¹ Ed. Dièls, H. and Schubart, W. (Berliner Klassikertexte ii, 1905)Google Scholar.

² The Middle Platonists (London, 1977)Google Scholar, hereafter cited as MP.

³ Antiochus and the late Academy (Hypomnemata lvi, Göttingen, 1978), hereafter cited as Ant., pp. 39, 219, 304Google Scholar.

⁴ But see intro. xxiv: ‘Der vorliegende Kommentar zum Theätet muss, wie die Schrift lehrt, vordem dritten nachchristlichen Jahrhundert entstanden sein. Auch fehlt darin jeder Hauch des Neuplatonismus. Dagegen ist er ein Muster jener eklektischen Philosophie, wie sie in dem ersten Jahrhundert v. Chr. und den beiden ersten n. Chr. herrschte’. Cf. also Schubart, in Papyri Berolinenses (Bonn, 1911), p. 31Google Scholar, where the papyrus is marked ‘Saec. II A.D.’.

⁵ Glucker, , Ant. pp. 39, 219, 220Google Scholar.

⁶ See Diels/Schubart, intro. xiv, for evidence of the earliness of the papyrus; cf. Dillon, , MP, p. 270Google Scholar: ‘The papyrus itself belongs to the first half of the second century’.

⁷ An. Morb. 31. 23–32Google Scholar. 5 (Marquardt).

⁸ Libr. Prop. 97. 9–11 (Müller)Google Scholar.

⁹ In 1906 appeared the work which pioneered the Gaius-School theory, Sinko, T., De Apulei et Albini doctrinae Platonicae Adumbratione (Krakow)Google Scholar.

¹⁰ Following Freudenthal, J., Der Platoniker Albinos und der falsche Alcinoos (Berlin, 1875)Google Scholar.

¹¹ For the connections see Diels, , intro. xxx–xxxiiiGoogle Scholar, and Dillon, MP, p. 271Google Scholar.

¹² MP, pp. 306–38.

¹³ ‘Parisinus Graecus 1962 and the writings of Albinus’, Phoenix 28 (1974), 320 ffCrossRefGoogle Scholar. and 450 ff.

¹⁴ Platonis Opera, ed. Hermann, C. Fr. (Teubner), vi. 147–51Google Scholar. Page references to Prol., to the Didascalicus, and to the anon. Prolegomena ad Platonis Philosophiam from this volume will be ‘given with an H.

¹⁵ 7. 14–20; cf. Didasc. 28 (p. 181. 16–182. 7 H) and Apul. De Plat. 23 (p. 126. 4 ff. Thomas) in addition to Albinus, Prol. 5 (p. 150. 10 H), 6 (p. 151.4 H)Google Scholar. The doctrine is early enough to appear in Arius-Eudorus (, Stob.Ecl. 2, p. 49. 8 Wachsmuth)Google Scholar, ordinary enough to appear at , D.L.i 3. 78Google Scholar.

¹⁶ 46. 43, 47. 44 ff; cf. 23. 7, 47. 20; cf. Prol. 6 (p. 150. 21–3 H).

¹⁷ Before writing K, A had written on Tim. (35. 10–12) and Symp. (70.10–12), and he promises a work on Phd. at 48. 7–11; the work is probably not yet written, though it may be that the I student who reads Tht. is not yet expected to have read Phd.

¹⁸ Prol. 3 (p. 148 H). While there is a possibility that omissions have been caused by problems £ of transmission (see Schissel, O., Hermes 62 (1931), 215–26)Google Scholar, Albinus may have deliberately passed over Phdr., for he adhered to the view that the rational soul alone is immortal (Proclus, , In Tim. 3.234. 9–18 Diehl)Google Scholar, and it is very difficult to reconcile this view with the myth of Phdr. (245 c ff). One doubts whether he saw the work as being spurious, for he quoted 237 be at p. 147. 8 (though without naming the work); but it is possible that he regarded both Phdr. and Tht. as educationally unsound. Neither work would seem to have been prominent in the programme of education which ‘he sketches at sections 5–6.

¹⁹ Proclus, , In Tim. 1. 340Google Scholar. 23 ff. (Diehl).

