^{page 41 note 1} Beloch, K. J., Die Attische Politik seit Perikles, P. 324.Google Scholar

^{page 41 note 2} e.g. West, A. B. in A.J.P. xlv (1924), 158–60.Google Scholar

^{page 41 note 3} Wade-Gery, H. T. in C.Q. xxiv (1930), 34.Google Scholar

^{page 41 note 4} Sealey, Raphael in Proceedings of the African Classical Associations i (1958), 74Google Scholar; Mattingly, H. B. in Historia xii (1963), 264.Google Scholar

^{page 41 note 5} Gomme, A. W., Commentary on Thucydides, ii. 363.Google Scholar

^{page 41 note 6} Lewis, D. M. in J.H.S. lxxxi (1961), 119.Google Scholar

^{page 42 note 1} In the speech Against Alkibiades it is stated that Hipponikos died while a general at Delion ([And.] 4. 13). This is probably a mistake, arising from confusion between Hipponikos and Hippokrates, but it cannot be definitely disproved; see Burn, A. R. in C.Q. N.S. iv (1954), 139.Google Scholar Therefore in my main argument I assume that Hipponikos was not a general in 424/3, but I add an occasional footnote to point out what difference it makes if he was.

^{page 42 note 2} Cf. Gomme, , Comm. on Th. iii. 526.Google Scholar As he says, ‘the three stand and fall together’; in Th. 4. 65. 3 the words … apply to them all, and prove that none of the three was expected to leave Sicily in the summer of 424. This refutes the suggestion of Sealey (P.A.C.A. i. 84) that only Pythodoros and Sophokles were reelected for 424/3 and Eurymedon was not.Google Scholar

^{page 42 note 3} This odd suggestion is made by Gomme, , Comm. on Th. iii 506.Google Scholar

^{page 43 note 1} Sealey, (P.A.C.A. i. 83)Google Scholar rejects altogether the idea that by-elections were held, on the ground that ‘no regular practice of holding bye-elections to replace fallen or disgraced strategoi is attested in Athens’. Instead he suggests (attributing the suggestion to Wade Gery; though actually Wade-Gery, , in C.Q. xxiv. 38,Google Scholar n. 3, regarded this as the less probable alternative) that Thucydides was not a general of the normal kind in 424/3 but held a special command analogous to Kleon's special command at Pylos in 425. The objections to this view are overwhelming:

(a) Thucydides calls himself and Eukles (4. 104. 4) without giving any hint that the one had a different status from the other. But when Kleon's special command at Pylos is described (4. 28–29), Thucydides never calls him .

(b) If there were fourteen generals in 424/3 but no by-elections, that would mean four special commands in one year. But no ‘regular practice’ of this sort is ‘attested in Athens’ either.

(c) When generals were killed or dismissed in the course of their year of office, what would one expect but by-elections? In wartime it is unwise to do without generals. It might be reasonable to allow one or two vacancies in a board of ten to remain un filled, but what if the number of vacancies was larger? It is not true to say that the death or dismissal of a general was a rare event, so that no arrangement for appointing suffecti was necessary. Athenian politics and warfare being what they were, it was not unusual for a general to be dismissed or killed; and in 424/3 at least four generals were dismissed and one was killed. In 406 nine generals were dismissed; and what happened then? There was a by-election (X. Hell. I. 7. I).

^{page 43 note 2} And Hipponikos, if he was a general this year.

^{page 43 note 3} If Hipponikos was a general this year, he must be the one.

^{page 44 note 1} Four, if Hipponikos was also a general and was killed at Delion.

^{page 44 note 2} It is implicit in the arguments of West (A.J.P. xlv. 151),Google Scholar Wade-Gery, (C.Q. xxiv. 34, n. 2),Google Scholar and Lewis, (J.H.S. lxxi. 125).Google Scholar

^{page 44 note 3} To save space I henceforth refer to the tribes by numbers only, after giving the names here for reference: I Erekhtheis, II Aigeis, III Pandionis, IV Leontis, V Akamantis, VI Oineis, VII Kekropis, VIII Hippothoontis, IX Aiantis, X Antiokhis.

