This paper owes much to the comments of Professor A. Andrewes, Mr W. G. G. Forrest and Mr G. E. M. de Ste Croix on earlier drafts. I am grateful to them all, but especially to Mr de Ste Croix for his generous reception of this attack on his views.

¹ de Ste Croix, G. E. M., Historia v (1956) 1–23Google Scholar (cited here as Ste Croix).

² Sealey, B. R. I., Essays in Greek Politics (New York: Manyland, 1967) 111–32Google Scholar (cited here as Sealey).

³ I concentrate here on constitutional matters. I do not, of course, imagine that other aspects of the revolutions are unimportant.

⁴ For example: those responsible for oligarchic propaganda after Pisander's first visit to Athens really intended to seize power for themselves (viii 66.1); Phrynichus was activated by fear of Alcibiades and belief that an oligarchic government would not allow Alcibiades to return to Athens (viii 68.3); when divisions appeared among the Four Hundred, Theramenes and his supporters claimed that because they were afraid of Alcibiades and the Athenians at Samos, and because harm might result if they tried to make peace with Sparta on their own, they wanted the constitution to be based on the Five Thousand, but their real motive was personal jealousy (viii 89.2–4); when discontent with the Four Hundred broke out openly, men called for the transfer of power to the Five Thousand, but they did this only for safety's sake and really wanted a restoration of democracy (viii 92.11—cf. below, 120).

Sealey, 129–30, claims that the personal motive ascribed to Theramenes in viii 89 is supported by Lys. xii. Erat. 66 but conflicts with the opposition in principle to extremist policies voiced by Theramenes in a speech in Xen. Hell. ii. 3.48 (cf. also Ath 28.5). (In fact, viii 89 ascribes the motive to a set of men headed by Theramenes and Aristocrates; Lysias in representing Theramenes as an opportunist contrasts him with Aristocrates.) Sealey supposes that we must choose between these views and that Thucydides' view is correct: I believe that there is room for both—though of course considerations of one kind may have counted for more with Theramenes, and perhaps much more, than the other.

⁵ Although much that the Ath gives us on the resolutions reads as documentary material, some of the documents appear to be excerpted rather than quoted in full, and the excerpting may not be the work of our author. The speech made by Antiphon when brought to trial under the intermediate régime was known to Thucydides (viii 68.2), and doubtless supplied material to fourth-century writers; on the appointment of the συγγραφεῖζ Ath 29.2, conflicting with Th. viii 67.1, gives a version which had previously been given by Androtion, 324 F 43 (cf. also schol. Ar. Lys. 421), and Androtion's father may be the Andron who was a member of the Four Hundred but survived to attack Antiphon under the intermediate régime (decree ap. [Plut.] X. Or. 833 e, Harp. Ἄνδρων). In particular, the unparalleled piece of information in Ath 29.1, that Pythodorus was the formal author of a motion but the man who spoke to the motion was Melobius, most probably points to the existence of an earlier narrative used by the author of the Ath.

⁶ Ath 29.1–3.

⁷ Ath 29.4–30.1.

⁸ Ath 32.1.

⁹ Ath 30–32.1.

¹⁰ Th. viii 48.3, 53–54.1, 66, 72, 86.3.

¹¹ I have few quarrels with the earlier account of Busolt, G. (Griechische Geschichte iii 2 [Gotha: Perthes, 1904] 1456 sqq.)Google Scholar; but the twentieth century has felt obliged to devise new answers without new evidence, and its major departures from that account, including those which he later accepted (Griechische Staatskunde i [Munich: Beck, 1920] 69–78) seem to me to be unjustified.

¹² In this I disagree with Miss M. L. Lang, who believes that Thucydides and the Ath each report different stages, which the other omits, in the whole course of events: AJP lxix (1948) 272–89, cf. Cary, M., JHS lxxii (1952) 56–61Google Scholar, Lang, M. L., AJP lxxxviii (1967) 176–87Google Scholar.