²⁰ His treatment of the Academics at 6. 29–7. 14, 54. 43–55. 7, and 70. 12–26 shows that he considered Academic ‘scepticism’ as a device for attacking Stoics, etc., which was used in a manner consistent with common sense and ⋯ν⋯ργεια (70. 21–6). It did not lead to ⋯ποϰ⋯, nor preclude their having δ⋯γματα (in a weak sense, 54. 43–55. 7). Moreover, A shows genuine interest in Pyrrhonism (61. 10–63. 40). For A's use of Philo of Larissa's alleged criterion of ⋯ν⋯ργεια see 5. 34–6 (cf. Numenius fr. 8 Leemans = 28 des Places, Aenes. in Phot. cod. 212, 170a 37–8, Cic. Ac. 2.34, K 70. 21–6, Glucker, Ant. 72–4Google Scholar, and the author's ‘Philo of Larissa on agreement and the self-evident‘, Dionysus 5 (1981), 66 ff.)Google Scholar. A's brand of ‘scepticism’ is perhaps similar to that which he fathers on Plato, e.g. 58. 36–9: οὐ γ⋯ρ ⋯φεῖτα⋯ μοι οὔτε ψε⋯δος συγϰωρ⋯σαι οὔτε ⋯ληθ⋯ς ⋯φαν⋯σαι, or 59.12–21, where Plato is said to avoid open demonstration of his views, but to reveal his pleasure imperceptibly to the familiar reader.

²¹ This was originally held to make A's identification with Albinus difficult (Diels/Schubart, , intro. xxxiii)Google Scholar, but Dillon, rightly does not think the point conclusive (MP, p. 271)Google Scholar. The phrase appears wrongly in K3. 2–3 and 15. 21–3, correctly at Prol. 6 (150. 27 H). If the Meno, and this, passage especially, were not crucial to A's epistemology, one might dismiss the mistake as a slip of memory, but it is more likely that some deliberate ‘emendation’ is responsible. Another such I alteration to the end of Meno (99e6) occurs at Clem. Strom. 5. 83. 4 (Stählin), cf. 84. 1–2. Antiochus’ theory that νο⋯ς is the cause of all cognitive activity (Cic. Ac. 2. 30) may be behind J, both emendations, νο⋯ς being both the αἰτ⋯α λισμο⋯ which converts right opinion to knowledge and the initial divine impulse towards right opinion and virtue. Though A seems to interpret, Meno 98 ab correctly at col. 3, he may be using a text transmitted via Antiochus' school.

²² A n An. Morb. 92 clearly views the Academics as a different kind of philosopher from the fj Platonics, closer to the Sceptics (ibid. 60), as also in Opt. Doctr. (aimed at Favorinus).

²³ He took the work's theories on the human body and its ailments seriously enough for both a commentary on the medical passages and a Plato-based attack on some self-styled Platonists, (Scr. Min. ii. p. 64Google Scholar. 19 ff. Müller). But where it suits him Galen will remind one that Plato did not think everything in Tim. to be certain, Plac. 9. 794–5.

²⁴ Mor. 999c ff, on which see below p. 166.

²⁵ See Glucker, Ant. p. 209, who points out that only the spurious De Mus. (1131 f.) refers to ‘Platonics’, who are in any case the immediate followers of Plato, not a later sect.

²⁶ On Taurus see Dillon, , MP, pp. 237–47,Google ScholarDorrie, H., ‘L. Kalbenos Tauros’, Kairos 1–2 (1973), 24–35Google Scholar. We are given greater insight into Taurus as a person than as a philosopher in Gellius; for the commentary on Gor. see N.A. 7. 14. 5.

²⁷ Compare Dillon's, comments on its remains at MP, pp. 246–7Google Scholar ‘fairly detailed exegesis’ with those on K at 270: ‘maintains a level of stupefying banality’.

²⁸ Dillon, , MP, pp. 240–2Google Scholar, makes much use of Gellius N. A. 12. 5. 7, where we see a rather Antiochian emphasis on οἰκείωσις σις and τ⋯ πρ⋯τα κατ⋯ φύσιν. But note that K rejects the former as a suitable basis for justice, cols. 6–7.

