¹ A Historical Commentary on Thucydides i (Oxford 1945) 401.

² Some of the most recent treatments where the ancient evidence, earlier bibliography, and full discussion of the problems may conveniently be found are: Gomme, , Commentary i 397–401Google Scholar; Meritt, B. D., Wade-Gery, H. T., and McGregor, M. F., The Athenian Tribute Lists (ATL) iii (Princeton, N. J., 1950) 111–13, 200–1Google Scholar; Highby, L. I., The Erythrae Decree, Klio, Beiheft 36 (Leipzig 1936) 81–98Google Scholar; Schaefer, H., RE 18 (1949) 2563–78Google Scholar, s.v. ‘Pausanias’ 25, 21 (1951) 195–6, s.v. ‘Pleistarchos 1’, 205–9, s.v. ‘Pleistoanax’; Flacelière, R. C., ‘Sur quelques points obscurs de la vie de Thémistocle’, REA lv (1953) 5–28CrossRefGoogle Scholar, esp. 5–14; Lenardon, R. J., ‘The Chronology of Themistocles' Ostracism and Exile’, Hutoria viii (1959) 23–48Google Scholar; Forrest, W. G., ‘Themistokles and Argos’, CQ n. 5. x (1960) 221–41.CrossRefGoogle Scholar

³ The usually accepted date for Tanagra is 457, see Gomme, , Commentary i 313, 325–6Google Scholar; the editors of ATL iii 171–3 argue for a year earlier, 458. Diodoros (xiii 75.1) places the death of Pleistoanax in 408/7 after a reign of 50 years, and so his accession in 458/7; but at xiv 89.1 the exile of his son the younger Pausanias is placed in 394/3 after a 14-year reign. This is a year too late: see Xen. Hell. iii 5.25 for the defeat at Haliartos in 395/4 and his flight to avoid the death sentence on his return. The same discrepancy of a year may have affected Diodoros' dates for Pleistoanax, so that his accession and Pleistarchos' death may have been in 459/8 rather than 458/7; cf. Ed. Meyer, , Forschungen ii (Halle, 1898) 503, 509–11Google Scholar on possible reasons for this and other difficulties in this portion of Diodoros' Agiad king-list. Nikomedes' command at Tanagra raises an interesting question, why did he command rather than Archidamos, the Eurypontid king? It is tempting to conjecture that Archidamos was needed at home because the Helot revolt was not yet completely quelled, so that we have here an additional reason for not emending Thucydides' δεκάτῳ ἔτει (i 103. i); cf. Gomme, , Commentary i 402–13.Google Scholar Professor Andrewes has pointed out to me that it is also surprising that the young and inexperienced Pleistoanax commanded the important invasion of Attica in 446 rather than Archidamos, and makes the interesting suggestion that Archidamos ‘was Kimon's friend and had no stomach for ventures against Athens’, and that Kimon's earlier dismissal from Ithome could be due as Thucydides suggests to the actual feeling among the Athenian troops rather than to any distrust of Kimon.

⁴ This seems to be the meaning of the schol. ad. Her. ix 85, and the gloss in Strabo's Geography published by Diller, A., AJP lxii (1941) 499–501Google Scholar; see the recent discussions by Billheimer, A., TAPA lxxvii and lxxviii (1946 and 1947) 214–20, 99–104Google Scholar; Chrimes, K. M. T., Ancient Sparta (Manchester, 1949) 84–136Google Scholar; Boer, W. den, Laconian Studies (Amsterdam, 1954) 248–61.Google Scholar This interpretation is confirmed by Agesilaos' remark in 379 B.C. quoted by Xenophon, (Hell. v 4.13)Google Scholar that it was more than forty years since he had come of military age and, as other men of similar age (i.e., over sixty) were no longer liable for military service outside the country, he would not undertake the campaign. Plutarch (Ages. 1.2) says that the heirapparent (but apparently not other members of the royal family) was exempted from the ἀγωγή because it was a training in obedience rather than in command. Whatever training the heir or a young king received, there is no reason to think that he was not ready for service at as early an age as other citizens.