¹³ viii 67.2–68. 1 init., 69. 1 init.

¹⁴ Thucydides, having mentioned them in his earlier account of the oligarchs' propaganda (viii 65.3), and knowing that they in fact played no part in the régime of the Four Hundred, is virtually silent on the Five Thousand; the Ath, having a later account of them in the ‘immediate’ constitution (ch. 31), is silent on the Four Hundred.

¹⁵ Ath 29.5 fin. We meet one of them in Lysias, xx. Pro Polystrato.

¹⁶ Ath 30.1, 32.1 init.: they are said to be appointed by the Five Thousand (cf. below, n. 21).

¹⁷ viii 67.3.

¹⁸ Th. viii 92.11, 93.2, cf. 89.2 (all quoted below, p. 110). The implication of Lys. xxx. Nic. 8, that the list was published, is not necessarily to be trusted; on the evidence of the Ath, see below, n. 21.

¹⁹ Ath 30–1.

²⁰ The last sentence of Ath 31 (εἰζ δὲ τὸν … οἱἑκατὸν ἄνδρεζ) seems to belong not to the ‘immediate’ constitution of ch. 31 but to the ‘future’ constitution of ch. 30 (cf. Cary, M., JHS lxxii [1952] 57)CrossRefGoogle Scholar. Perhaps the two constitutions were issued together as we have them, and this sentence is an amendment proposed by a man who seriously wanted the καταλογεῖζ, there and then, to begin to prepare for the introduction of the ‘future’ constitution. (But in the other references to the future in ch. 31 I would see an indication that other oligarchs intended to maintain the ‘immediate’ constitution for some time.)

²¹ Ath 32.1: apparently at an assembly of the Five Thousand, by whom the ἀναγραφεῖζ had been appointed. According to Thucydides the list of the Five Thousand was never published (cf. above, n. 18); but this is not necessarily confirmed by the wording of Ath 32.3, 33.2. I favour the view that the resolutions at Colonus were carried in several stages and that, having decided that there was to be an assembly of Five Thousand, those present deemed themselves to be the Five Thousand for the purpose of taking such further decisions as were needed to bring the new constitution into effect; with this ‘evidence’ of their activity, the author could not deny that the Five Thousand had existed.

²² viii 69–70.1.

²³ Th. viii 97.2, Ath 33 (extracts quoted below, p. 122).

²⁴ The decree of Demophantus ap. And. i. Myst. 96 emphasises that the boule of the restored democracy is ἡ βουλὴ οἱ πεντακόσιοι οἱ λαχόντεζ τῷ κυάμῳ: appointment of another Four Hundred is unlikely, and it will be enough if one of these conditions was not satisfied (cf. Hignett, C., History of the Athenian Constitution [O.U.P., 1952] 372Google Scholar; more cautiously, Ste Croix, 22 with n. 98). The fact that Alcibiades, before the fall of the Four Hundred, recommended that the boule of five hundred be restored (Th. viii 86.6, quoted below, p. 119) does not, of course, prove that the boule of the intermediate régime numbered five hundred.

²⁵ Ath 30.2 provides for an enlarged board of Hellenotamiae to handle both Athenian and Delian League funds, a change which took place in or shortly before 410 (see my Athenian Boule [O.U.P., 1972] 99 n. 4), and for the amalgamation of the two major boards of sacred treasurers, a change which took place probably in 406 (Ferguson, W. S., Treasurers of Athena [Harvard U.P., 1932] 4–7, 104–9CrossRefGoogle Scholar, Thompson, W. E., Hesp. xxxix [1970] 61–3)Google Scholar.

²⁶ Ap. And. i. Myst. 96–8. On the date of the restoration see Meritt, B. D., Athenian Financial Documents of the Fifth Century (U. of Michigan P., 1932) 105–7CrossRefGoogle Scholar; I am not convinced by the arguments of Hatzfeld, J., REA xl (1938) 113–24CrossRefGoogle Scholar, for an earlier date.

²⁷ Sealey, 125.