^{page 44 note 4} Wade-Gery's arguments for this resto ration are given in C.Q. xxiv 33–39.Google Scholar Some of his arguments are bad ones: one of them is that all ten generals must belong to differ ent tribes, and another is the very identi fication of Nikostratos son of Dieitrephes with Nikostratos the Skambonides which I am at present questioning. Nevertheless there are enough good arguments to make the restoration convincing: it fits conveniently in the inscription (where the name of a general is required; the suggestion of Lewis, , in J.H.S. lxxxi. 119,Google Scholar that the name of a Hellenotamias would do instead, is not persuasive; payments elsewhere in this inscription seem to be made to generals, some times in conjunction with Hellenotamiai, but not to Hellenotamiai alone), it does not conflict with any known evidence about the tribes of other men who were generals in the same years as Eurymedon, and the only other Eurymedon known in this or the next generation was of Myrrhinous (Diogenes Laertios 3. 42). (The suggestion of Gomme, , Comm. on Th. iii. 628, to the effect that Eurymedon the general may not have been the father but merely a maternal uncle of the known Eurymedon of Myrrhinous, is un convincing, because Athenians were often named after their fathers but seldom after their maternal uncles.) It will be observed that the arguments for saying that Eury medon was of Myrrhinous, which I accept, are similar to the arguments so far mentioned for saying that Nikostratos son of Dieitrephes was a Skambonides, which I regard as inadequate. The difference in my attitude to the two cases results from one vital difference in the facts: Nikostratos was a common name, but Eurymedon a very rare one.Google Scholar

Earlier Busolt (Hermes xxv [1890], 571)Google Scholar argued for the restoration or in I.G. i². 297.4, which he thought belonged to the year 427/6. West, (A.J.P. xlv. 151)Google Scholar accepted it, and proceeded to argue on this basis that Eurymedon was of IV. But meanwhile Bannier, W. (Berl. Phil. Wschr. xxxv [1915], 1613)Google Scholar had shown that the inscription should be dated to 414/3. Consequently the restoration of Eurymedon's name in the early part of it is no longer appropriate, and West's argument is invalid. Cf. Meritt, B. D., Athenian Financial Documents, pp. 88–92,Google Scholar and S.E.G. x. 229.

^{page 45 note 1} It has sometimes been suggested that two generals might be elected in a single year, but on this point I agree with the sceptical view of Hignett (History of the Athenian Constitution, pp. 349–51). Sealey, (P.A.C.A. i. 66)Google Scholar only repeats old arguments with which Hignett has already dealt. See also Lewis, in J.H.S. lxxxi. 118.Google Scholar

^{page 45 note 2} See Sealey, in P.A.G.A. i. 67 and 82Google Scholar, Dover, K. J. in J.H.S. lxxx. 66,Google Scholar Lewis, in J.H.S. lxxxi. 119–21.Google Scholar

^{page 46 note 1} Lewis, (J.H.S. lxxxi. 119–21) suggests that Lakhes was dismissed from the general ship in the middle of 426/5, and that Pythodoros may have replaced him at a by- election and thus have belonged to VII. I find this hypothesis unconvincing, not merely because there is no good evidence that Lakhes was dismissed (Th. 3. 115. 2 says only that Pythodoros took over from him the command of the ships in Sicily), but because it would require an unnatural order of events. If the Athenians suspected Lakhes of misconduct in Sicily, the natural thing to do would not be to set about holding a byelection before recalling him, but to send at once one of the generals already in office. They sent Pythodoros, and when they sent him they already had the intention of sending Sophokles too (Th. 3. 115. 5). I conclude that, even if Lakhes was dismissed, Pythodoros and Sophokles were already in office before his dismissal, and so did not belong to the same tribe.Google Scholar

^{page 46 note 2} This little chain of argument has one weak link. It is possible (even if not very probable) that Hipponikos was a general not in 426/5 but in 427/6 (cf. my Andokides: On the Mysteries, p. 10, n.11 and Mattingly in Historia xii. 260–1).Google Scholar If so, we do not know who, if anyone, was elected in 426/5; and that leaves open the possibility that Pythodoros or Sophokles belonged to II or VII or X (but not X if Hipponikos was a general in 424/3; cf. p. 42, n. But since this possibility is a rather slight one, and since anyway it does not affect my even tual conclusion about Nikostratos, I have banished it to footnotes.