⁵ Busolt, G.–Swoboda, H., Griechische Staatskunde ³ (Munich, 1926) 691Google Scholar, n. 3; Cary, M., CR lx (1946) 29.Google Scholar Both point out that the prohibition in Plut. Lyc. 25.1 is against going to the Agora to buy household necessities under the age of thirty, and has nothing to do with attending meetings of the Assembly.

⁶ It will be remembered that Alexander commanded the left wing at the battle of Chaironeia in 338 B.C. at eighteen, and succeeded to the throne of Macedonia in 336 B.C. at twenty. There were two kings in Sparta; the Agiad was the senior royal house and normally held the more important commands as Pausanias did at Plataia in 479, and in the next year's operations with the Hellenic fleet at Cyprus and Byzantium. If the Agiad king or his regent was for any reason considered unsuitable, the other royal house could supply the commander as happened, e.g., in the early years of the Peloponnesian War when Archidamos led the annual invasions of Attica until 428 because Pleistoanax was in exile and his son was still a minor. The prominence of Agis in the later years of the war was similarly due to distrust of Pleistoanax and his own lack of self-confidence (Thuc. v 17).

⁷ See Xen. Mem. i 2.35 for thirty as the age of full responsibility. It may be objected that the real reason that advisers were sent with Pleistoanax in 446 was not his youth but distrust of his heredity or capacity. This may well be true, but the reason given was his youth, and it would not have been accepted as plausible if he had been in fact thirty or over.

⁸ Thuc. ii 21. 1; Plut. Per. 23.1; Gomme, , Commentary ii 74Google Scholar suggests that the effective bribe was not so much the famous ten talents as Pericles' offer to surrender Megara, Troizen, and Achaia which Athens later ceded in the Thirty Years Peace, and that the young king was no match for so wily a bargainer.

⁹ Pleistoanax' exile lasted from late summer 445 to late summer 427; Thuc. v 16.3 and Gomme's commentary ad loc.

¹⁰ The last year of the reign of Archidamos was 427; by 426 his son Agis led the annual invasion so that it would seem that Pleistoanax, although allowed to return from exile in 427, inspired no great confidence. Kleomenes, uncle of the younger Pausanias, commanded for only the one year, 427, because of Archidamos' frailty.

¹¹ Greek men normally married at thirty or shortly thereafter; see, e.g., Hesiod, , Works and Days 695–7Google Scholar; Solon, fr. 27 (Edmonds) 9–10; Plato Rep. 5.460e, Laws 6.785b, 6.772d; Arist. Pol. 1335a. For Sparta there is the specific evidence of Plut. Lyc. 25.1 that men under thirty did not go into the Agora at all but had their household necessities supplied by relatives and lovers this surely implies that a Spartan was unmarried and did not have his own household under thirty, but that after thirty he would usually be married.

¹² This is the date most recently proposed by Forrest, W. G. in CQ x (1960) 237–8Google Scholar (‘473 at the latest’).

¹³ Diod. xi 69 places the murder of Xerxes and the accession of Artaxerxes in 465/4. Evidence from oriental sources enables greater precision: the terminus post quem for Xerxes' death is August 4–8, 465 and it cannot have been much earlier than December 17th, since a Babylonian legal text has been found, dated to the month of Kislimu of the 2 ist year of Xerxes, which in 465 began on Decem ber 17th. News of Artaxerxes' accession had reached Egypt by January 2, 464. See Parker, R. A. and Dubberstein, W. H., Babylonian Chronology ² (Brown University Studies 19, Providence, R. I., 1956) 17Google Scholar and Horn, S. H. and Wood, L. H. in JNES xiii (1954) 9.Google Scholar I am indebted to Professors F. M. Heichelheim and R. F. G. Sweet for calling to my attention the latter reference.

