¹ Dinsmoor, W. B.The Archons of Athens in the Hellenistic Age (1931)Google Scholar, The Athenian Archon-list in the Light of Recennt Discoveries (1939); Pritchett, W. K. and Meritt, B. D.Chronology of Hellenistic Athens (1941)Google Scholar.

² GG I² ii (1913) 155–70. Other lists: von Schöffer, RE II (1895) 585 ff.Google Scholar; Kirchner, Prosopographia Attica (1903) II 631–52Google Scholar; F. Hiller von Gärtringen IG I² (1924) 267–301.

³ V 20. 2.

⁴ Aristotle, for example, assumes that his readers can refer to an archon-list (Ἀθ. πολ. 17. 2). Cf. T. 39.

⁵ 566/5: Hippokleides (T 4), unless the notice about the Panathenaia is a later parenthesis. 510/9 (?): Skamandrios (T 7). See below, pp. 104, 113.

⁶ See below, pp. 109–12.

⁷ e.g. Wilamowitz, Aristot. u. Athen. I 8 f.Google Scholar; Beloch GG I² ii 157.

⁸ See below, p. 80.

⁹ On the ἀναρχίαι of Ἀθ. πολ 13. 1, see below, pp. Other examples are given in Jacoby, Apollodors Chronik 172Google Scholar. The year,of Pythodoros (404/3), nominee of the Thirty, was afterwards reckoned an ἀναρχίαι: cf. Xen., Hell. II 3. 1Google Scholar (), Diod. XI V 3. 1 (), [Plut.] , pp. 835 f. (), ‘Suidas’ s.v. . Dion Chrysostomos 71 (XXI, περί κάλλους) refers to ἀναρχίαι as frequent at Athens (;), and there is an example in an inscribed archon-list of the first century A.D. (IG III 1014). No doubt some examples have been lost.

When an illegal archon was replaced before the end of the year by a legally appointed one, it was the latter's name which was entered on the list (cf. Ἀθ. πολ. 33. 1, quoted below, n. 244).

¹⁰ V 20. 2.

¹¹ See below under Kreon, etc., pp. 88 f.; Aristaichmos, p. 92; Dropides and Phormion, p. 99; Solon and Komeas (text of Ἀθ. πολ. 14. 1), pp. 95–9; Aristeides, p. 117; Nikodemos, p. 118. But most of these and other discrepancies can be removed without drastic steps.

¹² See below, pp. 116–9.

¹³ See below, p. 83.

¹⁴ See below, pp. 109–12.

¹⁵ See below, pp. 93–5.

¹⁶ They cover the period 892–667; and events were dated by the names of the individual limmu. See S. Smith, in CAH III 3Google Scholar.

¹⁷ See e.g. Tod, M. N. in JHS LXII (1942) 56 f.Google Scholar, on recently discovered Attic inscriptions. For writing on materials other than stone in archaic Greece, cf. Kenyon, F. G.Books and Readers in Ancient Greece and Rome 42–4Google Scholar.

¹⁸ Hdt. VIII 41. 1, 51. 2, 53. 2, IX 12. 2; Thuc. I 89. 3.

¹⁹ Cf. Plut. Solon 25. 1, and Holland, L. B.AJA XLV (1941) 346–62CrossRefGoogle Scholar.

²⁰ See above, p. 78.

²¹ See below, pp. 104, 106, 109.

²² IG I². 16 f., 4. 26 f.; see below, p. 118; for the Salamis decree, p. 119.

²³ Ἀθ. πολ. 14. 1 (see below, p. 104).

²⁴ T 8 (see below, p. 113).

²⁵ See below, pp. 113 f., 115ⁿ.

²⁶ I 1. I, II 3. I, 77. 1, IV 46. I, etc.

²⁷ VI 53. 3, 54. 1, 55. 1, 60. 1.

²⁸ VIII 51. 1.

²⁹ See above, p. 80.

³⁰ Thuc. I 97. 2; for annalistic treatment in latter part of Hellanikos cf. FGrHist 4 F 171 (407/6), Pearson, L.The Local Historians of Attica 14 f.Google Scholar and Gomme, A. W.Commentary on Thucydides I 6. n. 3.Google Scholar

³¹ See above, pp. 77 f.

³² II 2. 1; cf. III 68. 5, IV 102, and see below, p. 117ⁿ.

³³ See above, p. 80.

³⁴ F 46 (FHG I 376): death of Kleon in 422/1 (Alkaios). VOL. LXVIII.

Cf. L. Pearson op. cit. 85.

³⁵ F 97 (FHG I 400): (Θεοδώρου is a pretty certain correction for Πυθοδώρου). Cf. L. Pearson op. cit. 121 f.

³⁶ TT 31, 32. See below, p. 119 (statue of Hermes ἀγοραῑος erected under Kebris; big fall of snow under Lakrateides).

³⁷ See above, p. 80.

³⁸ See Jacoby, F.Apollodors Chronik 39–51Google Scholar.

³⁹ FGrHist 228 FF 1 (T 12), 2; cf. FF 3, 10, and 43 f. (TT 71 f.). See below, p. 102.

⁴⁰ See below, p. 85.

⁴¹ FCrHist 6.

⁴² The adoption of the theory of Lenschau, T. (Philologus XCI = n. F. XLV (1936/1937) 396–411)Google Scholar that Olympiads 1–49 formed an annual series running from 632/1 to 584/3, after which the festival became quadrennial, would not oblige us to lower accordingly the dates of all events dated by Qlympiads 1–49: since all Greeks, at least from the time of Hippias of Elis, believed the festival was quadrennial from the start.

⁴³ Polyb. XII 11. 1; cf. FHG I 232Google Scholar, Christ-Schmidt, Griech. Litteraturgesch. II 220Google Scholar, Laqueur, RE VI A 1199 f.Google Scholar

⁴⁴ FGrHist 241.

⁴⁵ Cf. Jacoby, F.Apollodors Chronik (Philologische Unter-suchungen XVI (1902)Google Scholar) and FGrHist 244.

⁴⁶ FGrHist 245.

⁴⁷ The list he used was complete from Kreon (T 39).

⁴⁸ FGrHist 251.

⁴⁹ e.g. those giving the durations of the several tyrannies and exiles of Peisistratos (14 f.).

⁵⁰ See below, pp. 98 f., 104, 107, 115 f., 118. Objections to any emendations based on the assumption that cyphers became corrupted were raised by Pomtow (in Rh. Mus. LI (1896) 569)Google Scholar; cf. Hammond, N. G. L. (in JHS LX (1940) 73 f.)Google Scholar. Pomtow said that of 244 numbers in Chapters 1–41, all except five were written out in full; and that ordinal numbers were all written out in full. Hammond (if, as I suspect, we ought to read ‘ordinal’ for ‘cardinal’ on the top line of p. 74 of his article, as it is ordinal numbers that he is discussing) also makes this latter statement. But cyphers do sometimes stand for ordinal numbers in our text, as A. W. Gomme had already pointed out (JHS XLVI (1926) 176¹²Google Scholar: for ἐνάτης (47. 4) and for πέμπτη (54. 7); cf. for τετ ράδι (32. 1). And in any case the fact that most numbers in our present text are written out in full does not mean that they were not cyphers in earlier copies.

⁵¹ Cf. e.g. 22. 1–6.

⁵² See Kirchner, Rh. Mus. LIII (1898) 383¹Google Scholar for examples of inclusive reckoning with cardinals.

⁵³ Cf. Böckh, CIG II 305 f.Google Scholar, Wilamowitz, Philologische Unter-uchungen IV (1881) 251 f.Google Scholar; Dopp, E.Quaesliones de Marmor Pario (1883) 49, 60Google Scholar; Kirchner loc. cit. 382 f., Hermes XXXVII (1902) 441Google Scholar; Jacoby, Rh. Mus. LIX (1904) 80–4Google Scholar, Das Marmor Parium (1904) passim, FGrHist 239, especially comm. in II D 671; Munro, J. A. R.CR XIX (1905) 269Google Scholar; Laqueur, R.RE XIV (1930) 1886–8Google Scholar; Dinsmoor, W. B.The Archons of Athens, etc. (1931) 46 f.Google Scholar, 50, 54, 85–8, Athenian Archon-list, etc. (1939) 21, 37Google Scholar; Pritchett and Meritt op. cit. xx.

⁵⁴ Laqueur loc. cit. alone doubts the certainty of this: see below, p. 84.

⁵⁵ Cf. Dinsmoor and Pritchett-Meritt locc. cit., especially Dinsmoor, Athenian Archon-list, etc., 37.Google Scholar

⁵⁶ So Dinsmoor, Archons of Athens, etc., 46Google Scholar.

⁵⁷ See the table on p. 86. The conclusions about epochs A 32–50 will be explained as they arise in the paper. Those about A 51 ff. can be checked from Jacoby's editions.

⁵⁸ It is very improbable that any of the figures from the two lacunae were other than ‘orthodox,’ considering the uniformity presented by A 67–79 and B 1–26. Böckh, Dopp and Wilamowitz were unable to assess correctly the proportion of unorthodox to orthodox figures, as B had not been discovered when they wrote.

⁵⁹ The errors within the intrusion (in 58, 62, 63 and the apparently ‘orthodox’ 55) may (as already suggested) have been due to the mason. The accuracy of A 67–B 26 is not a serious argument against this.

⁶⁰ See below, p. 106.

⁶¹ Cf. e.g. I 74.

⁶² See below, pp. 88–90, 100; cf. 109, 117.

⁶³ Dionysios in Books V–X (Ol. 68–82, cf. TT 51–5, 59 f.); Diodoros in Books XI–XX (Ol. 75–119).

⁶⁴ i.e. in Books I–IV; cf. TT 44–50.

⁶⁵ II 24. 7 (T 86), IV 5. 10 (T 88), 13. 7 (T 89), 15. 1 (T 90), 23. 4 (T 91), 27. 9, VI 5. 3, 9. 5 (T 93), VII 25. 4, VIII 27. 8, 39. 3 (T 94), 45. 4, X 2. 3, 3. 1, 5. 13 (T 95), 7. 4 (T 96).

⁶⁶ Cf. Apollodors Chronik FF 13–15, 17, 36, 4 6 (ter), 47, 51 f., 52, 54, 56 (sexies), 69, 76 (bis), 95.

⁶⁷ See below, under Therikles, p. 109.

⁶⁸ VI 3. 8 (battle of Plataia in Ol. 75: actually in Ol. 75. 2 = 479/8). Of the other four events in Pausanias dated by a simple Olympiad number, two fell certainly (cf. IV 24. 5, VII 16. 10) and one probably (cf. IV 23. 10 = T 92; see below, p. 90) in the first years of their Olympiads; the date of one cannot be checked (cf. VII 16. 6).