²⁹ For some fragments see Baudry's, J.collection (Paris, 1931)Google Scholar; from passages in Proclus, In Tim., we learn of his exegesis of Tim. and Phdr. (the latter at 3. 277. 15 Diehl).

³⁰ But there is no evidence for his presumed Athenian connections, and he has ‘a rather curious name to be borne by an Athenian’ (Dillon, , MP, p. 248)Google Scholar. I assume that ὑπομνήματα were not designed for wide circulation, judging both from the word itself and from their uselessness to persons not possessing a text. Their obvious place was within the philosophical school, until such time as fame demanded their wider availability.

³¹ Extant remains of Middle Platonists have been partly shaped by the interests of Neoplatonists and Christians who preserve them. But Atticus' division of philosophy shows clearly that ethics and theology-directed metaphysics/physics were his main concerns, while logic was merely a tool (Eus, . P.E. 509 b = fr. 1 Baudry)Google Scholar.

³² The anti-Aristotelian elements are patent, as is the religious motivation (cf. Dillon, , MP, p. 253:Google Scholar ‘the general tone is one of furor theologicus’). For his linguistic endeavours (Proclus, In Tim. 1. 284. 13Google Scholar) and other features of his commentaries see Dillon, , MP, p. 257Google Scholar.

³³ See Dillon, J., ‘Harpocration's Commentary on Plato: fragments of a Middle Platonic Commentary”, Calif. Studs, in Class. Ant. 4 (1971), 125 ffGoogle Scholar.

³⁴ Proclus, In Tim. 1. 255. 5 ff. (Diehl)Google Scholar.

³⁵ ibid. 1. 227. 13 ff. (Diehl).

³⁶ I know of no evidence that could assist us to fix a date.

³⁷ He wrote a polemical work against the New Academy, partially preserved by Eusebius, (P.E. 14. 727Google Scholar a ff. = frr. 1–8 Lang = 24–8 des Places).

³⁸ As can be seen from his whole approach to τ⋯ ⋯γαθ⋯ν, asking first τ⋯ τ⋯ ⋯ν, then fiercely dismissing the claims of body and matter (frr. 11–13 Lang = 2–4a des Places).

³⁹ Cronius seems to have followed Numenius closely; his scant remains follow those of Numenius in Lang's collection (Brussels, 1937).

⁴⁰ Simpl. In Categ. saepe. On him see Praechter, K. (Hermes 57 (1922), 481 ff.)Google Scholar and MP, pp. 233–6.

⁴¹ Tim. dominates at 154. 22–9 H. Tht. 184–6 may underlie 154. 29–32, but the doctrine is too ordinary to affirm as much.Phil. 38b ff. is the inspiration for 154. 35 ff. and 155. 12–15 and Soph. 263e for 155. 15–17. However, there are few signs of first-hand use of the dialogues in the composition of this work.

⁴² A's concept of the criterion δἰ οὑ is not evident in col. 2, but it seems likely that it is to be identified with ᾧ κρινο⋯μεν τ⋯ πρ⋯γματ at 2. 26–8. When this latter is accurate the lasting παραδϰ⋯ of things well judged becomes knowledge. It is natural to assume that A demanded accuracy of both senses and intellect; Tht. 184–6 would allow him to think of the latter as a kind of criterion δἰ οὑ apprehending things δἰ αὑτ⋯ς (185el). Tht. 184d4 (ῄ δι⋯ το⋯των … αἰσθα⋯μεθα however, would sanction the view that we judge though the senses but with (dat.) the mind, and, as an Academic, A is much more likely to attribute accuracy to mind than to senses (cf. Diels, , Box. 396 b 17–19)Google Scholar. Thus A's criterion through which is either both senses and mind or only senses; but in Didasc. 4 (p. 154. 12–18H) it is only λ⋯γος φυσικ⋯ς.

⁴³ We meet μν⋯μη at 4 (155.29 H), betraying the influence of Phdr. 249c–254b, where it occurs 6 times. The more usual term ⋯ν⋯μνησις occurs in the context of arguments for immortality at 25 (177. 37 H).

⁴⁴ 3. 1–7: we know P when we know δι⋯ τ⋯ P.