¹⁴ Contra W. G. Forrest who (in the article cited in note 12) believes that there was a long interval between the two, that Pausanias died not later than 473 and that the charges against Themistokles were laid in the winter of 469/8, after Kimon's return from the Eurymedon. He finds support for this in Ephoros, through Diodoros xi 54. Some time, of course, must be allowed for the charges against Themistokles to be elaborated and acted upon, but this is more likely to be months than years.

¹⁵ See Gomme, , Commentary i 437–8Google Scholar for the names of his accusers: Leobotes of the Alkmeonidai; Pronapes, perhaps the hipparch of IG i² 400 who was a colleague of Kimon's son Lakedaimonios, and a Lysandros. Plut. Them. 24.4 says that Kimon himself prosecuted the friend of Themistokles who smuggled his wife and children out of Athens. This evidence suggests that Kimon and those who shared his views wished to preserve good relations with Sparta, and having no love for Themistokles willingly sacrificed him to that end. The charge of Medism was, one suspects, not the real ground of complaint; indeed, the evidence to support it does not appear in any of the accounts and may not have existed; cf. Forrest's discussion (see n. 12) pp. 232–41.

¹⁶ Gomme, , Commentary i 397–9Google Scholar believes that 467 or 466 are the latest possible dates, but doubts that Pausanias' trial could have been postponed so long; see below pp. 144 for discussion of this point.

¹⁷ Gomme, A. W., The Population of Athens in the Fifth and Fourth Centuries B.C. (Oxford: Blackwell, 1933) 75–82Google Scholar, Note C, points out that statistics compiled from, e.g., the Prosopographia Attica, showing a great preponderance of males, are misleading because the majority of known names are those of people in public life, ‘a field in which women did not compete’. Census figures from modern Greece show the sexes as approximately equal, with males predominating slightly in the age groups below thirty, and females predominating thereafter; male death rates are higher than female up to the age of 35. See Demographic Yearbook 1959 issued by the Statistical Office of the United Nations (New York, 1959) Table 5, p. 161, and Table 26, pp. 572–3.

¹⁸ Gomme, , Population of Athens 79Google Scholar: the average infant death rate (under one year) in Mediterranean countries in the 1920's was 120 per 1,000 births, and it was presumably higher, or at least as high, in antiquity. Greek girls married young, about fifteen (Xen. Oec. 7.5; cf. Arist. Ath. Pol. 56.7), or at Sparta perhaps a little older (Plut. Lyc. 15.3 advises that they be ἀκμαζούσας καὶ πεπείρους). Demographic studies show that the fertility rate of girls under fifteen is very low, that it rises slowly between fifteen and nineteen, and then rapidly to a peak between twenty and twenty-nine; after thirty it declines again. In Greece in 1956 the number of live births to mothers under fifteen was 20, to mothers fifteen to nineteen, 5,160, to mothers twenty to twenty-four, 38,792, and to mothers twenty-five to twenty-nine, 57, 618; Demographic Yearbook 1959, p. 11, and Table 11, p. 261. I am grateful to three colleagues in the Department of Political Economy in the University of Toronto for information about and discussion of these problems: K. F. M. Helleiner, N. Keyfitz, and Andrew Watson. They warned that a ten-year period between marriage and the conception of the third son who reached maturity is on any calculation of probabilities a minimum estimate; see below p. 150 f. for the suggestion that Pausanias married c. 475.

¹⁹ Jacoby, F., Die Fragmente der Griechischen Historiker (FGrH) 70Google Scholar Ephorus F 191.37–46; cf. Gomme, , Commentary i 399–400Google Scholar and Meritt, , Wade-Gery, , and McGregor, , ATL iii 158–9.Google Scholar Gomme and the editors of ATL give good reasons for rejecting a seven-year rule in Byzantium based on the emended text of Justin's epitome of Trogus 9.1.3 and accepted by, e.g., Beloch, K. J., Griechische Geschichte ii. 2 (Strassburg, 1916) 185–8Google Scholar, Lehmann-Haupt, C. F., Klio xvii (1921) 59–73Google Scholar, and Heichelheim, F. M., Zeitsch. f. Num. xl (1930) 22–4Google Scholar, who uses as a supporting argument the existence of iron money in Byzantium. Diodoros (xi 44–6) puts his whole account of Pausanias' later career under this year, but which episode he intended to assign to it is anyone's guess.