⁶⁹ Cf. Apollodors Chronik FF 64 (Ol. 116 for 116. 2 or 3), 72 A (114 for 114. 3), 74 (127 for 127. 2) and 79 (130 for 130. 3). For the three instances where the number to be supplied is 1, cf. FF 52 (Ol. 88), 53 A (108), 73 (123). There are besides twenty-one events dated by simple Olympiad numbers, to which exact years cannot be assigned.

⁷⁰ This may easily be seen in the Bodleian manuscript (collotype reproduction, Oxford, 1905) or in the version by R. Helm.

⁷¹ e.g. Kalliades' archonship assigned to Ol. 74. 3 or 4 instead of Ol. 75. 1.

⁷² Ol. 1. 1 = Abr. 1240 (Arm.); = Abr. 1241 (Hieron.).

⁷³ Cf. Jacoby, F.Die attische Königslisten in Klio II (1902) 406–39Google Scholar.

⁷⁴ For discussion see F. Jacoby op. cit. 409–15, Das Marmor Parium 162 f.; Beloch GG I² ii 156 f.

⁷⁵ loc. cit.

⁷⁶ Pausanias IV 17. 2, 10.

⁷⁷ id. 23. 4.

⁷⁸ id. 20. 1. That these figures should be reckoned inclusively is required by general Greek usage, which Pausanias elsewhere followed (cf. IV 5. 10, 13. 7 and X 2. 3, 3. 1; the exclusive reckoning of IV 13. 7, 15. 1 is probably an aberration).

⁷⁹ Cf. Jacoby locc. cit. He claims that a single error of one Olympiad runs through the whole of Pausanias' early chronology, accounting not only for this discrepancy, but also for his mistake about the victories of Chionis (see below, p. 90) and the numbering of the Pythiads (p. 100). But the error here can be made one of four (not five) years only by assuming that, of the two figures in Pausanias IV 17. 2, 20. 1, one was reckoned inclusively and one exclusively—which is improbable. Nor is there any connexion between this error and the error about Chionis, other than that both point to carelessness: for the former arises from a failure to relate the archon-list correctly to the victor-list, the latter from a miscopying of the victor-list. And whereas it is true that Pausanias starts the series of numbered Pythiads in Ol. 48. 3 instead of Ol. 49. 3, he knows that the first ἀγὼν στεφανίτης of the new series fell in Ol. 49. 3, and errs only in calling it Pythiad 2. His ἀγὼν χρηματίτης, on the other hand, which he calls Pythiad I, he dates five years too late. It would be wrong, therefore, to assume that we can correct Pausanias by reducing by four years the dates of all his archons down to Erxikleides (see below, p. 109).

⁸⁰ Böckh and Le Paulmier read Τλησια on the Marble (T 16), but Munro's careful examination of the stone established the reading Λυσια… (CR XV (1901) 356)Google Scholar. Similarly, the reading Λυσιας restored by Schubart and Walz to the text of Pausanias ‘e vestigiis Vindobonensis a’ (T 90) cannot sway us very strongly, as even if the editors in question did not succumb to the temptation to read more into the ‘vestigia’ of one MS. than could reasonably be inferred from them, we should still have to explain how the name became corrupted to the less common Τλησιας in the other MSS. Jacoby suggests that the two names may be variants of one another, which seems a little far-fetched; and he himself prefers to solve the problem by reckoning the Marble's epoch A 32 inclusively and 33 exclusively, so that Lysiades could follow immediately after Kreon (see Klio II (1902) 411¹Google Scholar; Das Marmor Parium 163; FCrHist 239 comm. ad loc.).

^80a So Wilamowitz, Aristot. u. Athen II 72Google Scholar; on Pausanias' victor-list, see below, n. 82.

⁸¹ See above, p. 89, n. 79.

⁸² In other respects the same version of the victor-list appears to have been used by Dionysios, Pausanias and Eusebios. Dionysios agrees with Eusebios as to the victors of Olympiads 7, 27, 35, 41, 50, 61, 68 and then continuously: Pausanias agrees with Eusebios as to the victors of Olympiads 1, 4, 6, 9, 14, 15, 18, 23, 25, 33, 37, 38, 41, 48, 58, 65, 72, etc. The discrepancy between Pausanias and the orthodox list is confined to this instance, and shortly before and after the era of Chionis' victories there is complete and continuous agreement. We cannot believe that a complete list circulated authoritatively till the time of Eusebios showing only three victories of Chionis in the στἀδιον, while an inscription at Sparta recorded four. That the compiler of the list (Eratosthenes) distinguished between Charmis and Chionis is indicated by the little observations attached to them (T 103): it is therefore more likely that Pausanias' error was due to someone's misreading Charmis as Chionis. If his list had Chionis down four times, and this had impressed itself on his mind, it would be easy for him to read the wrong way round an inscription which told of three victories in the στἀδιον and four in the δίαυλον. See further Jacoby, in Klio II (1902) 410²Google Scholar.

⁸³ That by Ol. 29 Pausanias means Ol. 29. 1 is not certain, but suggested by the wording.

⁸⁴ See below, p. 115ⁿ, on re-election to the archonship.

⁸⁵ Töpffer, Attische Genealogie p. 280Google Scholar assumes a different family; Berve, in Hermes: Einzelschriften II (1937) 22²Google Scholar declines conjecture. See below under Hippokleides p. 104 and Miltiades, archon 524/3, p. 110.

⁸⁶ Beloch GG I² ii 158 f. assumes that Miltiades was archon only once, and arguing from the confusion in Pausanias about Chionis, he adopts a date of his own, Ol. 28. 2 = 667/6, because, he says, the narrative of events in Pausanias (IV 23. 4–10) implies that Miltiades' archonship followed immediately upon Autosthenes’. It is true that no gap of four years is marked: but the passages referring to the archonships of Miltiades and Autosthenes as having occurred in successive Olympiads are close enough for Pausanias to have known what he was doing: especially as the two Olympiads are distinguished by the second and third victories of Chionis.

⁸⁷ See below, under Damasias, archon 582–0, pp. 102 f. Von Schoffer, in RE IV 2035 f.Google Scholar cites Hertlein as having maintained (Neues Corresp.-Bl. 1895, 49 f.Google Scholar) for no apparent reason, that Dionysios meant to write Ol. 35. 1 = 640/39.

⁸⁸ Cf. ΑΘΗΡΑΔΑΣ

ΑΡΥΤΑΜΑΣ

ΧΡΥΣΟΜΑχΟΣ.

⁸⁹ See below, under Philombrotos, Solon and Dropides, pp. 92–9.

⁹⁰ On the whole question, see Müller, FHG II 14Google Scholar, Beloch GG I² ii 159, and Jacoby, in RE VIII 1927–30Google Scholar.

⁹¹ Hdt. V 71. 2; Thuc. I 126. 8, 11, 127. 1.

⁹² Hdt. V 72. 1; Thuc. I 126. 12.

⁹³ Epitoma Heraclidis 2 (ed. Kenyon).

⁹⁴ Hdt. V 71. 1; Thuc. I 126. 3.

⁹⁵ Ἀθ. πολ. 1–4, cf. epit. Heracl. 2; Thuc. I 126. 5.

⁹⁶ On the thesis of Beloch GG I² ii 302 ff. that Kylon's venture really belongs to the middle of the sixth century, cf. Busolt, Griech. Staats. 599¹Google Scholar, De Sanctis Ἀτθίς 280 ff., Adcock, F. E. in CAH IV 661 f.Google Scholar, Kahrstedt, in RE XV 125 f.Google Scholar, and on the argument from the use of a δημοτικόν below p. 115ⁿ.

⁸⁷ The Megakles who was party leader in the years preceding Peisistratos' first usurpation of 561/0 and successful wooer of Agariste about 572 was son of Alkmeon (Hdt. VI 127.4; Aristot. Ἀθ. πολ. 13.4). An Alkmaion was leader of the Athenian force in the Sacred Wa r which terminated in 586/5 (Delphian inscription noted by Plutarch Solon 11): presumably this is the man. Hdt. tells us Megakles' father, Alkmeon, enriched himself at the Lydian court (I 125—but we should understand Alyattes in place of Croesus) and that he won the prize at Olympia for chariot-racing with four horses. This victory is confirmed by Isokr. περὶЗεύγους 25, and may be the same as the victory ascribed to Megakles (in that case mistakenly) as having taken place in 592 (schol. on Pind. Pyth. VII). Now, this same Alkmeon, the father of Megakles the rival of Peisistratos, whose oruit seems from these various pieces of evidence to have fallen about 590, is himself called by Hdt. (VI 125. 1) ‘son of Megakles’. That a Megakles should have been archon in 636/5 or 632/1 is, therefore, not improbable.

⁹⁸ Ἀθ. πολ. 4. 1.

⁹⁹ ibid.—cf. 5. 2 —.

¹⁰⁰ Though, as I have said above (p. 78), this does not mean that no material for time-notes was available.

¹⁰¹ An alternative but unlikely solution would be to hold that the tradition about Drakon placed his ἀκμή or fortieth year (cf. Jacoby, Apollodors Chronik 43–6)Google Scholar in 640/39, but his legislation in the archonship of Aristaichmos some twenty years later; that these two dates were confused in Diodoros, but that a trace of the distinction is preserved in ‘Suidas’, who says Drakon legislated γηραιός ὢν.

¹⁰² Karst, J.Eusebius' Werke V 186Google Scholar (versio Armenia—but cf. the Series Regum p. 151 which gives forty-five years); Helm, R.Eusebius' Werke VII 98Google Scholar (versio Hieronymi). Both versions place the accession of Alyattes in 611/0 or 609/8. See also below on Euthydemos, pp. 108 f.

¹⁰³ Hdt. I 25.

¹⁰⁴ FGrHist 239 A 36 commentary (in II D p. 687).

¹⁰⁵ A majority of the MSS. of the versio Hieronymi give Ol. 45. 1 = 600/599, a variant is Ol. 45. 2 = 599/8. The Ol. 46. 2 = 595/4 of the versio Armenia is impossible (see below under Philombrotos, this page).

¹⁰⁶ On the meaning of μζτά, in this passage, see below under Dropides, p. 99.

¹⁰⁷ See assembly of the evidence in Jacoby, Apollodors Chronik 165–78Google Scholar. I have omitted one or two unimportant items.

¹⁰⁸ The desired (46) may lie behind either the (47) (so Jacoby, Apollodors Chronik 167Google Scholar, citing a similar error in Σ ad Demosth. XLV 64) or the (56) (so Beloch GG I² ii 163) of the text as we have it.