⁴⁵ Stoic theory of common/natural notions is assimilated to Platonic recollection theory as early as Cic. T.D. 1. 57, partly becuase Plato uses ἔννοια and ⋯ννοεῖν in the relevant passage of Phd. (73c8, c9, 74a6, b6, c8, dl, d9, e2, 75a 1, a6x2, all, 76a3). Now if recollected Ideas = simple knowledge, while simple knowledge is the subject of Tht., it is odd that discussion of recollection is deferred until A's work on Phd. (48. 7–11).

⁴⁶ For its omission from Albinus’ classification and programme of study see above, p. 162 and n. 18. D.L. 3. 62 speaks of teachers who had begun with Tht., but this could have been well before the time of D.L. (or his source). Tht. actually comes seventh in a list of eight works with which teachers of Plato had begun, so that teachers who used it first were either not recent or not important. Clearly semi-sceptics are more likely to have used it than dogmatists.

⁴⁷ Antiochus, at Ac. 2. 29Google Scholar; but others of his time surely agreed.

⁴⁸ Socratic ‘midwifery’ is discussed by Plut.Mor. 999c ff. and Max. Tyr.Or. x. 4a–h; see also Alex., PhiloHer. 247Google Scholar, Mut. 144, and the ‘maieutic’ type of dialogue at Albinus, Prol. 3Google Scholar (148. 36 H), 6 (150. 34 H). Lines 176a–d are well known from the time of Eudorus (, Stob.Eel. 2. 49Google Scholar. 8 Wachsmuth) and Alex., PhiloFug. 63, 82Google Scholar, etc. (see n. 15). The wax-tablet section is known to Alex., Philo(Deus 43Google Scholar, Her. 181, Mut. 212), and influences Plut. frr. 215 h, 217 j (Sandbach, Loeb), Didasc. 4 (154. 35–155. 12 H). Galen, Plac. 1. 631Google Scholar is influenced by 184 b–186 e.

⁴⁹ See above, n. 17.

⁵⁰ For these terms see Glucker, Ant. pp. 206–25Google Scholar, with my review (Prudentia 12 (1980), 109–18)Google Scholar.

⁵¹ See above, n. 22.

⁵² See above, n. 20.

⁵³ This follows from (a) his sympathies (see n. 20), and (ft) his One-Academy theory (55. 2–7): in supposing himself to follow Plato he must also suppose that he follows later Academics.

⁵⁴ See Glucker, , Ant. p. 292Google Scholar; on p. 293 he notes that Galen refers to no Academic sect late in life, giving the impression that it had died out.

⁵⁵ One recalls the mild Academic attitude which Plut. adopts in Comm. Not., where he tries to show that it is not the Academics who overturn κοινα⋯ ἔννοιαι, ⋯μολογ⋯α, and ⋯ν⋯ργεια (on the connection between these terms see part 2 of my article cited in n. 20). Note Plutarch's distrust of ‘private notions’ (1062a, cf. AT 46. 34). Both authors see ‘recollection’ as important, and both apparently discussed it in relation to Phd. (Plut. frr. 215–17 = , Olymp.In Phd. p. 155Google Scholar. 24 ff. Norvin, cf. K 48. 7–11). Plutarch's view of Socratic midwifery is close enought to K to attrct five references to K in Cherniss', H. notes to Q.P. i (Loeb, 1976)Google Scholar.

⁵⁸ K 58. 39–59. 2 uses Tht. 151c7–d3 to refute those who detect εἰρωνε⋯α in the midwifery passage. The dogmatist view of Tht. was threatened by phrases such as ἄγον⋯ς εἰμι σοφ⋯ας (150c4), which seemed to picture a sceptic ‘Socrates’. The dogmatist resource was to appeal to εἰρωνɛ⋯α, and K answers by showing that ‘Socrates’ makes no attempt to be modest in this passage, since he likens himself to a God (a) in respect of his εὔνοια, and (b) in respect of his not accepting falsity nor hiding truth. Since A saw likeness to God as the foundation of virtue (7.14–20), such a claim to likeness could show no modesty. Plutarch also sees use of God's name as a sign of arrogance rather than εἰρωνɛ⋯α, quoting 151c 5–d 3, the very passage used by AT: as if he remembered that he had seen the passage used to refute the modesty theory.

