar.

¹⁴ Wilamowitz, , Aristot. u. Ath. i. 149Google Scholar. The whole question as to how far by the time of Aristotle or even Thucydides Themistocles had won his way into the fabulous is beyond the scope of this article.

It is certainly exaggerated by Wilamowitz in the work just cited. It is one thing to show that a historical character has become the victim of unhistorical anecdote; it is another to determine whether or to what degree the anecdotes in question are free to violate the historical setting.

¹⁵ Unsere Zeit, 1891, ii. pp. 28–9; cp. Seeck, O., Klio, iv. (1904), pp. 302–3Google Scholar.

¹⁶ Thucydides quotes (i. 138) relatives of Themistocles as stating that his bones were brought back to Athens and secretly buried; but it does not follow that the historian was able to get full information about the life of the dead statesman from this source. The only relative of Themis tocles known to have remained in Athens, his son Kleophantos (Plato, , Meno, 93)Google Scholar, was notoriously interested in nothing but horses and athletics.

¹⁷ Note, too, that Ath. Pol. 18 tacitly corrects Thuc. vi. on several points in the Harmodius story, and that it has been claimed by Weil, , Journ. d. Sav. 1891, p. 203Google Scholar, that Thucydides himself in i. 20 appears to realise the mistakes of the Book VI. account, which is probably the earlier. Thucydides, is also corrected by the Ath. Pol. (31–3)Google Scholar on points of detail about the four hundred : Weil, ibid. p. 204.

¹⁸ Ephialtes is much the more prominent all through the chapter. Where both are mentioned together, Ephialtes is put first. Themistocles has merely a share in the re sponsibility, The same inference is to be drawn from c. 41, In these last words Mathieu, , Bibl. École Hautes Etudes, 216 (1915), p. 64Google Scholar, wrongly finds traces of a tradition according to which Ephialtes was not aided by Themistocles.

¹⁹ Compare the remarkable omissions in Thucydides' synopsis of the history of Athens between the Persian and Peloponnesian wars, i. 97 f. ‘There is nothing about the political measures of Ephialtes or Pericles, nothing about the divisions of opinion on the question of sending help to the Spartans at Ithome, nothing about the ostracism of Cimon or the political activity* of Thucydides the son of Melesias: events either closely connected with external affairs, or so important that they might have seemed to demand mention in the most cursory sketch of the period.’ Forbes, Thuc. i. p. cxvii.

²⁰ The duration of ostracism is given as ten years by Plato, , Gorg. 616Google ScholarPlut., D.Cim. 17Google Scholar, Nic. 11, Corn. Nep., passim, pseudo-Andoc. iv. 5Google Scholar, schol. Aristoph. Vesp. 947; cp. Theopomp., F.H.G. i. p. 293Google Scholar, Cimon was recalled from ostracism If the sentence had been for five years we should expect not but Carcopino, , Bib. Fac. Lett. Paris, xxv. p. 117Google Scholar. Diod. xi. 55.2 gives it as five and Philochorus fr. 796 as originally ten, later five. If Diodorus is not confusing with the Syracusan petalismos, which he also describes, he might be explained by Philochorus, but note that the last victim of ostracism, Hyperbolus, was ostracised for six years, Theopomp., F.H.G. i. p. 294Google Scholar. Cimon seems certainly to have been sentenced for ten years. If, therefore, the length of the sentence was ever shortened it was presumably after the time of Themistocles.

²¹ Plut., Them. 23Google Scholar, de Exil. 15 (Moral. 605 E); Krateros, F.H.G. ii. 619Google Scholar, fr. 5. For Alcmaeonid hostility to Themistocles, see also Plut., Praec. Ger. Rep. 10Google Scholar (Moral. 805 C), Aristid. 25.

²² Cp. Thuc. v. 26.

²³ Plut., Them. 26Google Scholar.

²¹ So Wilamowitz, Aristot. u. Ath. i. p. 150Google Scholar, n. 47.