¹⁰⁹ See above, p. 78.

¹¹⁰ Cf. e.g. Bauer, A.Lit. u. hist. Forsch. zu Arislot. Ἀθ. πολ. (1891) 45–9Google Scholar; Blass, F.Lit. Centralblatt (1893) 1713Google Scholar; von Wilamowitz-Möllendorff, U.Aristot. u. Athen (1893) I 10Google Scholar; Sandys, J. E.Aristotle's Constitution of Athens pp. 49–51, 55Google Scholar; Schröder, O. in Philologus LIII = n. F VII (1894) 720 f.Google Scholar; Busol GG II² (1895) 301³; Kirchner, J. in Rh. Mus. LIII (1898) 380–6Google Scholar; Jacoby, Apollodors Chronik (1902) 165–72Google Scholar; de Sanctis, G.Ἀτθίς² (1912) 204 ff.Google Scholar; Beloch, GG I2 ii (1913) 161–5Google Scholar; Linforth, I. M.Solon the Athenian (1919) 265–8Google Scholar; Gomme, A. W. in JHS XLVI (1926) 178Google Scholar; Aly, in RE III A (1927) 948 f.Google Scholar; Hammond, N. G. L. in JHS LX (1940) 72 f.Google Scholar I am also indebted to T. E. Wright, E. Lobel, F. Jacoby and L. J. D. Richardson for giving me their views privately on 13. 1, 2. A conspectus of the views of the various writers cited is given in a table below, p. 103.

¹¹¹ See below, pp. 102 f.

¹¹² Cf. Aristotle himself elsewhere (Ἀθ. πολ. 22. 6), and Hdt. VII 20.1, etc.

¹¹³ The archonship was then the chief administrative post at Athens, and would not be left vacant if this were avoidable.

¹¹⁴ Four years: Bauer, Wilamovvitz, Busolt, de Sanctis, Beloch locc. cit. Three years: Schröder, Kirchner, Jacoby, Linforth, Hammond locc. cit.

¹¹⁵ The perfect participle should mean that the ταραχή had been going on till the moment of Solon's departure, when it ceased; and there is some support for this view in c. II, where it appears that Solon went away to put a stop to criticisms and agitation by the various parties. On the other hand, ἔτι implies that the ταραχή went on longer than you might expect, i.e. that Solon's reforms did not, after all, put a stop to it; on this view the ἁναρχίαι were instances of the continued ταραχή, and τεταραραγμένης, though a perfect participle, actually looks forward (so Hammond). Looking at the difficulty another way, we should imagine that ταραχή and ἡσυχία are contrary in meaning. A city could hardly, we suppose, be at the same time in a state of both; unless, by a stretching of the term ταραχή, we make it describe a state of ἡσυχία punctuated by periods of anarchy; otherwise we could only conclude that Aristotle thought of a period of ταραχή succeeded at a definite point of time by ἡσυχία: and this point of time could only be the assumption by Solon of the office of archon—which is hard to reconcile with ἔτι. I make no claim to settle this point.

¹¹⁶ On the argument that the archonship of Simon fell in 590/89, and that therefore the first ἀναρχία must be dated 589/8, see below under Simon, pp. 100 f.

¹¹⁷ In favour of a four-year interval: Bauer, Wilamowitz, Busolt, de Sanctis locc. cit. For three years: Schröder, Kirchner, Jacoby, Beloch, Hammond locc. cit. It is to be observed that the absurdity of reckoning καὶ πάλιν ἔτει πέμπτῳ exclusively is marked by the adherence of Beloch to the latter group: see above, n. 114.

¹¹⁸ So Wilamowitz, Busolt, Kirchner, locc. cit.

¹¹⁹ So de Sanctis, Beloch, locc. cit.

¹²⁰ So Blass, Schröder, Jacoby, Linforth, Hammond locc. cit.

¹²¹ Hammond (72⁹) says that the meaning ‘is clear from general Greek usage and from the example in Aristotle Politics (1275 a 23).’ This alleged parallel, first adduced by Blass and later by Jacoby (op. cit. 171), but challenged by de Sanctis (loc. cit.) cannot be accepted. The passage runs: —some offices can be held again by the same man only after a stated interval. The plural χρόνων here is due to the reference being to several offices, each one having its own prescribed interval between two tenures by the same man. It is hard to find other and more convincing parallels. In Ephoros (FGrHist 70 F 9) we find the expression διὰ τοσούτων meaning ‘after so long an interval,’ but it is uncertain what word should be understood with τοσούτων. A similar example to that from the Politics is to be found in Plato Tim. 22 d ( = recurring at long intervals (on several occasions)). As for ‘general Greek usage,’ the remaining passages cited by Jacoby loc. cit. (Aristot. 38, 833 a 19 (), [Plut.] . 194 (), Hypoth. ad Soph. El. ed. Pearson 1. 10 (), and Eus. χρονογρ. (ed. Schöne) I 162 ( = after the year of power ascribed to Philippos) merely illustrate that διά with the genitive in expressions of time means ‘after an interval of. It was necessary to point this out in order to show that the translation ‘immediately’ was inadmissible: but it does not bear on the anomalous use of the plural χρόνων. Linforth alone notices this anomaly, and suggests, not very convincingly, that it is ‘because he was thinking of the several terms of office involved in this thrice recurring period’. Another explanation, which I owe to L. J. D. Richardson, would be that the strangeness of the expression is due to the process whereby it was substituted for the phrase ἔτει πέμπτῳ. The stages of Aristotle's thought might be as follows: (i) ἔτει πέμπτῳ; (ii) (iii) (not Greek); (iv) (by analogy with ἔτῶν). Oragain, we might regard this passage as furnishing an early example of the use of χρόνος as equivalent to ἔτος: cf. Plat, νόμοι 778 b (apparently a genitive of time after which, though without διά), and LS⁹ χρόνος I 2 c, where a i/B.c. papyrus is cited. It is common in Byzantine and modern Greek: cf. E. A. Sophocles Greek Lexicon of the Roman and Byzantine Period s.v.

¹²² See below, pp. 102 f.

¹²³ See above, pp. 78 f.

¹²⁴ See below, pp. 104–6.

¹²⁵ So Kirchner (591/0) loc. cit., Beloch op. cit. 165.

This view involves omission or mistranslation of (see above, pp. 94f.). The text is also accepted without question by Linforth (266), but as he allows for exclusive as well as inclusive reckoning, and for several possible dates for Komeas, he arrives, on the basis of this passage, at 594/3–591/0 as the period within which Solon's archonship fell.

¹²⁶ So Bauer, Wilamowitz, Jacoby, de Sanctis, Gomme, Aly locc. cit.

¹²⁷ Busolt GG II² 258³; cf. Jacoby op. cit. 171 f.

¹²⁸ Beloch GG I² ii 165.

¹²⁹ loc. cit.

¹³⁰ See above, p. 78.

¹³¹ See below, p. 99.

¹³² Cf. e.g. Ἀθ. πολ. 17. 1, 2.

¹³³ Ἀθ. πολ. 5. 2.

¹³⁴ Linforth op. cit. 286 doubts the authenticity of the VOL. LXVIII. law referred to in T 2. However, even if it is in fact a later invention, it is evidence for the tradition that Solon's legislation was passed during his archonship.

¹³⁵ See below, p. 99.

¹³⁶ See below, p. 106.

¹³⁷ Cf. Hdt. I 29, V 113. 2, Aristotle Ἀθ. πολ. 11. 1, Plut, . Solon 25–8Google Scholar.

¹³⁸ e.g. he visits Amasis of Egypt, whose reign began in 569, and Philokypros of Cyprus, whose son Aristokypros fel in battle in 497 (Hdt. and Plut. locc. cit).

¹³⁹ Hdt. I 29–33. The reign of Croesus probably ran from 555 to 541 (or, according to Eusebios, from 560 to 546), and the visit of Solon is made to fall well on in it, after the enslavement of the Ionians, and two years (46. 1) or a little more, allowing for the death of Croesus' son, before he began the enquiries that led directly to his war on Persia and deposition. Herakleides of Pontos says Solon lived long after the first accession of Peisistratos (T 69), but is probably wrong (see below, under Hegestratos, p. 106).

¹⁴⁰ For example, Plutarch (Solon 23. 3) says he instituted a reward of 100 drachmas for (Athenian) victors at the Isthmia, which were first made a national festival about 581 (T 110, Solinus VII 14 (Mommsen)). Cf also Milne, J. G.CR LVII (1948) 1–3Google Scholar, who argues that the ‘increase in measures and weights and units of currency’ (Ἀθ. πολ. 10. 1) took place not very long before the usurpation of Peisistratos, after Solon had journeyed to Egypt to review trade prospects.

^140a See below, pp. 104–6.

¹⁴¹ For the former emendation, which has won wide acceptance (see above, n. 126), cf. Jacoby's convincing parallels in Apollodors Chronik 171. For the latter (mentioned by Gomme loc. cit. as palaeographically the less probable of the two, and by Schachermeyr, in RE XIX 165Google Scholar) cf. Demosth. κατὰ Μειδ. 154, where the same two alternative emendations have been suggested for by different editors. For the objections of Pomtow and Hammond to all emendations presupposing the expansion of cyphers, see above, n. 50. It is unlikely that the comedian Eupolis, in whose Δῆμοι the scholiast found the name, stated that he was archon after Solon. The phrase is a comment of the scholiast's, based on other evidence.

¹⁴² It is unlikely that the comedian Eupolis, in whose Δῆμοι the scholiast found the name, stated tha the was archon after Solon. The phrase μετὰ Σόλωνα ἄρξας is a comment of the scholiast's, based on other evidence.

¹⁴³ As is supposed by Kirchner, Rh. Mus. LIII (1898) 386Google Scholar and PA 4573, and Wilamowitz, Aristot. u. Athen I 7⁹, 179⁸⁴Google Scholar

¹⁴⁴ Cf. (406/5) Hypoth. ad Aristoph. Βατρ. I, Dionysios VII 1. 5.

¹⁴⁵ This Dropides was probably grandson of the archon of 645/4. Kritias (in T 1) is made to speak of him as πρό παππος, but this, as Kirchner, (RE V 1720)Google Scholar points out, makes the generations too long.

¹⁴⁶ See above, p. 87.

¹⁴⁷ See above, pp. 96–8.

¹⁴⁸ See below, this page

^148a That Delphian archives existed we know from Plutarch (Solon 11. 2). There would be no difficulty in relating the Athenian and Delphian lists accurately, for the Delphian archons, like the Athenian, took office at the summer solstice (cf. Thomson, G. in JHS LXIII (1943) 55)Google Scholar.