⁵⁷ Note 999c: οὐ γ⋯ρ εἰρωνευ⋯μεν⋯ς γε κα⋯ πα⋯ζων. Without κα⋯ πα⋯ζων the passage would argue, like K, that to call upon God's support for one's practice is hardly a sign of modesty. But with these words it emphasizes instead that one does not call upon a God unless one is serious.

⁵⁸ 48. 7–11, and see n. 17 above.

⁵⁰ Severus in Proc.In Tim. 1. 255. 5 ff. (Diehl); Didasc. 4 (154. 22–9 H) and 14 (169. 20–6 H); ‘Platonics’ in S.E. Math. 7. 141–4, cf. Cic. Ac. 1. 30–2.

⁶⁰ Discussed by Plut, . Mor. 1001cff.Google Scholar; Didasc. 7 (162. 7–20 H).

⁶¹ frr. 215–17 (Sandbach, Loeb) = , Olymp.In Phd. 155Google Scholar. 24 ff. and 212. 1–26 (Norvin). •2K 2. 18–21, 15. 2–23, 17. 25–32. Note here that my argument is unaffected by R. G. Hoerber's theory that the secondary titles of the dialogues antedate Thrasyllus. K seems unaware of the existence of any established tradition relating to a secondary title of Tht. In any case Hoerber can only produce evidence of the earliness of a few such titles, and even then they need not be official titles (see Hoerber, , ‘Thrasyllus' Platonic canon and the double titles’, Phronesis 2 (1958), 10–20)CrossRefGoogle Scholar. Indeed K 48. 10 refers to Phd. by its long-established secondary title.

⁶³ A's introduction is historical, and seems to deal chiefly with Protagoras. But to judge from Albinus' Prologos (and from D.L. 3. 49 ff.) Thrasyllus' activities had generated much interest in the arrangement and classification of dialogues. The anonymous author of P. Oxy. 3219 (ed. M. Haslam. XLV, pp. 29–39) shares interests with the source of D.L. 3. 52–6 (see Haslam's notes), ^ and would presumably have been interested in classification, while Didasc. 6 (p. 158.24 H) shows awareness of the classifications; anon. Prol. in Plat. Phil. 17 ff. shows interest. One suspects that A's phrase ⋯ν ταῖς ζητ⋯σεσιν (59. 7, 13) would have read ⋯ν τοῖς ζητητικοῖς διαλ⋯γοις if it had 1 post-dated Thrasyllus.

⁶⁴ Galen's criterion is ἴσθησ⋯ς τε κα⋯ ν⋯ησις ⋯ναργ⋯ς (plac 9. 778 et saepe).

⁶⁵ 61. 15–46. For Aenesidemus’ use of the δαιν μενον see D.L. 9. 78, 10; also S.E. Math. 8. 215–16. The Pyrrhonist principle at 61. 12 (οὐδ⋯ν ἄν τις καθοιστικ⋯ς δογματ⋯ζοι) occurs at D.L. 9. 71, and may be linked with Aenesidemus on the basis of D.L. 9. 106 (οὐδ⋯ν ⋯ρ⋯ζει δογματικ⋯ς) οὐδ⋯ν το⋯το δογματ⋯ζοντος (61. 38–9) utilizes the Aenesideman phrase οὐδ⋯ν το⋯το (see Janacek, K., Eirene 14 (1976), 99)Google Scholar. Though the appearance of the moment is said to be Pyrrho's criterion, Aenesidemus' interest in maintaining Pyrrhonism's difference from dogmatic philosophies is reflected in the claim that Pyrrho did not try to show that the δαιν μενον exists, i because of equal arguments to the contrary (cf. Aenes. at D.L. 9. 106: δι⋯ τ⋯ν ⋯ντιλογ⋯αν) and i his treatment of all presentations as equal (cf. Aenes. in Phot. cod. 212, 169 b 40 ff.). Here the, term for ‘equally strong’ arguments (ἰσοκρατ⋯ς) at 61. 26 suggests that Sextus' term (ἰσοσθεν⋯ς) was not yet standardized, while the verb ⋯ξομαλ⋯ζειν (61. 28) sounds authentic (cf. ⋯νωμαλ⋯α) I in Aenes., D.L. 9. 78) but is not Sextan. 63. 1 ff., on the claim that all things are relative, leads, to a discussion of συνθεωρο⋯μενα (63. 11–20) and κρῖνον as expected from Sextus' account of the eighth Aenesideman trope (PH. 1. 135–6). Material from tropes 1, 5 (S.E. = 7 D.L.), and, 6 is also present. Aenesidemus' sensation-reason dichotomy (cf. Phot. cod. 212, 169b20, S.E. Math. 8. 40) also appears.