²⁰ The action of the ephors was apparently resented by a strong faction in Sparta; indeed there may have been doubts about the evidence alleged against him since there was no trial and he had no opportunity to defend himself. At any rate the ephors were forced to give him proper burial and to allow him honours after his death: Delphi later ordered his tomb to be removed to the spot where he died and two bronze statues of him to be dedicated to Athena Chalkioikos; Thuc. i. 134.4 with Gomme's commentary ad loc. and Paus. iii. 17.7. Herodotos also reflects a tradition favourable to Pausanias; see How, W. W. and Wells, J., A Commentary on Herodotus (Oxford, 1912) ii 12Google Scholar for references.

²¹ Her. vi 71–2; cf. Plut. Them. 20.1, Arist. 22.2, Mor. 859 D; Paus, iii 7.9. The Thessalian expedition is not dated by Herodotos but, if it be granted that the setting of Plutarch's tale about Themistokles' plot to burn the Greek fleet at Pagasai is historical (the plot itself may be rejected as a fabrication of the later anti-Themistoklean propaganda), E. M. Walker's date of 479 (CAH v 466) seems to me the most probable. To Walker's arguments for 479 might be added: (1) the phrase ἀπηλλαγμένου Ξέρξον of Them. 20.1 is more appropriate to 479 than later; (2) a naval expedition from Sparta to Thessaly seems unlikely, but there is less improbability in the Hellenic fleet (minus the Athenian contingent) stopping in Thessaly en route from the Hellespont (Her. ix 114 says that they departed ἐς τὴν Ἑλλάδα and Thuc. i 89.2 ἐπ' οἴκου but neither wording precludes their stopping in Thessaly on the way); (3) when Pausanias was recalled from Byzantium in 478, why was Dorkis (Thuc. i 95.6) sent to succeed him unless Leotychidas who had commanded in 479 was already in disgrace?

For the dates of the reigns of Leotychidas and Archidamos see Gomme, , Commentary i 405–7.Google Scholar Professor Andrewes suggests to me that Diod. xi 48.2 where Leotychidas' death is placed in 476/5, archon Phaidon, is simply a mistake: that Diod. confused Phaion, whom he gives as archon for 469/8—the correct year of Leotychidas' death—with the earlier Phaidon, and that there is no question of confusion between exile and death.

²² Recent discussion of these Peloponnesian events may be found in Gomme, 's Commentary i 401–8Google Scholar; Meriti, , Wade-Gery, , and McGregor, , ATL iii 158–80Google Scholar; Andrewes, A., ‘Sparta and Arcadia in the Early Fifth Century’, Phoenix vi (1952) 1–5CrossRefGoogle Scholar; Forrest, W. G., CQ n.s. x (1960) 221–41.CrossRefGoogle Scholar Argos, always hostile to Sparta, was for all or part of this period under the regime which had seized control after Sepeia in 494 B.C.; there had been synoikisms at Elis (471/0) and Mantinea, accompanied probably by the establishment of democracies in both places; Sparta had to fight two battles against the Arcadians at Tegea and Dipaia about this time; and the Helots revolted at the time of the great earthquake c. 465.

²³ As is maintained by, e.g., Lenardon, R. J., Historia viii (1959) 26 ff.Google Scholar, with n. 15 for references to earlier advocates of this date, Highby, Busolt, and Jacoby. It should be added that Cicero (Lael. 12.42; Brut. 10.41) does support 471 or 470 as a year of crisis in Themistokles' life; he says expulsus esset which may mean either the ostracism or later condemnation.

²⁴ So Forrest, W. G., CQ x (1960) 241Google Scholar; cf. Gomme, , Commentary i 401Google Scholar for similar scepticism about the value of Diodoros' date. It is accepted as the date of the ostracism by Flacelière (see n. 2), and the list of earlier writers cited by Lenardon (see n. 24) 26, n. 14. Lenardon dates the ostracism even earlier, c. 474/3.