¹⁴⁹ See above, pp. 93 f.

¹⁵⁰ See below, p. 102; and above, pp. 83–6.

¹⁵¹ Simon 590/89: Wilamowitz, Aristot. u. Athen I 10f.Google Scholar (with hesitation), Busolt GG I² 697, Kirchner, J. in Rh. Mus. LIII (1898) 380, 386Google Scholar, de Sanctis Ἀτθίς² 203 ff., Beloch GG I² ii 143–5. Simon 591/0: Schröder, O. in Philologu LIII = n. F. VII (1894) 717–24Google Scholar, Jacoby, Apollodors Chronik 169Google Scholar, Das Marmor Parium 102–5, 165 f., Hammond, N. G. L. in JHS LX (1940) 74Google Scholar.

¹⁵² Cf. Jeanmaire, H.Couroi et Courètes 392 ff.Google Scholar, 401 ff. on the scope and historicity of this early festival; see also Lenschau, T.Philologus XCI = n. F. XLV (1936/1937) 396 f.Google Scholar

¹⁵³ One of the scholia gives the interval between the two ἀγῶνες as six years (T 77); this is probably a mistake for the interval between the capture of Kirrha and the final defeat of the Kirrhaians. The parallel scholium (T 75), though clumsily worded, seems to say that the final defeat of the Kirrhaians came after six years, and that the ἀγὼν στεφανίτης was later than that.

¹⁵⁴ T 110. The dates offered by the various MSS. are Ol. 49. 3 = 582/1, 49. 4 = 581/0, 50. 1 = 580/79 and 50. 3 = 578/7. It is unlikely that Eusebios was referring to anything but the first celebration of the numbered series. For discussion of the issue between the scholia and Pausanias see Busolt loc. cit. (with authorities cited, especially Bornemann, L. in Philologus L = n. F. IV (1891) 242–7)Google Scholar; also Wilamowitz, Kirchner and Beloch locc. cit. Schröder 719 suggests that Pausanias' mistake is due to his assuming that the ἀγὼν στεφανίτης fell in the same year as that in which the resistance of the Kirrhaians was finally overcome, instead of in that in which, six years before (according to the scholia), Kirrha was captured. As Kirrha was captured in the archonship of Simon in 591/0 (as Schröder held), the extinction of Kirrhaian resistance was in 586/5 = Ol. 48. 3, where Pausanias places the ἀγὼν στεφανίτης. But this explanation, though it accounts neatly for Pausanias' misdating of the ἀγὼν στεφανίτης, does not explain why he numbered the Pythiads wrongly.

¹⁵⁵ On the regulation of the calendar by means of an eight-year cycle see Lenschau loc. cit. and Thomson, G. in JHS LXIII (1943) 57–9Google Scholar.

¹⁵⁶ A 53: 215 years since the accession of Gelon, in the archonship of Timosthenes (478/7). A 55: 208 years since the accession of Hieron, in the archonship of Chares (472/1).

¹⁵⁷ Hammond (74) says that the assigning of Damasias' first year to 582/1 is ‘confirmed by the previous entry in the Marmor Parium on the ἀγὼν στεφανίτης, which gives the multiple of three (a nine-year interval) required between the two Pythian festivals’. The argument seems to be: Simon was archon in 591/0; Damasias' first year was perhaps 582/1; but this gives an interval of 3 × 3 = 9 years between the two festivals, which is what we should expect; therefore Damasias' first year was certainly 582/1. But why should we expect the festivals to be separated by an interval of nine years, or any other multiple of three ?

¹⁵⁸ See above, pp. 93–5.

¹⁵⁹ Alternatively, we may use Ἀθ. πολ. 13. 1, 2, together with the probability that Damasias' first year was 582/1, as additional ground for believing the Apollodorean date for Solon (see above, pp. 95, 98).

¹⁶⁰ Cf. the synchronization of Gylidas and Simon in the same scholia (see above, pp. 99 f.).

¹⁶¹ See above, p. 100.

¹⁶² We must here dispose of the view (held by Busolt, GG II ²301³Google Scholar, Wilamowitz, Aristot. u. Athen I 11Google Scholar and de Sanctis Ἀτθίς² 204 f.; cf. von Schoffer, in RE IV 2036)Google Scholar that the words in the Marble mean that the second year of Damasias is intended. are elsewhere used in the Marble, as we should expect, to distinguish archons of the same name (cf. A 36, 43, 48, 49, 64); and the reference here is to the earlier Damasias who was archon in 639/8 (see above, p. 91). The proper expression for a second archonship held by the same man is τό δεύτερον, and τό δεύτερου could not possibly stand for this (so Kirchner, J. in Rh. Mus. LIII (1898) 380 ff.Google Scholar; for the use of τό δεύτερον, cf. IG II² 389 and Dinsmoor, W. B.The Archons of Athens, etc., 19)Google Scholar. Yet another view, hazarded in spem secundam by de Sanctis (loc. cit.), is that the Parian, finding the name Damasias written twice in succession in his archon-list, wrongly assumed that these were two different men, and that the took place during the archonship of the second of the two: but this is equally untenable, for the Attic traditions from which the Marble partly derives must have included the facts known to Aristotle and stated in Ἀθ. πολ. 13. 1. It is unlikely, too, that the Parian directly consulted an archon-list (see above, p. 85).

One of the motives for these suggestions was the desire to harmonize the data of Ἀθ. πολ. 13. 1, 2 (as then commonly interpreted) with the dates of Solon and Damasias: combined with the excision of this interpretation of τό δεύτερου made it possible for Damasias archonship to follow directly upon the second ἀναρχία (see table on p. 103).

¹⁶³ See above, p. 78, n. 9.

¹⁶⁴ How did Aristotle know about the irregular archonship of Damasias, and particularly its exact length? Adcock, F. E. (in CAH IV 60)Google Scholar suggests it was by simple inference from the archon-list, which may have read , . The ten must have held office each for a month, which leaves two months for Damasias at the beginning of the year. But if he resigned office in the middle of an official year, it must have been by force. This argument is rather doubtful: all lists may not have read like that: and it is a pure guess that the ten held office each for a month. More probably oral tradition preserved these facts.

¹⁶⁵ On the text see Kenyon, F. G. in CR XIV (1900) 413Google Scholar: the Berlin papyrus reads not [τέττβρ]αςocs as was formerly supposed, but [πέν]τε, and thus agrees with the London papyrus.

¹⁶⁶ Rev. de Philol. n.s. XLVIII (1924) 144Google Scholar; cf. Wade-Gery, H. T. in CQ. XXV (1931) 79, 89³Google Scholar.

¹⁶⁷ loc. cit.

¹⁶⁸ So Meyer, E.Forschungen II 537–41Google Scholar and de Sanctis Ἀτθις 265 f.; cf. Ἀθ. πολ. 13. 2 . This hypothesis is however characterized by Glotz, G.Histoire Grecque I 443⁹⁰Google Scholar as ‘bien aventureuse’.

¹⁶⁹ On the character of Damasias' tyranny and the year of the ten archons, cf. von Schöffer, in RE IV 2036 f.Google Scholar, Kahrstedt, J.Staatsgebiet und Staatsangehorige in Athen 245Google Scholar, and Gernet, L.Rev. de Philol. 3° s. XII (1938) 216–27Google Scholar.

¹⁷⁰ Cf. Mommsen, A.Heortologie 120 f.Google Scholar

¹⁷¹ Hdt. VI 126–9.

¹⁷² Hdt. VI 35. 1, 127. 4, 128. 2.

¹⁷³ For recent discussion see Berve, H.Miltiades (Hermes: Einzelschriften II (1937)) 2–4.Google Scholar (1) The date of the wooing must fall before the death of Kleisthenes of Sikyon, whose rule of thirty-one years (Nik. Dam. in FHG III 394) embraces the Sacred War (Paus. II 9. 6, X 37. 6)—i.e. before about 565; on the other hand, it must be late enough to allow for the political activity of Kleisthenes, son of Megakles and Agariste (Hdt. VI 131. 1), in 508/7–i.e. after about 585. The wooing also takes place in the same year as an Olympic festival at which Kleisthenes has just won the chariot-race (Hdt. VI126. 2); and it is not unlikely that this victory would be near in time to his other chariot-victory at the Pythia in 582/1 (Paus. X 7. 6). 576/5 would therefore be a reasonable date. (2) In the text of Marcellinus the insertion of a Miltiades as father of Hippokleides is clearly a gloss. How the text ought to be emended (if at all) to agree with the statement of Hdt. that Miltiades the oikist was son of Kypselos is not so clear. To insert this Kypselos between Hippokleides and Miltiades is impossible, for Miltiades is already active in the Chersonesos before the fall of Croesus (Hdt. VI 37. 1), and there is not room for another generation between him and Hippokleides. Perhaps Kypselos was brother of Hippokleides' father Teisandros: for we know that Hippokleides was connected in his descent with the Korinthian Kypselids (Hdt. VI 128. 2). If he had no children and Kypselos was a relatively unimportant figure, it would be easy for a line of direct descent through Hippo kleides to become current. On the other hand, Wade-Gery, H. T. (in CAH III 570, 704 f.)Google Scholar tentatively identifies Kypselos and Hippokleides the archon, from whom consequently he separates Hippokleides the suitor, who must be roughly contemporary with Miltiades the oikist, son of Kypselos.

¹⁷⁴ See above, pp. 98 f.

¹⁷⁵ See partisans of the two dates listed by Schachermeyr, F. in RE XIX 171 f.Google Scholar

¹⁷⁶ See above, n. 141.

¹⁷⁷ See, for example, below, p. 118.

¹⁷⁸ Jacoby referred in 1904 to the already ‘fast unübersehbare Litteratur’ (Das Marmor Parium 167). The most recent full treatment, with references to earlier treatises, is that of Schachermeyr, in RE XIX 164–74Google Scholar.

¹⁷⁹ The same scholiast who quotes Eratosthenes' opinion tells us also that Aristotle said 41: Jacoby, Das Marmor Parium 170Google Scholar recommends emendation to 51 ( to ) rather than to 49 (ἔν to ἐννέα), and suggests that Ἀθ. πολ. 19. 6 may easily be emended correspondingly, to read 51 (ἓν καί for ἑνὸς δεῑ).

¹⁸⁰ See below under Philoneos and Harpaktides, pp. 109, 112 f.

¹⁸¹ This is more probable than Beloch's view (GG I² ii 161 f.), that the differences arose from varying calculations as to the length of 1½ generations.

