¹ Laws iii 692d, and, especially, 698e: Does the vague reference to other Spartan difficulties perhaps refer to the formation of the Arkadian League (see below), or, more generally, to their difficulties with Kleomenes?

² Pausanias iv 23: After the capture of Ira in their second war with Sparta (which Pausanias dates to the 28th Olympiad—668 B.C.) the Messenians who had not been captured by the Spartans decide to found a colony, and accept the invitation of Anaxilas of Rhegion (who was tyrant there from 494 to 476) to help him conquer Zankle; this they do (‘in the 29th Olympiad’) and change the name of Zankle to Messene. The chronological confusion was pointed out by Bentley in the Dissertation on the Epistles of Phalaris, etc. (1697, enlarged edition 1699). See also Diodorus xv, 66:

³ Strabo viii 4, 10:

⁴ Thuc. vi 4, 6: The reason is different, but not incompatible.

⁵ Frag. Gr. Hist. IIIa (1943), pp. 109–81. With some diffidence I make the following general remarks about Jacoby's discussion. There seems to be no guarantee that ‘A’ is the only intermediate source used by Pausanias, or that Rhianos and ‘the vulgate’ plus some Myron were ‘A's’ only sources. Moreover, if Pausanias' sources are as thoroughly mixed in his narrative as Jacoby holds, it is obviously difficult to feel certain about their identification and to know exactly what comes from which. Jacoby's argument seems to depend on the assumption that our information about the sources available to Pausanias, and used by him, is approximately complete. The attempt at (Quellenforschung must, of course, be made, and Jacoby's analysis is very attractive; it is still a question how much real confidence one can have in the results in detail. But when every allowance is made for the uncertainties, Jacoby's conclusion that Rhianos dated the revolt of the Messenians led by Aristomenes to the early fifth century seems highly probable. It seems to me much less clear that this early fifth-century revolt must be pronounced a fiction in toto. Whether the Messenian revolts were two or three in number is surely a literary rather than an historical question—there must, in three centuries of oppression, have been more revolts than that, although perhaps only two or three of them could properly be called ‘wars’. Both Plato and Rhianos put a Messenian revolt early in the fifth century, and it is surely more likely that they had some tradition of an insurrection at that time to go on than that they had none. This a priori consideration is supported by various definite indications, especially now by the date of the change of name at Zankle, and by the probable date of the Olympia dedication: see notes 6 and 7 below.

⁶ ‘Rhegion, Zankle-Messana and the Samians’, by Robinson, E. S. G., JHS LXVI (1946), pp. 13–21.CrossRefGoogle Scholar Robinson shows good reason to believe that the Samians who seized the town in 494/3 struck only five numbered and probably annual issues of coin before being ousted by Anaxilas, an event which will thus have occurred in 489, or possibly in 488.

⁷ ‘Comments on some archaic Greek inscriptions’, by Jeffery, L. H., JHS LXIX (1949), pp. 25–38CrossRefGoogle Scholar, see pp. 26–30. Miss Jeffery shows that the letter forms of the inscription can hardly be dated as late as 465; thus the dedication should not be related to the revolt which followed the earthquake. The letter forms certainly seem too early for a date near the middle of the century, but unfortunately there is little comparative material available.

⁸ ‘The Growth of Spartan Policy’, by Dickins, Guy, JHS XXXII (1912), pp. 1–42CrossRefGoogle Scholar, see pp. 31–2.

⁹ Her. vi 74.

¹⁰ Beloch, (Gr. Gesch. II. 1, 2nd ed., p. 36)Google Scholar says: ‘Wahrschein lich haben ihn die Ephoren aus dem Wege geräumt, im Einver ständnis mit seinem Stiefbrudern, Leonidas und Kleombrotos.’ Similarly, Mitchell and Caspari (Grote, George, A History of Greece, ed. M, . & C., , London n.d.-1907, p. 176Google Scholar) say ‘it may be suspected that Herodotus' account of Kleomenes' death covers a piece of foul play on the part of the ephors’, and Monro, expresses the same suspicion in CAH IV, pp. 261–2.Google Scholar

¹¹ ‘Arkadische Münzen’ in ZfN IX (1882), pp. 18–41, and ‘Nachmals das altarkadische Gemeinwesen’ in ZfN XXIX (1912), pp. 139–46.