⁶⁶ I think it is wrong to accept, with Burkhard, U., Die angebliche Heraklit-Nachfolge des: Skeptikers Aenesidem (Bonn, 1973)Google Scholar, the correction made by Natorp at PH. 1. 222 (κατ⋯ ≶τοὺ¨ς> περ⋯…) rather than that of Heintz (κατ⋯ ≶τοὺ¨ς> περ⋯…) PH. I 1. 222 for several reasons, (a) Sextus cites Timon as an authority on this point (223), and it would be remarkable that he should cite others, particularly in PH where much material is unoriginal ky but acknowledgement is only made in a striking case of debt (e.g. to Aenes. at 1. 180). Why I mention Menodotus at all (hapax in Sextus), let alone use the periphrasis ol irepl… ? Because v Sextus has a oraois in mind? But why use a arams as an authority rather than its single ^ authoritative leader? One corrects a oraais. (b) That Aenes. elsewhere (Phot. cod. 212) points to differences between Academics and Pyrrhonians proves nothing, for the Academics in question t were Philo's school. Photius’ conclusion that Aenes. also overthrows Plato (170b 37–9) is based on Photius’ dogmatic view of Plato, and is unrelated to criticism of the Philonian school. That V. school is criticized because it is the school of Tubero, to whom the work is addressed (169 b 32–5). (c) The extreme view of Plato as a sceptic does make him out to be almost Pyrrhonist: K 54. 38–43 shows us that he was seen as free from all doctrine, while Prol. in Plat. Phil. 10–11 (205. 2–207. 10 H) opposes those who used the term ⋯φεκτικός (205. 3, 11, 207. 9, 12 H), asserting that Plato πρεσβεύει ⋯καταληψίαν (205. 4, 14, 31 H). There are strong grounds for supposing that Prol. is replying to Aenes. in particular; see below, pp. 170 ff. (d) Aenesidemus lived when it was important to establish ancient authorities for one's views; D. L. 9. 71–3 gives a list of Pyrrhonist predecessors from Homer to Plato; Aenesidemus had himself made some kind of appeal to the thought of Heraclitus (PH. 1. 210–12), which was probably part of a wider appeal to ancient authority, (e) Sextus cannot be following Aenes. in refuting the view that Plato is Pyrrhonian, since nobody prior to Aenes. could have espoused that view; if Aenes. called Plato ‘sceptic’, then this may have been misunderstood by Sextus as being equivalent to ‘Pyrrhonist’, when it in fact meant only ‘prepared to look at both sides of the issue’.

⁶⁷ Cic. Ac. 1. 46, 2. 15, 74; S.E.Math. 7. 141–4 (from Antiochus' Canonical). The dispute is attested by S.E. PH. 1. 221 and D.L. 3. 51. To judge from Ac. the debate was comparatively unsophisticated around 87 B.C., but one imagines that rival claims of Philo and Antiochus to be the true followers of Plato soon gave rise to detailed quarrels over interpretation, as found in A: and anon. Prol. in Plat. Phil. 10–11.

⁶⁸ The Fourth Academy, or such of its members as accepted the meaningfulness of ⋯νάργεια as well as the impossibility of certainty, along with Cicero, Plutarch, A, etc.

⁶⁹ 54. 31–43, on Tht. 150c6: δι⋯ τ⋯ μηδ⋯ν ἔχειν σοφόν For Plato as a follower of Socrates' doubts see Ac. 2. 74.