¹⁸² Perhaps the μάλιστα is a concession to the fact that he is dealing in archon-years and his statements cannot therefore be accurate as to fractions of years (or even necessarily correct to the nearest natural year): a curious concession to make here, as it is made nowhere else in the Ἀθ. πολ.

¹⁸³ The date of Harpaktides, archon at the expulsion of Hippias, is fixed almost beyond doubt in 511/0; while that of Philoneos, archon when Hippias succeeded Peisistratos, falls by pretty general agreement in 528/7 (cf. Schachermeyr loc. cit., below, pp. 109, 112 f.; n. 184 for calculations displacing Philoneos).

¹⁸⁴ By rejecting some of the data, we can in fact arrive at dates both earlier and later than these. The addition sum presented by Ἀθ. πολ. 17. 1 and 19. 6, 33 + 17 = 49 (if this last figure is sound), can be justified if we assume that all three figures are inclusively reckoned. This implies the dates 559/8 for Komeas and 527/6 for Philoneos (511/0 for Harpaktides being regarded as fixed). The objections are: (1) the testimony of the Marble; (2) μάλιστα in Ἀθ. πολ. 19. 6 and the eighteen years of the Politics; (3) Eratosthenes' fifty years and the possible alternative reading fifty-one in the Ἀθ. πολ.; (4) the improbability that Philoneos was later than 528/7 (see below, p. 109). Again, we could take the thirty-three years of Peisistratos, the eighteen years assigned to Hippias in the Politics, and the fifty-one years for both conjectured by Jacoby in Ἀθ. πολ. 19. 6, and, assuming them all to be exclusively reckoned, date Komeas in 562/1 and Philoneos in 529/8. The objections to this view are: (1) the testimony of the Marble; (2) ‘about seventeen’ in Ἀθ. πολ. 19. 6; (3) Eratosthenes’ fifty years and the forty-nine in the unemended text of Ἀθ. πολ. 19. 6.

Gomme, A. W. in JHS XLVI (1936) 177 f.Google Scholar also reaches the conclusion that the duration statements do not enable us to decide between 561/0 and 560/59. He regards it, however, as improbable that the employment of different methods of reckoning account for the discrepancies in these statements, and prefers to derive them from traditions about the time of year at which the different events took place (cf. μάλιστα in Ἀθ. πολ. 19. 6).

¹⁸⁵ See below, pp. 107–9.

¹⁸⁶ See above, pp. 83–6.

¹⁸⁷ It is, ofcourse, possible that the words are not part of the citation from Phanias, but an editorial addition of Plutarch's.

¹⁸⁸ The story was already current in Aristotle's time that Solon tried vainly to obstruct Peisistratos' first coup d'état (Ἀθ. πολ. 14. 2, Plut., Solon 30Google Scholar, Valerius Maximus (T 64)): Aristotle further knew who was archon when Solon died, but does not tell us his name: he merely implies that it was a good many years before Peisistratos himself died (Ἀθ. πολ. 17. 2). That Phanias and Herakleides, both contemporaries and acquaintances of Aristotle's, should differ so sharply, is strange. There is a further difficulty about the place of Solon's burial: Aristotle and other good authors, according to Plutarch, said his ashes were scattered about the island of Salamis (T 69); Diogenes Laertios, expanding this and quoting from Kratinos to confirm the story, says he died in Cyprus and at his own request his bones were brought to Salamis, burned, and scattered there (T 34); Valerius Maximus says he opposed the tyranny of Peisistratos, died in old age an exile in Cyprus, and was not buried in his native land (T 64); while Aelianus says that he survived a short time the usurpation of Peisistratos, that they buried him publicly at Athens, and that his tomb became surrounded by other buildings (T 84); finally, ‘Suidas’ refers to the tradition that he died in Cyprus (T 125). We know that legend grew up rapidly round the figure of Solon (cf. the meeting with Croesus); and as we have considerable apparent discrepancy here, both as to the date and place of his burial, can we be sure that any version is true ? May it not be the case that, after the date and circumstances of Solon's death had long been forgotten, it was inferred from fragments of his poems (e.g. F XIII (Linforth p. 144)) that he outlived the establishment of the tyranny; but that he must have died very soon afterwards, as he would by then have been very old—perhaps even under the very next archon? (so, for example, Jacoby, in JHS LXIV 50⁶⁴)Google Scholar. We must not, on the other hand, overstress the contra dictions of the testimonia. The quoted statements of Aristotle and Kratinos about the ashes say no more than that they were scattered in Salamis. If Aristotle said that, he evidently did not think it inconsistent to refer to the known date of Solon's death in another part of his writings (Ἀθ. πολ. 17. 2): in other words, even if Solon did die in Cyprus and his ashes were brought and scattered in Salamis, there is no reason why the ceremony should not have been performed with the official recorded cognisance of the Athenian State. But if so, then it is not surprising that Phanias and Aristotle know what year he died in. Later on, Athenians who wondered where the tomb of Solon was and who had not heard of the story that his ashes were scattered in Salamis, assumed that his tomb must have been built over: this version of the story being preserved by Aelianus. Herakleides' statement that Solon lived many years after the usurpation of Peisistratos may be due to the desire to overcome the chronological difficulty of the alleged meeting with Croesus.

¹⁸⁹ See below, this and f. page.

¹⁹⁰ See the discussion, with citation of previous literature by Schachermeyr, F. in RE XIX (1937) 167–72Google Scholar.

¹⁹¹ Rh. Mus. LI (1896) 560–76Google Scholar, esp. 574 ff.; cf. Jacoby, Apollodors Chronik 183 ff.Google Scholar, Das Marmor Parium 171, and Kirchner, in RE VII 2606Google Scholar.

¹⁹² Hdt. I 60. 1. See the pertinent remarks of Cornelius, F.Die Tyrannis in Athen 6Google Scholar, and Gomme, A. W. in JHS XLVI (1926)Google Scholar, as against Pomtow loc. cit. and Adcock, in CQ XVIII (1924) 174 f.CrossRefGoogle Scholar

¹⁹³ In fact, the great majority of scholars cited by Schachermeyr 171 f. accept a duration of five years for the first tyranny.

¹⁹⁴ Hdt. V 65. 3.

¹⁹⁵ Hdt. I 62. 1, Ἀθ. πολ. 15. 2.

¹⁹⁶ Hdt. I 60. 1.

¹⁹⁷ Hdt. I 61. 2.

¹⁹⁸ The concept of legal φυγή goes back at least to Solon (T 2): that it would involve leaving Attika is implied by the συνοικισμός and by the ancient courts ἐν Φρεαττοῑ and ἐν Πρυτανείῳ (cf. Demosth. κατ' Ἀριστοκρ. 76–9, Plat, νόμοι IX 873 e, Harpokr. s.v. ἐν Φ, Pollux όνομαστ VIII, 120).

¹⁹⁹ See above, p. 80.

²⁰⁰ Otherwise great confusion would have occurred with some archons of the fifth century; Λυσίστρατος (467/6), Λυσανίας (466/5 and 443/2), Λυσίθεος (465/4), Λυσικράτης (453/2 ) and Λυσίμαχος (436/5)—six similar names in 32 years, and three in three years. Cf. also Φρασικλῆς (460/59), Φιλοκλῆς (459/8).

²⁰¹ See above, p. 106.

²⁰² The sending of an embassy with gifts may be referred to; and it has been reasonably conjectured (by Schubert, Könige von Lydien 92Google Scholar, cf. Töpffer, Beiträge 88 ffGoogle Scholar., Jacoby, Das Marmor Parium 171Google Scholar) that the accession of Croesus is intended: for the Parian gives us 605/4 or possibly 604/3 (see above, p. 92) as the date of Alyattes' accession, and this allows forty-nine years for his reign, which is the tradition conserved in Eusebios (cf. Karst, J.Eusebius' Werke V 188Google Scholar, R. Helm id. VII 102). That Eusebios puts the first year of Croesus in 562/1 or 560/59 need not disturb the dating of Euthydemos, for there were at least two Greek traditions about the date of the fall of Sardis.

²⁰³ Apollodors Chronik 183 ff. He points out that the parenthetical citation from Pamphila embraces καὶ πρῶτον ἔφορον γενέσθαι.

²⁰⁴ See above, p. 87.

²⁰⁵ See above, pp. 104–8.

²⁰⁶ See above, p. 106.

²⁰⁷ See above, n. 82.

²⁰⁸ VI 103. 3.

²⁰⁹ See above, pp. 105 f.

²¹⁰ See below, pp. 112 f.

²¹¹ See above, pp. 77 f. The commentary of Meritt, B. D. in Hesperia VIII (1939) 59–65Google Scholar should be followed for all the names.

²¹² See below, p. 110.

²¹³ Hdt. V 69. 1, VI 131. 1.

²¹⁴ Hdt. I 64. 3, V 62. 2, VI 123. 1, cf. Aristot. Ἀθ. πολ. 19. 3.

²¹⁵ Hdt. says (V 62. 3) that while in exile they repaired the temple of Apollo at Delphi: on archaeological grounds MissRichter, G. (in AJA XLI (1937) 162)Google Scholar is inclined to date the completion of the temple to the period 514–10. MissGuarducci, M. (in Annuario n.s. III–IV (1941/1942) 121 fGoogle Scholar.) says Meritt's hypothesis of a reconciliation and second banishment of the Alkmeonids is not required by the stone, which could in fact be read Π[λεισθεν]ες: but admits that this name is met, outside Sparta, only in Thasos at the end of the fifth century.

²¹⁶ Dionysios does not actually say that it was the first year of the sixty-fourth Olympiad. But it is the only date from the period before Ol. 68 for which the year of the Olympiad is not specified by Dionysios, and as the construction (ἐπί with the genitive) is the same as that used for the regular series beginning Ol. 68. 1, the assumption, that here Ol. 64. 1 was intended, is reasonable. Some assistance, though not absolutely precise, is afforded by the context. It was twenty years after the attack on Kyme in Ol. 64, that Aristodemos established himself there as tyrant (VII 5. 1), and he had been tyrant some fourteen years () when the envoys arrived from Rome (12. 1). Now these envoys were sent in Ol. 72. 2 = 491/0, the treatment of which year in Dionysios (VII 1–19) is interrupted by this digression on Kymaian affairs (that the year is Ol. 72. 2 is certain from the preceding and following sections and is also stated in 1. 5). Now if both those figures (20 and 14), one ordinal, one cardinal, are reckoned inclusively, we arrive at Ol. 64. 2 for the attack on Kyme and archonship of Miltiades. But Dionysios sometimes reckoned exclusively, with both cardinal and ordinal figures (cf. VII 1. 5), and the second figure (14) is stated to be approximate (όμοῦ τι). The most we could concede therefore would be that Ol. 64. 2 (but not Ol. 64. 3 or 4) is a possible alternative. It must be remembered that if Miltiades moves, the whole block from Onetor to Peisistratos moves with him.