^{page 46 note 3} Not X, if its general at the beginning of the year was Hipponikos.

^{page 47 note 1} If Pythodoros or Sophokles belonged to II or X (cf. p. 46, n. 2), so as to leave no general in this year known to have belonged either to VII or to VIII, that would open up further alternative solutions for Aristeides:

(iv) Aristeides may have belonged to VII, and VIII have remained unrepresented throughout the year.

(v)Aristeides may have belonged to VIII, and VII have remained unrepresented throughout the year.

(vi) If Aristeides died in the course of the year, he may have belonged to II or IV or X (whichever of these was not represented by Pythodoros and Sophokles), still leaving either VII or VIII unrepresented through out the year and leaving either VII or VIII represented by another general.

But these three alternatives are all fairly unlikely, because they depend on the assump tion that Hipponikos was not a general in 426/5.

^{page 47 note 2} If solution (iv) or (v) of the problem of Aristeides (in n. i above) is right, Eukles may have succeeded Aristeides in VII or VIII. If solution (vi) is right, he may have re presented VII or VIII from the beginning of the year.

^{page 47 note 3} By Westlake, H. D. in Hermes lxxxiv (1956), 114–15.Google Scholar

^{page 47 note 4} But it should be observed that Westlake, (Hermes lxxxiv. 115, n.7) suggested 418/7 as a possible alternative date for the anecdote about Nikias and Sophokles. If this is right, the remaining evidence about 423/2 tells us nothing about Nikostratos’ tribe, and the possibility that he belonged to III remains open.Google Scholar

^{page 48 note 1} See Kirchner, J. E. in Hermes xxxi (1896), 256.Google Scholar

^{page 49 note 1} My distribution of the dialogue is the same as W. J. M. Starkie's, but my reasons differ from his. See p. goy of his edition of the Wasps.

^{page 49 note 2} See Lowe, J. C. B. in Bulletin of the Institute of Classical Studies ix (1962), 27–42.CrossRefGoogle Scholar

^{page 49 note 3} So Meineke, Bergk, and Starkie.

^{page 50 note 1} Similar arguments would still apply if Xanthias, not Sosias, were the slave who pretended to hear the audience's suggestions; it would still be necessary to assume that Amynias and the other three were sitting near the front. Thus, even if my distribution of the dialogue is rejected, the rejection does not invalidate my subsequent conclusion about Nikostratos.

^{page 50 note 2} For a fuller list see Pickard, A. W. Cambridge, The Dramatic Festivals of Athens, PP. 275–8.Google Scholar

^{page 50 note 3} Ar. Knights 575 implies that generals customarily receive this privilege now, although in earlier generations they did not.

^{page 50 note 4} Russo, C. F., Aristofane autore di teatro, p. 195, suggests that Aristophanes may have made a prior arrangement with these members of the audience. But this cannot be right, because the jokes are jokes against them, to which they would not have sub mitted voluntarily.Google Scholar

^{page 50 note 5} In Wasps 1267 he is called but this does not mean that Amynias son of Sellos is a different man from Amynias son of Pronapes (Wasps 74), since only means ‘son of boaster’. So Meineke, , Frag ments Comicorum Graecorum, ii. 585, followed by others.Google Scholar

^{page 50 note 6} Kaibel, G. in Hermes xxx (1895), 441–5.Google Scholar

^{page 51 note 1} The weakest is the one based on Wasps 466–77, where eight lines separate Amynias from Brasidas, and it is not obvious that the two are meant to be closely connected.

^{page 51 note 2} It does not necessarily follow that Sosias and Derkylos (Wasps 78) were generals in 423/2, since they may have held other offices; but clearly it is possible that they were. It may be just worth pointing out that Nikias is only three letters different from the sus picious reading and that Nikias is known to have been a general in 423/2 (Th. 4. 129. 2, 133. 4).

^{page 51 note 3} I am grateful to Professors K. J. Dover and H. D. Westlake for helpful comments on this article.