⁷⁰ 53. 37 ff. on 150c4 (ἄγονός εἰμι σοφίας) suggests that the phrase had been interpreted absolutely by some. Similarly 54. 14 ff. on 150c6, 55. 34 ff. on 150c8–dl (οὐ πάνυ τι σοφός), 55. 45 ff. on 150d 1–2 (οὐδέ τί μοί ⋯στιν εὔρημα…ἔκγονον).

⁷¹ Note 63. 48–64. 3: ⋯κολουθεî τῷ πάντα ῥεîν τ⋯ μηδέν εἶναι πάγιον μηδ⋯ τ⋯ αὐτό Also 64. 28–36 on 152b 1: ⋯πακολουθήσωμεν.

⁷² 158b8–c 1. Of K all that is extant here is fr. 2 (on 157e–8a). But Cic. Ac. 2. 47–53, 88–9, shows that the sensations of the dreamer, drunkard, and madman were much discussed, and serious study of Tht. would quickly bring Plato's attitude to light, supporting the sceptic view that there is no obvious difference between these and healthy sensations.

⁷³ Ant. pp. 38–9; he claims that the ultimate source cannot be Pyrrhonian, failing to see the importance of the term ephectic in anon. Prol. L. G. Westerink, in his note on Olymp. In Phd. 6. 14, also sees the arguments as stemming from the second Academy.

⁷⁴ I cannot here discuss how far suspension of judgement was an ideal of Arcesilaus, but surely he did not refrain from προσδιορισμός (205. 10–12), nor did he wish to show Plato overturning number in Tht. (205. 20 H). That anon, is not refuting Carneadean probabilists is evident from I the fact that no provision had been made for grading impressions.

⁷⁵ Against me see Burkhard (above, n. 66, with my reply). He fails t o discuss all evidence for the sceptic view of Plato.

⁷⁶ The Ephectic sect is the Pyrrhonist sect in th e introductions to the Categories-commentaries of [Ammonius], Olymp., Elias, and Philopon, and probably in the lectures of Ammonius which inspired these commentaries. But Ammonius was also the likely origin of comments on the ephectic view of Plato in the late Plato-commentaries, for see Olymp, . In Phd. 8. 17Google Scholar (Westerink) = 51. 1–12 (Norvin): he wrote a book against the ephectic claim that Plato doubted the immortality of the soul.

⁷⁷ Besides the ten famous ‘tropes’ (D.L.9.87 ff.) we have the eight ‘tropes’ against cause-theory i (, S.E.PH. 1. 180)Google Scholar. The five ways of showing Plato to be ephectic are not referred to as ‘tropes’, but each is numbered, and one receives the impression of a very methodical exercise in forced interpretation. There is a similarity between the fourth ‘trope’ here and the way in which Aenes. counters the notion of truth in , S.E.Math. 8. 40Google Scholar.

⁷⁸ Compar e pp. 205. 6–7 H (⋯πιρρήματά τινα ⋯μφίβολά τε κα⋯ διστατικά) with Aenes. in Photius 169 b 39–40 (τ⋯ μ⋯ν τίθενται ⋯διστάκτως, τ⋯ δ⋯ αἴρουσιν ⋯ναμφιβόλως), cf. 170a29/32. Compare also p. 206. 26 (τ⋯ διαπορεîν ⋯δός πρ⋯ς τ⋯ καταλαβεîν) with , S.E.PH. 1. 210Google Scholar (⋯δ⋯ν εἶναι τ⋯ν σκεπτικ⋯ν ⋯γωγ⋯ν ⋯π⋯ τ⋯ν Ἡρακλείτειον φιλοσοφίαν).

⁷⁹ Photius cod. 212, 169b 38 describes Tubero as both an Academic and a συναιρεσιώτης of Aenes., probably meaning that they had taken the same side in the internal dispute between Philo and Antiochus. At D.L. 9. 116 Aenes. is said to have studied under Heraclides, for which Pappenheim (AGP 1 (1888), 37 ff.) proposed to read Heraditus. Since many mistakes in D.L. stem from abbreviations (seeGoogle ScholarMeyer, J., Diogenes Laertius and his Hellenistic Background (Hermes Einzelschriften 40 (1978)), 25–7Google Scholar), this would seem plausible.