²¹⁷ The most recent full discussion, with references to other views, is that of Berve in Hermes: Einzelschriften II (1937) 5–10Google Scholar. He did not, of course, know about the inscribed archon-list, which confirms us in believing that a Miltiades was archon in 524/3. He rejects the identification of the archon with any member of the Philaid family: the Peisistratids always filled offices with people from their own clique (Thuc. VI 54. 6), and Hippias would never have appointed this Miltiades archon so soon after he had compassed his father's murder. Moreover, Miltiades was too young to have been archon in 524/3—his birth must fall between 545 and 540. Per contra, Kleisthenes, also a member of a hostile family, was archon the previous year (Berve did not know this); we hear positively from Hdt. that the Peisistratids were effectively concealing their complicity in the murder of Kimon, and favouring his son Miltiades (VI 39. 1); and there is no reason why Miltiades, son of Kimon, should not have been born as early as (say) 555, even though he was his father's second son. He would, in this case, be only 65 in the year of Marathon. And even if he were born later, there is no reason why, especially under the tyranny, he should not have held the office of archon in his twenties. (The minimum age of 30 is attested positively only for the late fifth century and afterwards: cf. Busolt-Swoboda, Griechische Staatskunde 1070Google Scholar, Kahrstedt, U.Untersuchungen zur Magistratur in Athen 18 f.Google Scholar)

The sequence of events in the history of the family is as follows:

561/0: Peisistratos' first usurpation.

?: During one of Peisistratos' tyrannies, Miltiades, son of Kypselos, goes to Chersonesos (Hdt. VI 35–37). Berve's attempt to show (8 f.) that this must have been during the first tyranny, because Miltiades was impatient of the tyranny, and because Peisistratos was in exile from 556 to 546, whereas Miltiades went out before the fall of Sardis in 546, rests on the truth of one view of the tyrannies and exiles of Peisistratos and one view of the date of the fall of Sardis; moreover, why should not the tyranny have become intolerable to Miltiades by slow degrees ?

546/5 or 541/0: Fall of Sardis.

536/5: Kimon, son of Stesagoras, while in exile, wins chariot-race at Olympia (Hdt. VI 103. 2).

532/1: Kimon's second victory at Olympia—Peisistratos allows him to return ὑπόσπονδος (ib.).

528/7: Kimon's third victory at Olympia; Peisistratos dies, Kimon is murdered with complicity of Hippias; Miltiades, son of Kypselos, still alive in Chersonesos (103. 3, 4).

524/3: Miltiades, son of Kimon, archon in Athens.

Before or after 524/3: Miltiades, son of Kypselos, dies; Stesagoras, son of Kimon, becomes tyrant of Chersonesos (Hdt. VI 38. 1).

?: Stesagoras murdered; Peisistratids send out Miltiades, son of Kimon, to succeed him (38. 2, 39. 1).

?: Miltiades takes part in Darius' Skythian expendition (Hdt. IV 137 f., cf. VI 41. 3). Date may be about 512 (cf. How and Wells on Hdt. II p. 429): is certainly between about 517 and 511/0, but precision on this point does not matter here.

493: Miltiades escapes from Chersonesos to Athens—his son Metiochos is old enough to be in charge of a ship (Hdt. VI 41).

489: Death of Miltiades, after the failure of the Parian expedition (136. 3).

²¹⁸ Any other name would be unique, except Philliades, which occurs once on a vase (ARV, 16, no. 5); and Telliades would be possible (cf. Tellias).

²¹⁹ Hesperia VIII (1939) 62–5Google Scholar. He goes on to suggest, as an alternative year for the archonship of Peisistratos, son of Hippias (attested by Thucydides and an extant inscription (T 3)), one of the three years 499–6; perhaps the archon-ships of Peisistratos (in 497/6?) and Hipparchos (496/5) were acts of appeasement towards Persia after the disasters of the Ionic Revolt, in which Athens took part. A later date for Peisistratos' tenure may (he says) receive some support from an ostrakon found in the agora bearing the name , which indicates that he was politically prominent after 488/7, when ostracism began (Ἀθ. πολ 22. 3–6); but political activity is not impossible thirty-five years after a man's archonship. In any case, as Miss Guarducci points out (op. cit. 122 f.), the inscriber of the ostrakon may simply have meant it as a jocular reference to the great tyrant. A single ostrakon is not very weighty evidence. Cf. also Gomme, A. W. in AJP LXV (1944) 327 fGoogle Scholar.—heregards it as certain that the Peisistratids were in continuous exile from 511/0.

²²⁰ Hdt. I 61; cf. Thuc. VI 55, Aristot. Ἀθ. πολ. 17. 3—18. 1. According to Thuc., Hippias married Myrsine, daughter of Kallias, son of Hyperochides, and would have been the first of P.'s sons to marry.

²²¹ Cf. Schachermeyr, in RE XIX 171 fGoogle Scholar., and above, p. 107.

²²² Thuc. VI 54. 6, 7. The aorist participle () is to be understood ingressively (‘son of Hippias who became tyrant’): cf. T 22, Hdt. I 14. 1, V 92 ε 2, Isokr. Εὐαγόρας 39 and Forster, E. S.Isocrates' Cyprian Orations p. 91Google Scholar, LS ⁹s.v. τυραννεύω.

²²³ (T 3). There is still some disagreement among epigraphists as to how we should explain the relatively advanced style of the lettering and the words ἀμυδροῑς γράμμασι used by Thucydides to describe it. See Wilhelm, A.Beiträge zur gr. Inschriftenkunde 111 ffGoogle Scholar., Tod 8, Lauffer, S.AM LXII (1937) 110Google Scholar, Löwy, Sitz. Akad. Wien. CCXVI (1937)Google Scholar Abh. 4, 13, Welter, G.AA LIV (1939) 23 ffGoogle Scholar., and M. Guarducci op. cit. 118–20. The best explanation of ἀμυδροῑς γράμμασι, accepted by Wilhelm, Tod and Miss Guarducci, would seem to be that the paint, with which the letters had originally been filled, had worn off; and, with regard to the style of the letters, which have plainly not been re-cut, our knowledge of other dateable Attic inscriptions of this period is so slight, that we cannot say on epigraphic grounds alone that this inscription could not have been cut in 522/1.

²²⁴ Professor Meritt has pointed out to me that the tiny oblique mark, on the edge of the stone, just below the ∣ of . might be part of a ε: if so, the restoration would be less likely than , for example. The doubling of the ε could be paralleled from other official inscriptions, but not the use of ∣ for E∣. The mark, however, may not be a part of a letter: it is about 2 mm. higher than the supposedly corresponding stroke of the ε to the right. The average height of the letters is 12 mm., and the general alignment is very good (cf. the ΠΠ in the second name).

²²⁵ Hdt. VI 108. 4 and How and Wells ad loc., Thuc. III 68. 5, VI 54. 6. Cf. Miss Guarducci op. cit. 124.

²²⁶ Aristot. u. Athen II 301²⁰.

²²⁷ Dieterich, in RE I 1065 fGoogle Scholar.; cf. T 28.

²²⁸ It is, at any rate, consonant with what we know of his literary activity. He was writing as early as 498/7 (Pyth. X, cf. Scholia in Pind. ed. Drachmann, II 241 f.Google Scholar, Theiler, W. in Schr. Königs. Gesell. XVII (1941) 290Google Scholar), and as late as 448/7 (Ol. V, cf. Drachmann I 139, Theiler 261, 290). Possibly the sixty-six years, if not a mistake for 86, represent an inaccurate calculation of the time from his first composition to his death.

²²⁹ Diod. XI 79. 1 (T 43).

²³⁰ Per contra Beloch GG I² ii 163 believes that Abion was thought to be archon of Pindar's death and identical with Habron archon 458/7. That Pindar in fact lived a few years after that, he thinks, ‘tut nichts zur Sache’.

²³¹ Ἀθ. πολ. 19. 6, 22. 1.

²³² Hdt. V 55 f., Thuc. I 20. 2, VI 56. 2, 59. 4, Ἀθ. πολ. 18. 3, 19. 2. The scholiast on Aristophanes (T 80) appears to be in some confusion about the interval between the murder of Hipparchos and the expulsion of Hippias, unless his text is corrupt. ‘Three years’ and ‘four years’ may be exclusive and inclusive computations of the interval; but the ‘six years’ ascribed to Hdt. answer to nothing either in Hdt. or anything else in the tradition. Hence Dindorf's conjectural emendation, transferring the reference to the duration of the whole tyranny of the family.

The date in Eusebios' chronicle for the murder of Hipparchos, 520/19 or 518/7 (T 117) is wildly wrong.

²³³ The Parian's inaccurate method is illustrated in this epoch by his tacking of two chronologically distinct events—the murder of Hipparchos and the expulsion of Hippias—to one archon.

²³⁴ See above, pp. 83–6.

²³⁵ Thuc. VI 59. 4. Cornelius, F.Die Tyrannis in Athen 4Google Scholar suggests that Thuc.'s reckoning is in natural years, not archon-years. This too is a possibility: see below, n. 253.

²³⁶ Hdt. V 94 f., How and Wells ad loc. The capture of Sigeion must be earlier than 528/7 (death of Peisistratos), and later than c. 540, as Hegesistratos, its first tyrant (Hdt. loc. cit.) was born after 561/0 (Ἀθ. πολ. 17. 3). He accompanied Argeian troops which joined Peisistratos for the battle of Pallene (ib. § 4), but was not necessarily adult at the time. Beloch (GG I² ii 167) thinks the capture of Sigeion was c. 550.

²³⁷ Thuc. VI 57. 4, Aristot. Ἀθ. πολ. 18. 4, Diod. X 17. 2, Polyain. I 22, Eusebios (T 116).

²³⁸ So Belcch loc. cit., Cornelius, F.Die Tyrannis in Athen 91Google Scholar, Wade-Gery, H. T. in CQ XXVII (1933) 23Google Scholar. (He says that the name suggests the tyrants' circle, and that Skamandrios may have been the last archon nominated by the tyrannical administration. But it is hard to believe that even the tolerant Athenians would have allowed a nominee of Hippias to take office after the end of the tyranny.)

²³⁹ See above, pp. 83–6.