²⁵ This date accords quite as well as any other with the meagre evidence, which is fully set forth by Busolt, , Gr. G ². III. i. p. 113Google Scholar.

²⁶ Diod. xi. 54 speaks of two trials, the first at Athens before the ostracism, ending in acquittal, the second at Sparta, after the ostracism, resulting in Themistocles' flight and condemnation. Diodorus' evidence is not decisive; he assigns the events of a number of years to the single year 471–470 B.C., and makes the unlikely statement that the trial that drove Themistocles to Asia took place at Sparta; but an early trial and acquittal can be easily reconciled with the order of events suggested above.

²⁷ Artaxerxes in 462 B.C. might still be newly on the throne' from the point of view of Thucydides writing after the close of his long reign of 40 years. The version that brings Themistocles to Persia before the death of Xerxes may be dismissed (so e.g. Bauer, , Forsch, p. 69Google Scholar) as a poetic emendation of the facts. The flight to Persia is indeed dated 471 B.C. by Diod. xi. 54–6, and 472 B.C. by the Armenian version of Eusebius, but their evidence is weak: on Diod. see note preceding; Euseb. is probably based on Diod. The flight is probably misdated by either writer to the year required by his chronology for the ostracism (in which case we have here a further possible explanation of the double dating with a ten years' difference already noticed in the chronology of Themistocles, by Munro, J. A. R., C.R. vi. (1892) pp. 333–4Google Scholar. On Cic., de Amie. 14, 42Google Scholar, which has been thought to confirm Diod. and Euseb., see below, p. 177. Wilamowitz, , Aristot. u. Athen, I. pp. 143–4Google Scholar and Busolt, Gr. G. III. i. pp. 113Google Scholarn, 128, accept 471 B.C., but their arguments are flimsy, based on the assumption that the three authorities who alone give a definite date to the flight are based on contemporary documents, notably the and copies made by Krateros of Athenian decrees. But because Krateros is known to have published the charge brought against Themistocles, it hardly follows that Diodorus derived from him the date of Themistocles' flight. As regards the it is rather remarkable that they are never once mentioned in connexion with Themistocles. If they are to be used at all as evidence, that is one point that must be taken into account. Can the explanation be that the trial and condemnation took place, as the dating proposed in this paper implies, during a comparatively brief setback in the progress of the party to which Themistocles belonged, and that consequently his name never got posted up?

²⁸ So e.g. Walker, E. M., C.R. vi. p. 96Google Scholar and Kenyon ad loc.

²⁹ Ath. Pol. 25. 2.

³⁰ This year could easily be regarded, especially on an inclusive reckoning, as “about 17 years after the Persian wars,” which is how the Constitution dates the beginning of Ephialtes', attacks, Ath. Pol. 25.1Google Scholar. See further, Hertlein, , Korrespondenz Blatt f. d. Gelehrten-u. Realschulen Wuerttembergs, 1895, pp. 2–3Google Scholar.

³¹ The word is rendered by the translators (Th. Reinach, Haussoullier, Poste, Dymes, Zuretti, Ferrini, Poland, Kaibel and Kiessling, Erdmann) by such words as associé, concours, co-operation, conjunction, cooperatore, compagno, Unterstützung, beteiligt, Mitwirkung. But the Greek for this would surely be some such word as or

³² See previous note.

³³ The sentence might perhaps be completed in some such way as this: Kaibel, Stil u. Text d. Πολ. Αθ pp. 182–3Google Scholar states that the words imply that the missing words told of the death of Themistocles: but may not mean simply ‘was removed?’ cp. Ath. Pol. 25. 2

³⁴ Kenyon, , ad. Ath. Pol. 26. 2Google Scholar.

³⁵ Pace Kemach, Th., Rev. Ėt. Gr. 1891, p. 156Google Scholar. For the presumed change in the form of attack cp. Ath. Pol. 25. 3 and above n. 1 with F.H.G. II. p. 619.