¹² ZfN IX, p. 20: ‘Da ein politisches Centrum in Arkadien vor der Erbauung von Megalopolis nicht existiert hat, muss der Prägort der arkadischen Landesmünzen während der älteren Zeit gesucht werden bei einem der gemeinsamen Stammesheiligthümer des Landes.’ That coins were struck for, and somehow put into circulation at, religious festivals is a frequent assumption of numismatists; money, however, is surely more often brought to fairs and festivals than carried away from them—it is the visitors who do the purchasing.

¹³ ZfN VXXIX, pp. 144–5.

¹⁴ Babelon, E., Traité des monnaies grecques… II. 1 (1907), p. 860.Google Scholar

¹⁵ Gardner, Percy, A History of Ancient Coinage (Oxford, 1918), p. 381.Google Scholar

¹⁶ Seltman, Charles, Greek Coins (London, 1933), p. 97.Google Scholar

¹⁷ Head, B. V., Historia Numorum ² (Oxford, 1911), pp. 444 and 338.Google Scholar

¹⁸ Historians have regularly treated Kleomenes' activities in Arkadia as abortive: How, and Wells, , A Commentary on Herodotus, vol. 2 (Oxford, 1912), p. 93Google Scholarad vi 75, 1; Beloch, K. J., Gr. Gesch. II i (2nd ed.Strassburg, 1914), p. 36Google Scholar; Monro, J. A. R. in CAH IV (Cambridge, 1930), p. 261, etc.Google Scholar Grote merely says that Kleomenes ‘employed the powerful influence of his regal character and heroic lineage to arm the Arcadian people against his country’, making no reference to a league. Grote's editors, Mitchell and Caspari (l.c. note 1) say ‘a nucleus for an Arcadian League existed in the common religious cult of Zeus Lykaeus in connexion with which a federal coinage was issued at this period’, but they seem not quite to believe in the league, and do not refer to it again; Caspari includes a brief notice of the Arkadian coins in his ‘Survey of Greek Federal Issues’ in JHS XXXVII (1917), pp. 168–83, but he appears to accept Weil's early date for them, and he does not mention Kleomenes.

¹⁹ Tod, M. N., Greek Historical Inscriptions, I² (Oxford, 1946), no. 5.Google Scholar

²⁰ The Arkadian games for which they were supposedly issued, though of great antiquity (Aristotle ap. schol. Aristides, p. 105, ed. W. Frommel, 1826, says that they were earlier than the Olympic), were not otherwise of great importance—the scanty testimonia (chiefly two or three references in Pindar, and the scholia ad loc.) are collected in Immerwahr, W., Die Kulte und Mythen Arkadiens I (Leipzig, 1891), p. 5.Google Scholar

²¹ Her. viii. 71.

²² That Tegea, at least, had been hostile to Sparta for some years is clear from the fact that the Elean seer Hegesistratos took refuge there from the Spartans some time before Plataia (Her. ix 38), and that the exiled Spartan king Leotychidas spent perhaps ten years there, from 479/8 or 478/7 until his death in 469 (?)—Her. vi 72. The best discussion of Arkadian history at this period is by Andrewes, A. in ‘Sparta and Arkadia in the Early Fifth Century’, The Phoenix VI (1952), pp. 1–5.CrossRefGoogle Scholar Andrewes distinguishes three stages in Tegeo-Spartan relations—the 480's when Tegea was hostile to Sparta, the 470's when Tegea was friendly but Mantineia was ‘a source of trouble’ (Sparta doubtless disapproved of her new, probably democratic, synoikismos), and the middle of the 460's when Tegea was again hostile but Mantineia was friendly. We may perhaps equate Tegea with the Arkadian League, of which it was the strongest member. Then the ‘friendliness’ of the 470's (or is friendliness too strong a word?) was due, at the time of Plataea to Spartan respect for the strength of the new league, and at the end of the decade to the League's fear of the new Mantineia—Sparta found it necessary to treat the League with respect at the time of Plataea, while the League needed Spartan neutrality (and Sparta the League's) as Mantineia and Elis grew stronger through synoicism at the end of the decade. Perhaps the 470's should be regarded rather as a period of wary co-operation; in the 460's Mantineia exploited the essential hostility between the League and Sparta.