²⁴⁰ Cf. Munro in CR XV (1901) 357, Jacoby, Das Marmor Parium 14 fGoogle Scholar., 110, 173 f. For the name Λυσαγόρας cf. IG II 2165. I have recently examined the stone with Professor Meritt, who definitely confirms Munro's readings.

²⁴¹ See above, n. 200.

²⁴² That the new male choruses presuppose the tribal reforms of Kleisthenes (Wilamowitz, Hermes XX (1885) 66Google Scholar, cf. Jacoay op. cit. 110) was a theory based on Anth. Pal. XIII 28Google Scholar (Simonides?), which mentions a victory of a male chorus of the tribe Akamantis. But this does not mean that there were no contests before the ten tribes existed, And even if it did, we could hold that the institution of male choruses was dated too early by the Parian, without doubting that Lysagoras was archon in 509/8.

²⁴³ Hdt. V 74. 1.

²⁴⁴ Cf. the rather curious words of Aristotle on 412/1 (Ἀθ. πολ. 33. 1): .

²⁴⁵ See above, p. 78.

²⁴⁶ Cf. Wade-Gery, H. T. in CQ, XXVII (1933) 17–29CrossRefGoogle Scholar.

²⁴⁷ So Busolt GG II² 403ⁿ and Wilamowitz, Aristot. u. Athen II 81Google Scholar (that he names the year 506/5 (II 417ⁿ) is perhaps a slip); Beloch GG I² ii 166 suggests that Ἀλκμαίωνος got into the text of Pollux through confusion with the fact that Kleisthenes, the author of the reforms, was an Alkmeonid. But why ἐπὶ Ἀλκμαίωνος ?

²⁴⁸ Plut., Them. 23Google Scholar. 1. This Alkmeon might be a son of Alkmeonides son of Alkmeon (BCH XLIV (1920) 229 fGoogle Scholar.), and thus cousin of Kleisthenes.

²⁴⁹ In the foregoing paragraphs we have not discussed at all the circumstances under which Kleisthenes passed his reforms, or the status which he himself, as prime mover of the revolution, held: we have confined ourselves to those pieces of evidence which appear to bear directly on the names and order of the first four archons of the restored democracy. Conclusions about the latter have however sometimes been drawn from consideration of the former, and it will be as well to see how far such consideration may rightly affect our reconstruction of the archon-list.

First of all we may deal with the view of Ehrenberg, V. (Klio XIX. (1925) 106–10)Google Scholar, that Kleisthenes the reformerwa s himself the archon of 509/8. He says that no one in the sixth century could be without holding an official position, and scouts Busolt's suggestion (GG II² 402 f.) that Kleisthenes could have been a member of a college of νομοθέται. He holds that it was as archon that Kleisthenes introduced his legislation, and that his work was completed in the following year, when Isagoras was officially archon, but had, in fact, fled from the city. An alternative form of this view, advanced by Wilamowitz, (Aristot. u. Athen I 6, II 77⁸Google Scholar) and Jacoby, (Das Marmor Parium 174Google Scholar) suggests that Kleisthenes was the archon who replaced Isagoras in office after he had fled, and that he passed his laws as archon in the course of the remaining months of the year.. On the other hand, it is strange that all trace of Kleisthenes' supposed archonship has disappeared from the tradition. Aristotle calls him ‘champion of the people’ (Ἀθ. πολ. 20. 4), and would certainly have added that he was archon, if his name had been in the list at this point; and if he had replaced Isagoras and completed his year of office, it would have been his own name, and no Isagoras', which would, under the democratic régime, have survived on the list as that of the eponymous archon for the year, if we are to judge from the case of 412/1 (cf. Ἀθ. πολ. 33. 1, quoted above, n. 244). Wade-Gery (loc. cit.) holds that Kleisthenes introduced his laws as a private citizen, that they were approved by the βουλή and δῆμος and passed during the effective archonship of Isagoras; and that Aristotle knows that this was so because the laws were still extant in his time and bore the archon's name attached to them (cf. Ἀθ. πολ. 29. 3): now Isagoras' name could not have become annexed to the laws if he was a fugitive from the city at the time when they were passed. We might further add that the election of Kleisthenes to the archonship of 525/4 (not known by Ehrenberg) makes it somewhat improbable that he was elected for 509/8, though not impossible. In the fourth century re-election to the archonship was illegal (Ἀθ. πολ. 62. 3), and does not, in fact, occur either in the fourth or in the fifth century. The method of would, in any case, make chances of its occurring very slight, even if it were not illegal. But a Miltiades was apparently archon twice in the seventh century (see above, p. 90); and as, probably, all the archons down to 487/6 were elected (cf. Beloch GG I² ii 318 f., Busolt-Swoboda, Griechische Staatskunde 842 fGoogle Scholar., 887), cannot say with certainty that Kleisthenes could not have been re-elected (cf. Kahrstedt, Untersuchungen zur Magistratur in Athen 134–41Google Scholar). However, Ehrenberg's theory is totally lacking in positive evidence: there was no necessity for Kleisthenes to be archon in order to pass his reforms; and in view, too, of the claims of Lysagoras to the archonship of 509/8, we need have no hesitation in rejecting the view that Kleisthenes was archon in that year. Whether Isagoras, on the other hand, completed his year of office before he fled from the city, and if he did not, what arrangements were made for the filling of his position, is a further question. We can at least say that there is no evidence that Kleisthenes became archon in his stead.

The apparently different order of events in the narratives of Herodotos and Aristotle has caused a great deal of confusion among scholars. The account of Hdt. V 66, 69–73 relates the passage of the reforms before the intervention of Kleomenes, whereas that of Ἀθ. πολ. 20, 21 relates these events in the reverse order. In fact, an examination of the texts shows that there is no real discrepancy. Ἀθ. πολ. 20. 1–4 is nothing more than a précis of the Herodotean narrative, and it is clear that for the sequence of events Aristotle had no authority, other than Herodotos, from whom he could have checked his account in point of chronological exactitude. After completing this précis and adding one or two editorial comments (20. 4, 5), Aristotle goes on (21) to give a detailed exposition of the reforms, taken from another source, possibly a copy of the laws themselves (so Busolt and Wade-Gery locc. cit.). Various scholars have attempted to show that the Herodotean order must be correct, as against the supposedly differing order of Aristotle, and to harmonize the successive exchanges of power between Kleisthenes and Isagoras noted by Hdt. with the indications of the reconstructed archon-list. Among the more recent discussions is that of Schachermeyr, F. in Klio XXV (1932) 334–47Google Scholar (with all needful references to previous literature). He argues strongly that the new constitution was not only approved but actually in force by the time of Kleomenes' intervention, and in order to make time in the narrative for this to happen, he explains the words in Hdt. V 70. 1 (cf. Ἀθ. πολ. 20. 3 ) as referring to the election of the reforming archon Alkmeon for 507/6. The reforms are thus made to stretch into two archonships, and this accounts for their being dated by Aristotle and Polydeukes under different archons. His reasons for arguing that the new constitution was in force by the time of Kleomenes' intervention are: (1) The details in Ἀθ. πολ. 1 and Plut. Solon 12 about the formal expulsion of the polluted really belong not to Solonian times but to the expulsion carried out during Kleomenes' invasion. This is clear from the δημοτικόν attached to the name of their accuser Myron in the account of Plutarch: as the use of the δημοτικόν dated from the reforms of Kleisthenes, these must have preceded the expulsion (cf. Ἀθ. πολ. 21.4).

(2) The βουλή which resisted Kleomenes must have been the new Kleisthenic βουλή of 500.

(3) The 300 partisans of Isagoras (Hdt. V 72. 1), who are really identical with the 300 jurors of Plut. Solon 12, clearly belong to the new system often tribes and thirty τρίττυες.

(4) The 700 families expelled by Kleomenes are probably the new citizens enrolled by Kleisthenes (Aristot., Politics 1275 D 35 ff.Google Scholar).

We may readily agree with Schachermeyr that the passing and putting into force of the reforms may have extended into two years, in the sense suggested above (p. 114), that they must have come into operation at the beginning of an official year; and all available time between the approval of Kleisthenes' measures and the beginning of the next official year must have been needed for the preparatory work of registration. But to assert positively that the intervention of Kleomenes did not interrupt this preparatory work, but came when it was completed and the new régime was in full force, is more than the evidence entitles us to do. None of Schachermeyr's arguments is conclusive:

(1) Athenians must sometimes have been referred to by their villages before Kleisthenes' reforms: in fact, Hdt. I 60. 4 gives Phye a δημοτικόν, and Aristotle says of the period before Peisistratos (Ἀθ. πολ. 13. 5) . And the identification of Kleomenes' expulsion of Athenian families with the expulsion of Ἀθ. πολ. 1 and Plutarch Solon 12 is most doubtful.

(2) We need not assume that the new Kleisthenic βουλή of 500 was so vastly different in temper from the βουλή of 400, which it replaced. No doubt many of the supporters of the tyranny fled with Hippias, and their places would have been filled by men of a more independent spirit: so that it would not be surprising if the opposition to Kleomenes came from the βουλή of 400. According to the view of Wade-Gery (loc. cit.) this βουλή must, in any case, have discussed and approved the laws of Kleisthenes before they were submitted to the ratification of the people.

(3) We cannot believe that Isagoras chose his 300 partisanjurors in equal numbers from the city, inland and shore regions to suit Kleisthenes' new system! The oligarchic clubs seem a likelier source.

(4) This does not prove that the tribal reforms were completed.

It is not, therefore, necessary to suppose that the new constitution was already in force when Kleomenes came. Nor need we explain Hdt.'s remark about the ‘defeat’ of Isagoras as referring to the election of Alkmeon as archon for 507/6: it can quite well refer to the political ascendancy of Kleisthenes, which enabled him to get his reforms approved by the βουλή and δῆμος, in spite of the fact that Isagoras was archon. The attachment of the reforms to different archons' names by Aristotle and Pollux is explained by the fact that they were passed in Isagoras' archonship and became effective from the beginning of Alkmeon's.

Beloch (loc. cit.) also believed, like Schachermeyr, that the instalment of the Kleisthenic βουλή of 500 necessarily preceded the intervention of Kleomenes; and for this reason he argued that, if we do not reject Alkmeon from the archon list altogether, we must insert him before Isagoras. But his premise is, as we have seen, without grounds.

The order of events, so far as we can tell, may have been as follows:

510/9: Skamandrios archon (from pro-tyrant family, but personally harmless?).

509/8: Lysagoras archon.

Summer 508: Kleisthenes' attempt to control the elections for 508/7 in favour of the reform-party fails (), and Isagoras is elected (Hdt. V 66. 2, Ἀθ. πολ. 20. 1).