²⁵ As in Pausanias' case, one suspects that the real ground for Spartan apprehensions was the trouble with the Helots; they feared that Themistokles was encouraging Pausanias' intrigues with them rather than his Medism; cf. Thuc. i 102.3 and Kimon's dismissal from Ithome.

²⁶ ATL iii 112, n. 7, where the delay is accepted and corroboration found in Number 20 of the Epistles ascribed to Themistokles; also Lenardon (see n. 24), esp. pp. 37–9.

²⁷ P. 241 of article cited in n. 24.

²⁸ Commentary i 397–8. Gomme himself tentatively suggests that Themistokles may have written to Artaxerxes before the death of Xerxes, but is not happy about the suggestion because it convicts Thucydides of error. He rightly dismisses the testimony of the writers quoted by Plutarch (Them. 27.1–2) who had Themistokles meet Xerxes, rather than Artaxerxes. Their aim was dramatic effect rather than the solution of a chronological difficulty.

²⁹ Plut. Them. 24.4 quotes Stesimbrotos as saying that Themistokles fled to Hiero of Syracuse, asking for his daughter's hand in marriage and offering to make him ruler of Greece. Plutarch rejects a journey to Sicily as inconsistent both with the rest of Stesimbrotos' account and with all the other accounts. Hiero died in 467/6 (Diod. xi 66.4) and the news might have reached Themistokles on his arrival in Kerkyra where he had gone with the intention perhaps of taking refuge in Sicily. If Themistokles heard in Kerkyra of Hiero's death, then 467/6 is of course the earliest possible date for his arrival in Kerkyra; cf. Gomme, , Commentary i 400, n. 1Google Scholar, who mentions the passage as ‘a possible indication of date’. This has also been suggested as an explanation of the route—Kerkyra seems an extraordinary first stop on the way to Susa. But the route up the west coast was no doubt the safest for Themistokles, who was fleeing from the Athenians whose navies controlled the Aegean. Demaratos, the ex-king of Sparta, had taken a similar route some years before, Her. vi 70.

³⁰ ATL iii 160; cf. Gomme, , Commentary i 397–9.Google Scholar

³¹ Gomme, , Commentary i 390–1Google Scholar; ATL iii 175–6, esp. n. 57.

³² Flacelière (n. a), 5–14; cf. Frost, Frank J., ‘Thucydides i 137.2’ in CR n.s. xii (1956) 15–16Google Scholar points out that the prevailing summer winds in the Aegean are ‘northerlies’, the etesian winds, and that a ship sailing from Pydna caught in one of them would run for shelter between Naxos and Paros. Only a southerly gale of some duration would carry a ship off course as far north as Thasos, and even then a mariner would avoid the ‘iron’ coast of Thrace opposite Thasos. His objections to Thasos are valid if a straight voyage from Pydna to Ephesos or Kyme is in question, but Themistokles' captain may have been making for some northern port of call on the way, and found Thasos his easiest place of refuge.

³³ Gomme, , Commentary i 398Google Scholar; for the Epistles and their sources see Lenardon, R. J., ‘Charon, Thucydides, and “Themistokles”’, Phoenix xv (1961) 28–40CrossRefGoogle Scholar, with the bibliography there cited; for Charon as Thucydides' contemporary see Jacoby, F., Abhandlungen (Leiden, 1956) 178–206.Google Scholar

³⁴ E.g., Beloch, K. J., Griechische Geschichte i. 2 (Strassburg, 1913) 171–8, esp. 172Google Scholar, in his full and valuable discussion of the Agiad king-list puts him at not less than thirty; Schaefer, Hans, RE xviii. 2 (1949) 2563–79Google Scholar, s.v. ‘Pausanias’ 25, says ‘not older than 30–35’; Grant, J. R., ‘Leonidas' Last Stand’, Phoenix xv (1961) 21Google Scholar, ‘he may have been around thirty in 480’.