²⁸ Ste Croix, 9.

²⁹ viii 65.3.

³⁰ viii 97.1.

³¹ viii 67.3. ξυλλέγειν must surely mean, to convene a ξύλλογοζ, or meeting (cf. Th. ii 22.1 [ξύλλογοζ], Xen. Hell. iii 3.8 [συλλέγειν]).

³² Ath 29.5.

³³ Ath 30.1, 32.1 init.

³⁴ Ath 29.5.

³⁵ viii 65.3. The Ath's minimum of 5,000 suits the propagandist tendency of the official resolutions (and compare the claim that Polystratus enrolled 9,000: Lys. xx. Poly. 13); Thucydides' maximum of 5,000 suits the earlier stages of oligarchic propaganda.

³⁶ Cf. also Lys. xx. Poly. 13, 16, with Ste Croix, 8–9.

³⁷ viii 53.3.

³⁸ Notice particularly viii 53.1: μὴ τον αὐτὸντρόπον δημοκρατουμένοιζ. Pisander held out the bait that if it was not liked the new constitution could aways be changed, and insisted (53.3 cf. 54.1).

³⁹ viii 72.1.

⁴⁰ viii 86.3.

⁴¹ Dishonest, because it implies that those who used not to attend were the thetes who would now be forbidden to attend.

⁴² After μεθέξουσιν I understand τῶν πραγμάτων or τῶν ἀρχῶν (but some read τῶν πεντακισχιλίων as the object of μεθέξουσιν: e.g. Caspari, M. O. B., JHS vxxxiii [1913] 9)Google Scholar. I believe that rotation in the membership of the boule, after the manner of the Boeotian cities (Hell. Oxy. 16.2), was envisaged in the ‘future’ constitution, issued before the envoys set out from Athens. Meanwhile absolute power was retained by the Four Hundred, and on either view of the scheme approved at Colonus it would be appropriate for the envoys to give an assurance that these men would not retain their monopoly of power indefinitely.

⁴³ viii 86.6.

⁴⁴ viii 89.2.

⁴⁵ Ath 33.2.

⁴⁶ On Thucydides' detection of ‘real’ aims beneath the surface see above, 115–16.

⁴⁷ viii 92.11.

⁴⁸ viii 93. (I quote §§2 and 3. In 3 the reading ἐν Διονύσου is due to Dr D. M. Lewis, and will be accepted by Andrewes in vol. v of Gomme's Commentary.)

⁴⁹ viii 93.1, 3.

⁵⁰ Most MSS. of viii 94.1 have the phrase πᾶζ τιζ τῶν πολλῶν ὁπλιτῶν; but πολλῶν is omitted by C and ὁπλιτῶν was omitted at first by B, and editors have followed Stahl in deleting τῶν πολλῶν ὁπλιτῶν as a combination of two glosses.

⁵¹ Sealey, 125. He passes too easily from this observation to acceptance of Ste Croix' thesis.

⁵² Ste Croix, 9. Cf. below, 123–4.

⁵³ On relativity in the use of such terms as ‘democracy’ cf. below, 122–3, 125.

⁵⁴ Sealey, 112, 122–7.

⁵⁵ It is in fact not certain whether thetes were able, in law or in practice, to hold office in the democracy of the late fifth century: see below, 126–7.

⁵⁶ Th. viii 97.1–2.

⁵⁷ Ath 33.1–2. In common with Ste Croix (10 n. 45 with Appendix, 22–3; 4 n. 17). I shall base no arguments on Ath 34.1 init. or 41.2.

⁵⁸ P. 120.

⁵⁹ Ste Croix, 6–8.

⁶⁰ [Xen.] Ath. Pol. i 2 sqq.

⁶¹ viii 92.11, quoted above, 120.

⁶² Jones, A. H. M., Athenian Democracy (Blackwell, 1957) 7–1, 79–83. 166–9Google Scholar.

