¹ On the date of the trial see MacDowell, D., Andokides, On the Mysteries (Oxford, 1962), pp. 204–5Google Scholar, where he favours 400.

² On the chronology of this period see Meritt, B. D., ‘The Departure of Alcibiades for Sicily’, AJA 34 (1930), 125–52CrossRefGoogle Scholar; Athenian Financial Documents (Ann Arbor, 1932), pp. 152–79Google Scholar; “The Chronology of the Peloponnesian War’, Proc. of the American Philosophical Society, 115 (1971), 97–115Google Scholar; Dover, K. J., ‘Excursus on the Herms and Mysteries’, in A Historical Commentary on Thucydides iv (Oxford, 1970), pp. 265–88Google Scholar; MacDowell, op. cit. (n. 1), pp. 181–9; Aurenche, O., Les Groupes d' Alcibiade, de Léogoras et de Teucros, Remarques sur la vie politique Athénienne en 415 avant J.C. (Paris, 1974), pp. 155–8Google Scholar.

³ MacDowell, op. cit. (n. 1), pp. 178–80. Andokides indicates that his freedom was the issue when he decided to confess at 50(σεαυτ⋯ν σ⋯σον), 54 (εἰ οὖν τινι ὑμ⋯ν … ⋯γώ δ⋯ σωθε⋯ην … where Andokides' reasoning is illogical if he was not actually released from prison on the strength of his confession), 57(…τ⋯ ζ⋯ν περ⋯ πλε⋯ονος ποιησ⋯μενοι το⋯ καλ⋯ς ⋯ποθανεῖν), 59 (αὐτ⋯ς τε ⋯σῳζ⋯μην). The fact that at 66ff. Andokides does not state expressly that he was released from prison (as opposed to his relatives and other innocents) is easily explained by the fact that Andokides did not want it to appear that he had confessed merely to obtain his own release from prison. He wanted it to appear that his confession had had high moral purpose, namely to save his relatives and to release Athens from inner turmoil. In my opinion, this whole section of Andokides' defence speech would lose a lot of its point if the audience knew that he had not been released after confessing. We note at 54 that Andokides' enemies accused him of cowardice in accusing others in order to save his own skin. This accusation would have had no thrust if he had failed to save himself by confessing.

⁴ Meritt, ‘Departure of Alcibiades’ (art. cit. [n. 2] above), selects 21 June as the exact date. Dover favours early June; Weill, N., ‘Adoniazousai’, BCH 90 (1966), 675–98Google Scholar(esp 690) favours a date in July, on the strength of her proposed dating of the Adonia festival.

⁵ See e.g. Marr, J. L., ‘Andocides' Part in the Mysteries and Hermai Affairs 415 B.C.’, CQ 65 (1971), 326–38CrossRefGoogle Scholar .

⁶ MacDowell, op. cit. (n. 1), pp. 177–80, argues that Andokides' imprisonment lasted a lot longer than both Andokides himself and Thucydides would lead us to think. Mainly on the strength of the present passage(⋯γγὺς ⋯νιαυτ⋯ν) he constructs an elaborate hypothesis whereby Andokides confessed first to a part in the Herms affair but was not released then (in summer 415) because he failed to produce his slave to confirm his story. According to MacDowell he was kept on in prison for a considerable length of time until he confessed a second time, this time to a part in the Mysteries profanation. It should be emphasized that there is no ancient evidence for this hypothesis. It is true that Plutarch, (Life of Alkibiades 21)Google Scholar names the man who befriended Andokides in prison and persuaded him to confess as Timaios, whereas Andokides himself says that his cousin Charmides played this role in a single night (1.48), but how much can one responsibly build on the strength of this discrepancy? Marr (art. cit. (n. 5), 331) calls this part of MacDowell's reasoning ‘a good example of the perils of source reconciliation at all costs’. In fact MacDowell himself suspects the reading ⋯γγὺς ⋯νιαυτ⋯ν of being corrupt. Stays in prison in antiquity tended not to be long. Prison was, after all, a place for holding prisoners prior to trial, not a place for punishment. One might point to a second known term of imprisonment in Andokides’ career (in 411, when he was imprisoned by the oligarchs: see Andokides 2, On his Own Return, 13–16) as a possible source of the corruption in the present passage, and perhaps also of the discrepancy between Charmides and Timaios as the men named as Andokides' confidant in prison (the latter term of imprisonment was of some duration, as Andokides himself admits). See Marr, above n. 5, for more criticism of MacDowell on this point.

⁷ op. cit. (n. 1 above), p. 179.

⁸ Reading ⋯ζ ⋯πιβουλ⋯ς and construing with MacDowell, op. cit. (n. 1), p. 177. MacDowell rejects Taylor's proposed alteration to ⋯ζ ⋯πιβουλ⋯ς. There is a problem with MacDowell's reading, however, in that ⋯ζ ⋯πιβουλ⋯ς, which he takes with ἔδησεν ⋯αυτ⋯ν, is unnaturally distant from it. One might take ⋯ζ ⋯πιβουλ⋯ς with εἰσαϰθε⋯ς, but that gives a strained sense to the passive participle: ‘having been (= had himself?) deliberately led before the court’. Lamb, W. R. M., Lysias (Loeb Edition, Cambridge, Mass., 1967), p. 126Google Scholar, reads ⋯ζ ⋯πιβουλ⋯ς with Taylor, rendering (obscurely) ‘by a summary citation’ (but for what?).

⁹ ⋯γγυ⋯ω and τι⋯ω are not, however, synonymous terms, ⋯γγυ⋯ω refers to a pledge given that some juridical point will be met in the future (e.g. that someone will appear in court), whereas τι⋯ω is properly used of a penalty proposed after conviction. It seems to me certain that the context of [Lysias] 6.21–3, is before Andokides' imprisonment and subsequent confession, as it was this incident which led to both (23: ⋯κ το⋯του το⋯ τιμ⋯ματος ⋯δ⋯δετο … κα⋯ ⋯μ⋯νυσε δεδεμ⋯νος …). Thus, strictly speaking, the speaker is using τιμησ⋯μενος (21), τιμ⋯σασθαι (22), τιμ⋯ματος (23) metaphorically. According to his own narrative, Andokides is not yet being punished for anything. In going to prison he is merely paying the penalty for not fulfilling the terms of his pledge. Perhaps the speaker is using the strong word Tip.au) deliberately, in order to heighten the sense of Andokides’ guilt at this early stage of the proceedings. He speaks of him as if he were already a condemned criminal.

¹⁰ For what it is worth, I think MacDowell, op. cit. (n. 1), p. 178, is probably right that Andokides and [Lysias] must be referring to the same occasion in the matter of the promised slave. However, I do not believe that this entitles MacDowell to conclude that Andokides was re-imprisoned after confessing to the Boule because he failed to hand over his slave. MacDowell has not really come to terms with the fact that the speaker's context in [Lysias] 6.21–3, is before Andokides' imprisonment, not after (see previous note). If that is so, then it is Andokides who has distorted the facts when he relates the slave episode in connection with his subsequent confession. He had a good motive for so doing, too: he wanted it to appear that the Council had, in 415, accepted the account of his personal involvement in the Herms' mutilation which he gave at his trial in 400. In fact his confession in 415 probably involved admitting a greater degree of guilt than he was prepared to concede in 400. Shifting the slave episode to the main confession following imprisonment, and simply glossing over the fact that the unfortunate slave had in fact been removed from the scene before he could give evidence, served to lend credence to his story in 400.

¹¹ I would like to express my gratitude to S. C. Todd for many helpful suggestions in connection with this piece.