³⁶ Walker, E. M., C.R. vi. p. 97Google Scholar.

³⁷ Diod. xi. 70, apparently dates the fall of Thasos in the archonship of Archidemides, 464—3 B.C., but (pace Cauer, Hat. Aristot. p. 27Google Scholar) he may be like the moderns, merely making an inference from Thucydides.

³⁸ And perhaps also to recall Cimon from Sparta to take the command (Plut., Cim. 14Google Scholar). The chronology of this part of Cimon's career is difficult, but it seems on the whole most probable that the urgency of the situation in Thasos was the reason why Cimon came back from his first Spartan relief expedition in so great a hurry that he had not even time for the usual civilities to the states through which he passed en route.

³⁹ The MSS. vary between and The Teubner and new Oxford texts both print But is the difficilior lectio and has the support of a good group of MSS. It is read by Forbes.

⁴⁰ For examples of unsatisfactory chronology in Plutarch see his accounts of democratic developments at Athens, , Cim. 15Google Scholar, the two expeditions of Cimon to Sparta, , Cim. 16Google Scholar f., and the various occasions on which he returned from active service to Athens, , Cim. 14, 15, 17Google Scholar: cp. also Them. 5–6, where the choregia of Themistocles in 476 B.C. is mentioned just before the account of 480 B.C. and Salamis.

⁴¹ Plut., Per. 7Google Scholar.

⁴² Plut., Per. 16Google Scholar.

⁴³ Cic., de Orat. iii. 34, 138Google Scholar.

⁴⁴ Plut., Praec. Ger. Rep. 15Google Scholar (Moral. 812 C).

⁴⁵ Plut., Per. 7Google Scholar.

⁴⁶ Plut., Cim. 14Google Scholar.

⁴⁷ Plut., Per. 7Google Scholar.

⁴⁸ Pace Busolt, Gr. G ² III. i. p. 113Google Scholarn. The difficulties raised by Corn. Nep. mirisi, fin., which dates the death of Aristides “fere post annum quartum quam Themistocles Athenis erat expulsus’ need not be here discussed.

⁴⁹ Plut., Arist. 26Google Scholar.

⁵⁰ That Ephialtes had been the master of Pericles would have been forgotten the more easily since the position of or to Pericles was commonly ascribed to Damonides or Damon, , see Ath. Pol. 27. 4Google Scholar, Plut., Per. 9. 4Google Scholar. The latter quotes Plato Comicus on Damon:

⁵¹ Aristot., Pol. ii. 1274Google Scholar A.

⁵²

⁵³ Sandys, , Ath. Pol. ² p. 109Google Scholara.

⁵⁴ Pace Ruehl, Rhein. Mus. 1891, p. 433Google Scholar. As to why Plutarch may have omitted, see further (in spite of his mistaken chronology), Bauer, , Forsch, p. 82Google Scholar, and Nordin, , Stud. i.d. Themistoklesfrage, pp. 62–3Google Scholar.

⁵⁵ Plut., Them. 24Google Scholar.

⁵⁶ Plut., Them. 18, cp. 24Google Scholar.

⁵⁷ Cic., ad Fam. v. 12. 5Google Scholar.

⁵⁸ See Tyrrell and Jeans ad loc.

⁵⁹ Tyrrell quotes Brut. 43, ad Att. ix. 10. 3, de Amie. 42. The death of Themistocles is mentioned in the pro Scauro (54 B.C.), but in a context that deals with the subject of suicide.

⁶⁰ Cic., de Amie. 42Google Scholar.

⁶¹ See Cic., Brut. 43 (46 B.C.)Google Scholar.

⁶² Mention should perhaps be made of Boot's neat emendation ‘reditusque spe tenetur (cp. Purser Script. Class. Bibl. Oxon., ad loc.); but though ingenious this emendation is as untenable as the rest. The context requires a reference to an actual return.