⁶³ The noun is not used elsewhere by Thucydides, nor by Herodotus, Xenophon or Aristophanes. The verb συγκεράννυμι is used by Herodotus at iv 152.5, vii 151, and perhaps ix 37.4; by Thucydides at vi 18.6; by Xenophon at Cyr. i 4.1 and Cyn. iii 1; by Aristophanes, at Plut. 853Google Scholar. The range of uses given to the verb by these writers suggests that it may be unwise to insist that the noun must here mean ‘mixture’ in the most literal sense.

It is not universally accepted that ξύγκρασιζ here refers to the actual constitution at all: see, e.g., Donini, G., La posizione di Tucidide verso il governo dei Cinquemila (Turin: Paravia, 1969) 8–12, 94–5Google Scholar. But I reply to Ste Croix on the assumption that ξύγκρασιζ does refer to the constitution.

⁶⁴ Ste Croix, 7 with n. 31. This classification would no doubt have been resented by many democratically-minded hoplites.

⁶⁵ Elsewhere Thucydides distinguishes [oligarchic] and ‘democratic’ factions as ὀλίγοι and δῆμοζ (e.g. iii 27.2 [Mytilene], iii 72.2 etc. [Corcyra], v 82.2 [Argos]). It is usually impossible to discover where the line should be drawn, but in places where the ὀλίγοι were a small and exclusive clique δῆμοζ would presumably be applied to all their opponents, including men of some substance. In Syracuse, where Athenagoras defended democracy against Hermocrates, the defence of democracy put into his mouth seems less than extreme (vi 39.1: .

⁶⁶ Ste Croix, 9–10.

⁶⁷ Pp. 120–21.

⁶⁸ Sb. Berlin 1935, 52–3. Cf. Ferguson, W. S., believing that at the ἄλλαι ὕστερον πυκναὶ ἐκκλησίαι ‘the body assembled was obviously οἱ τὰ ὅπλα παρεχόμενοι’, with the implication that the assembly which deposed the Four Hundred was an open one (Mélanges Glotz [Paris: P.U.F., 1932] i 364–5Google Scholar, cf. CAH v [1927] 338). We cannot, of course, tell how efficiently restrictions on the membership of the assembly were enforced, on this occasion or on any other.

⁶⁹ Ste Croix, 9 n. 39.

⁷⁰ Cf. Andrewes, A., JHS lxxiii (1953) 2–9CrossRefGoogle Scholar.

⁷¹ Ste Croix, 10. On the πάτριοζ πολιτεία as an oligarchic ideal see Fuks, A., The Ancestral Constitution (Routledge, 1953)Google Scholarpassim.

⁷² Ath 29.3.

⁷³ Cf. Meiggs & Lewis 58A, 11 (ζητεῖν). But Andrewes tells me he takes προσαναζητῆσαι here to mean not ‘search for’ but ‘study’ or ‘investigate’, as at Th. viii 33.4.

⁷⁴ It may be true, as Fuks believes (The Ancestral Constitution, pp. v, 107), that the phrase πάτριοζπολιτεία came particularly to be associated with the more moderate oligarchs; but the phrase could of course be used as a slogan with great propaganda value by oligarchs of various shades wishing to make their views seem respectable.

⁷⁵ Notice the attitudes expressed in Iliad ii 188 sqq., and on ἰσηγορία in Athens see Griffith, G. T., Ancient Society and Institutions: Studies presented to Victor Ehrenberg (Blackwell, 1966) 115–38Google Scholar, Woodhead, A. G., Historia xvi (1967) 129–40Google Scholar.

⁷⁶ Th. ii 65.10, 70.4; and compare the accounts of debates such as that on Mytilene, iii 36.6–49.1, and that on the expedition to Sicily, vi 8.2–24.2.

⁷⁷ Ar. Ach. 633–42 (I take l. 642 to mean: showing what democracy in the allied cities—and by implication in Athens too—is really like); Eq. passim, e.g. 1111–20.