508/7: Isagoras archon. Kleisthenes appeals to the people, and they pass his programme of reforms, in face of Isagoras' official opposition (Hdt. V 66. 2, 69. 2, Ἀθ. πολ. 20. 1).

507/6: Alkmeon archon. First year of new régime. Where exactly in this framework to place the invasion and discomfiture of Kleomenes is somewhat uncertain. The incident was not protracted, the siege operations lasting only three days (Hdt. V 72. 2). Perhaps Isagoras may be expected to have sent for help to Kleomenes as soon as it was clear that the people would pass Kleisthenes' proposals, i.e. during his own archonship in 508/7 (in which case Isagoras did not complete his year of office and a substitute must have been appointed): or perhaps it was later, during the archonship of Alkmeon, after the new régime had already begun to operate. All one can say is that this intervention of Kleomenes was closely connected in the tradition with the reforms of Kleisthenes and must have taken place fairly soon after the reforms were first approved by the people.

²⁵⁰ Cf. Wilamowitz, Aristot. u. Athen II 82²⁴Google Scholar, Beloch GG I² ii 170.

²⁵¹ He appears as archon for 495/4 in Wilamowitz op. cit. I 114—apparently a slip.

²⁵² I 93. 3. A scholium on the passage says (ed. C. Hude) . Whether τὰ Μηδικά here means 490/89 or 480/79, we cannot say: in either case the statement is wrong, as the archons of 491/0 and 481/0 are known. Gomme, A. W. (AJP LXV (1945 323Google Scholar) thinks it strange that the archonship of Themistokles is not mentioned in the biographical traditions used by Plutarch, and that so long a gap—some twelve years—followed before he became prominent again (Hdt. VII 143. 1); and he suggests that the archonship may be fictitious, and that the office referred to by Thucydides may be something else, perhaps that of . But we really do not know enough about Themistokles' life to doubt the archonship, for which the official list was probably the ultimate authority. Cf. also Munro, J. A. R. in CR VI (1892) 333 fGoogle Scholar., and Sandys, J. E.Aristotle's Constitution of Athens 90 f.Google Scholar

²⁵³ No reasonable doubt can be entertained about the dates of the last fourteen archons of our period whose names survive, from Hipparchos (496/5) to Hypsichides (481/0). After him begins the undisputed series of names from Kalliades (480/79) preserved in Diodoros. The interval between the battles of Marathon and Salamis, as apparently given in Herodotos and Thucydides, has, however, troubled some historians, and J. A. R. Munro was even led to date Marathon one year earlier than everyone else, i.e. in 491 (CAH IV 232 f.). As by general agreement the battle was fought in September, this means 491/0, the year of Hybrilides: so that Munro, although he does not explicitly say so, means us to reject the testimony of Aristotle, the Parian and Plutarch, who say it was in the archonship of Phainippos (490/89). His arguments are two: (1) Herodotos' narrative of the events preceding the battle show that it fell in the year following that in which Mardonios' expedition failed (492): the year apparently inserted by Hdt. between these two is fictitious; (2) Herodotos, Thucydides and Aristotle all give the interval between the battles of Marathon and Salamis (480) as not ten, but eleven years.

But (1) Hdt.'s chronology of the period from the Ionian revolt to the Battle of Marathon is intended to be complete, although it contains a possible asyndeton at VI 94. 1 after the digression on the fate of Kleomenes and the Athenian war with Aigina (cf. Macan, R. W.Herodotus IV–VI II 71–7Google Scholar). But, apart from Hdt., we have no date for any of these events except the Battle of Marathon, on the date of which accordingly the chronology of the whole period depends. The proper solution for Munro, if he wants to withdraw a fictitious year from Hdt.'s narrative, is not to pre-date Marathon one year, but to post-date the Ionic Revolt one year.

(2) Munro refers to the passages on which he bases this conclusion (Hdt. VII I, 4, 7, 20; Thuc. I 18, 118, VI 59, VIII 68; Ἀθ. πολ 22), but does not proceed to exegesis. Ἀθ. πολ 22. 3–8 will be discussed in detail below (pp. 118 f.). If Munro wants to insist on the text of 22. 8 and to place the archonship of Hypsichides in the fourth year from that of Nikomedes, it will mean either rejecting the testimony of Dionysios, or moving the Battle of Salamis down to 479/8! As to Herodotos, there is no difficulty in equating his time-notes with the received dates of the two battles:

Sept. 490 (i.e. 490/89): Battle of Marathon.

490: News of Marathon reaches Darius (VII 1. 1).

487: In the fourth year Egypt rebelled (VII 1.3).

486: In the next year Darius died (VII 4).

485: In the next year Egypt was subdued (VII 7).

481: In the autumn of the fifth year from the subjugation of Egypt, and after four full years of preparation, the march on Greece began (VII 20. 1).

480: After wintering at Sardis the army set forth (VII 37. 1).

Sept. 480 (i.e. 480/79): Battle of Salamis.

Of the four passages of Thucydides referred to by Munro, only two have any bearing on the date of Marathon: one (I 18. 2) places the expedition of Xerxes in the tenth year from Marathon; and if inclusive reckoning is meant, this can be met only by reducing, not expanding, the accepted interval between the two battles, or by supposing the expedition of Xerxes here to mean the march forth from Susa; possibly, however, Thucydides used exclusive reckoning; the other (VI 59. 4) represents Hippias as accompanying the barbarians to Marathon in the twentieth year after his expulsion. This is a real difficulty. In archonyears, with inclusive reckoning, the figure should be twenty-two (511/0–490/89). Munro can justify the twenty on his theory only by assuming natural years and exclusive reckoning (511–491). Or perhaps both 10 and 20 are to be taken as round numbers: for Thucydides is fond of them (cf. I 13, 3, 4, 18. 1, 118. 2, VI 2. 5, 3. 2, VIII 68. 4).

²⁵⁴ Demetrios also said Aristeides was appointed by lot, which is in harmony with his own date for Ar.'s archonship, but not with the received date. Ἀθ. πολ 22. 5 says that the first appointments by lot after the tyranny were in 487/6; and idomeneus (early third century) confirms the received date for Ar.'s archonship when he says he was elected (T 71). It would have been surprising if Aristeides had been appointed by lot, after 487/6; for, the πρόκριτοι being 500 in number, the chances that a leading figure would be appointed were small. While election, however, was still the rule, quite a number of notable men became archon: Isagoras, Alkmeon, Hipparchos, Themistokles, Aristeides. Jacoby (FGrHist 228 F 44 comm.) suggests that Demetrios may have been referring to some ἀρχή of Aristeides' other than that of the eponymous archon. The fragment, he says, is apparently from his Σωκράτης, but he may have included lists of minor officials in his . In that case. Plutarch mistakenly took him to be referring to the eponymous archonship. In any case it is extremely unlikely that a serious discrepancy in this part of the archonlist would survive to the end of the fourth century, even if it had existed in the fifth.

²⁵⁵ Cf. Wilamowitz, Aristot. u. Athen I 25Google Scholar, Beloch GG I² ii 168.

²⁵⁶ See IG I² 3. 16 f., 4. 26 f., and above, p. 80: though the readings and are plausible, the date of the incription is still under discussion. Cf. SEG X 5.

²⁵⁷ Although the testimonium does not give the Olympiaddating, it can be inferred from the context, as D. is here writing an account of every year: and it is confirmed by the lapse of 270 years since the foundation of Rome, which D. dates in Ol. 7. 1 (T 44).

The MSS. of Dionysios read Νικοδήμου, and this is supported by the Berlin Papyrus of Ἀθ. πολ 22.4–7 (Ν.. οδημο): the London Papyrus reads Νικομήδους. Nikodemos is thus some-what better attested, and is also the more probable name (see Kirchner PA s.vv.). The mistake is an easy one and is paralleled in Paus. III 19. 4, IV 31. 12 (cf. I 29. 15).

²⁵⁸ Nepos Arist. 1. 5 says ‘sexto fere anno’ which is clearly a mistake.

²⁵⁹ So Wilamowitz, Aristot. u. Athen I 25 fGoogle Scholar.: per contra Judeich, in RE II 880 fGoogle Scholar., cf. Sandys, J. E.Aristotle's Constitution of Athens p. 88Google Scholar. Eusebios also had a date for the ostracism of Aristeides: the majority of MSS. of the versio Hieronymi give 484/3, a minority 483/2 (T 118).

²⁶⁰ See above, p. 117.

²⁶¹ See above, p. 92.

²⁶² Cf. Pearson, L.The Local Historians of Attica 111Google Scholar.

²⁶³ Michaelis, in Hermes XXI (1896 ) 493–5Google Scholar said Kebris was not an Attic name, and that the text was corrupt. Moreover, no statue would have survived the destruction carried out by the Persians. He therefore suggested that Akestorides, archon 474/3, is intended. Wilamowitz ib. 600² (cf. Aristot. u. Athen II 81¹⁴)Google Scholar thought there was nothin peculiar about the name Kebris; and that if the Persians destroyed the original statue, the Athenians no doubt replaced it. The statue is also referred to in Paus. I 15. 1 (which throws no light on the date of its erection), and in Lucian 33, where reference is made to its ancient hair-style, . With his we may compare Thuc. I 6. 3, but the indications are not precise enough to warrant a more exact date than that suggested in the text.

²⁶⁴ See above, p. 109.

²⁶⁵ IG I² I = SIG III 13 = Tod 11 = SEG III 1. The origina l fragment ran (ll. 11 f.) ¹²; . Luria, S. (Comptes rendues de I'Acad. des Sciences de Russie (1924) 134 ff.)Google Scholar read . Schweigert, E. (Hesperia VII (1938) 264)CrossRefGoogle Scholar, after the discovery of the new fragment, read ; and Meritt, B. D. (Hesperia X (1941) 301–7)CrossRefGoogle Scholar argued for some such reading as . In the same year, however, Roussel, P. (Revue Archiologique XVIII 213 f.)Google Scholar claimed that the new fragment did not exclude a proper name, and conjectured . The name Βολακλῆς occurs once in the fifth century, borne by a Delian (IG I² 377 1. 4), and once at the beginning of the fourth (IG II² 1929 1. 10). Wade-Gery, H. T., however, in CQ XL (1946) 104Google Scholar, prints Meritt's reading without comment, though adding, as a possible alternative, .

²⁶⁶ An asterisk by an archon's name means that the date is relatively uncertain: but there are of course many degrees of probability, even where no asterisk is put.

²⁶⁷ Brackets indicate events whose connexion with the archon was probably calculated, not recorded or remembered (see above, pp. 80–82).