²³ Similarly, when the Euboian League was founded and the Euboian federal coinage began—in 411/10—the independent cities which composed the League struck no coins for a period of some forty years or more; when they did begin to strike in their own names again, about 369 (?), Eretria did not do so—perhaps an indication that the league no longer had any real existence but was being maintained as a fiction by the city which had been its capital. I have discussed these matters in some detail in a forthcoming study of ‘The Euboian League and its Coinage’. The exact date when the Arkadian League issues came to an end, and the exact dates at which the individual Arkadian towns begin to strike again, are not, of course, known. Head in Hist. Num. ² assumed that there was an overlap; Weil considered that there was none.

²⁴ It is unfortunately impossible to tell exactly how long an interval separates the Spartan refusal to march at once (as delivered in Sparta to Pheidippides) and the actual setting out of their 2000 hoplites. One difficulty is that there is no reason to suppose that any state's calendar, in the early fifth century, was in step with the moon—see on this question Pritchett, W. K. sensible remarks in ‘Julian Dates and Greek Calendars’ in CP XLII (1947), pp. 235–43Google Scholar, especially p. 238. Indeed, the probability is strong that all calendars were wrong to some extent, and some very wrong indeed. Thus Herodotos' one actual date—Pheidippides’ arrival at Sparta on the ninth of the Spartan month—does not help. Plutarch's thrice-recorded date for Marathon (and for the celebration of the victory), Boedromion 6, is often doubted (see Jacoby, note 121 in JHS LXIV (1944), p. 62Google Scholar) but may well be right; if so, we have evidence that the Athenian calendar was out of step with the moon in 490/89, for Herodotos certainly implies that the battle was fought about the time of the full moon—the Spartans arrived at Athens three days after the full moon, marched to Marathon, and found the dead still unburied. As Herodotos also suggests that Pheidippides left Athens after the Persian landing at Marathon, that the battle occurred about a week after the landing, and that the Spartans arrived two or three days later, he appears to imply that the Spartans inarched out about a week or at most ten days after Pheidippides' arrival. His dates are not, however, explicit, and his relative chronology contains inherent difficulties (e.g. did Pheidippides really not leave for Sparta until Athens had received news of the Persian landing in Attica?—this difficulty is emphasised by Sotiriades in Πρακτ. Ἀκ. Ἀθ. VIII, 1933), so that no firm conclusions are possible.

²⁵ There is evidence of unrest among the helots on some occasion prior to 465 in Thuc. I 128, I, where the Athenians tell the Spartans to ‘drive out the curse of Tainaron’ It is possible that this incident was connected with the revolt of 490.

²⁶ It is true that most historians are unwilling to accept Herodotos' definite implication that Kleomenes' death ante dated Marathon (see Beloch, , Gr. Gesch. II. 1², p. 36Google Scholar; Monro, in CAH IV, pp. 261–2Google Scholar; Andrewes, A., ‘Athens and Aegina, 510–480 B.C.’, BSA XXXVII, 1936–1937, p. 4Google Scholar, etc.); for they feel that there is too little time between his coercion of Aigina (spring-summer 491) and the battle of Marathon, a period of rather more than a year, for his recorded movements and for a war between Athens and Aigina (Her. vi 87–93). Andrewes may well be right about the Aiginetan wars; but as far as Kleomenes himself is concerned, if his bribery of the Pythia was discovered in the summer of 491 and his flight from Sparta followed almost at once, his visit to Thessaly, his Arkadian activities, and his recall may surely belong to the autumn of 491 and the winter, spring, and summer of 490 without undue compression. Indeed, this rather close time table shows that the helot revolt must have been suppressed very quickly, and helps to explain why it did not leave clearer traces in the tradition.