⁷⁸ [Xen.] Ath. Pol. i 6–9. I cannot believe that this pamphlet was written earlier than 431: see, most recently, Forrest, W. G. G., Klio lii (1970) 107–16Google Scholar.

⁷⁹ Xen. Hell, i 7, esp. 12. On this occasion the emotions of the assembly were aroused in favour of Theramenes and his friends, against the extreme democrats.

⁸⁰ Cf. above, 122 with n. 60.

⁸¹ For a more cynical view see Hdt. v 97.2.

⁸² Ste Croix, 10–11.

⁸³ Ste Croix tells me that he believes the assembly was much less powerful than under the democracy. See below, 127.

⁸⁴ Ste Croix thinks it important (2, 21–2) that on his view of the intermediate constitution the surrender of a measure of power by the thetes was ‘voluntary’ in that they were members of the assembly which decided on this surrender and, remaining members, could when they chose revoke it. I doubt whether this would matter very much in practice.

⁸⁵ Lys. xx. Poly. 22 cf. 14. Andrewes reminds me that since the leaders of the intermediate régime had themselves been implicated in the extreme oligarchy the formal charge is likely to have been a side issue.

⁸⁶ Ste Croix, 11–12, cf. Ferguson, , Mélanges Glotz, i 358–60Google Scholar.

⁸⁷ Ste Croix, 12.

⁸⁸ viii 98.1, D.S. xiii 38.1. Elsewhere Thucydides seems to describe the intermediate régime as δημοκρατία (viii 68.2; cf. Antiphon, fr. B1.2, Maidment [Loeb]); no sense can be made of Diodorus' description of the intermediate régime.

⁸⁹ Ath 33.1.

⁹⁰ Meritt, Athenian Financial Documents 105–7Google Scholar.

⁹¹ As Ferguson pointed out (Mélanges Glotz, i 364[–5] n. 1) the ending of Thucydides' history in 411 may be partly responsible for this silence.

⁹² Ste Croix, 22. But cf. above, n. 84.

⁹³ Cf. The Athenian Boule 134–5.

⁹⁴ Cf. above, n. 24.

⁹⁵ Cf. above, 118 with n. 26.

⁹⁶ Cf. above, 125.

⁹⁷ See, conveniently, Andrewes, , JHS lxxiii (1953) 2–3CrossRefGoogle Scholar.

⁹⁸ By 458 they were appointed from the first two classes, and no change is attested before then. In Dem. Phal. 228 F 43 ap. Plut. Arist. 1.2 it is claimed that archons were appointed from the first class only: this can hardly be correct for 489/8, but might perhaps be correct for the time of Solon. (See Hignett, , History of the Athenian Constitution, 101–2Google Scholar, believing that there was no change between Solon's law and 458.)

⁹⁹ Ath 7.3–8.1.

¹⁰⁰ Ath 26.2.

¹⁰¹ Ath 8.1.

¹⁰² Ath 7.4 fin.

¹⁰³ Cf. Ste Croix, 1: he lists what evidence there is in n. 5.

¹⁰⁴ The language of Ath 7.4 suggests that in the author's day the law may have been evaded. Though we are often able to assign a man to a higher class, the evidence never allows us to assign an Athenian citizen reliably to the lowest class, and we have no way of discovering whether the law was in fact evaded, either in the late fifth century or (after appreciable monetary inflation) in the 320's. There was at any rate a tendency for offices to go to the rich—in the fourth century men from trierarchic families occupied more than their fair share of seats in the boule (The Athenian Boule, 4–6)—but we cannot say whether any thetes did in fact hold any offices in the late fifth century.

¹⁰⁵ See above, 123.

¹⁰⁶ The assembly of the intermediate régime may have been slightly less powerful than that of the democracy (I think it is possible that the Thesmophoriazusae should be dated to 410 and contains indications that the boule's powers were somewhat enhanced: I hope to consider this play elsewhere; meanwhile, see The Athenian Boule, 185–6, 190)Google Scholar. But a major surrender of power by the assembly is surely impossible in this situation.