⁸⁰ In his edition of anon, . Prol. (Amsterdam, 1962)Google ScholarWesterink, L. G. refers to Tht. 147d–8bGoogle Scholar, 198a–c, and 204b–e, ‘none of which is really pertinent’. One might also consider 195e–6c and 199b, where Plato fails to explain miscalculations, thus raising the question of the validity of arithmetic, and 152d2, etc., where nothing is allowed to be one; when combined with 204b–e, where the whole is just the sum of the parts, the absence of one abolishes arithmetic. Aenes. had an interest in the whole-part problem (, S.E.Math. 9. 337)Google Scholar, and at Math. 10. 216–17 numbers are regarded as mere multiples of a corporeal (hence flux-prone) unit.

⁸¹ I tend to reject the et which Halm adds between unum and constans, believing that the phrase may relate to the concept ἒν αὐτ⋯ καθ᾽ αὑτό at Tht. 152d 2, etc. Manutius' reading ne idem seems obvious; cf. K 64. 1–2: μηδ⋯ν εἶ ναι πάγιον μηδ⋯ τ⋯ αὐτ⋯.

⁸² Charmadas, however, was prepared to use the distinction between knowledge and right opinion (Cic. De Or. 1. 92).

⁸³ , S.E.PH. 1. 235Google Scholar; see my interpretation, loc. cit. (n. 20, above) sections 5–6.

⁸⁴ Perhaps Antiochus', excesses were in the realm of physics (Ac. 1. 24–9)Google Scholar, but see , S.E.Math. 7. 141–4Google Scholar. The orthodox view of Plato, at Ac. 1. 46 was fairly extreme, though not compared with the view refuted by anonGoogle Scholar. Prdl. (see nn. 66, 75).

⁸⁵ Stob, . Ecl. 2. 49. 25–50. 1Google Scholar (with Heeren's supplement), also 55. 6; on Eudoran origin see Dörrie, H., Platonica Minora (München, 1971), pp. 159–60Google Scholar.

⁸⁸ See 7. 14–20 and 58. 39–59. 2.

⁸⁷ After his introduction A moves quickly on to discuss moral questions on the basis of semi-relevant material in Tht. (cols. 5–11).

⁸⁸ A looks critically at Stoic, and to a lesser extent Epicurean, material in cols. 5–11.

⁸⁹ See n. 56 above.

⁹⁰ Olymp, . In Phd. 155. 24 ffGoogle Scholar. and 216. 1–26 (Norvin); see n. 55 above.

⁹¹ Glucker, Ant. pp. 90–7Google Scholar challenges the notion that Antiochus could have influenced Alexandria directly though some influence may have come about through his pupils Dio and Aristo. I add that I find no evidence of A ignoring Antiochus, or failing to mention him where it was relevant to do so.

⁹² See Glucker, , Ant. pp. 124–34Google Scholar, on the relationship of Ammonius to Plutarch.

⁹³ Mor. 392a–393a, cf.Whittaker, J., ‘Ammonius on the Delphic E’, CQ. n.s. 19 (1969), 185–92CrossRefGoogle Scholar. For K see 63. 48 ff.

⁹⁴ Mor. 393ef, 413ef, 435e; cf. K 58. 39–59. 2.

⁹⁵ See Mor. 385 cd. For him τ⋯ ζητεῖ is the source of philosophy (cf. Meno 86 bc), while the source of search is τ⋯ θαυμάζειν and τ⋯ ⋯πορεῖν (cf. Meno 84 bc). Note too Ammonius' approval of the ζητήσεις that have sprung from the Delphic maxims (385d). K too sees ζητήσεις as the essence of Platonism, 58. 23, 59. 7, 13.

⁹⁶ Mor. 746b shows the semi-sceptic B35 of Xenophanes to be a favourite quotation of Ammonius; the Delphic maxims which he praises invite and receive (387 f., cf. 431 a: τò ἂγαν τ⋯ς πíστεως) an Academic interpretation. As for self-awareness, Ammonius relates it to awareness of our mortal nature, including our inability to apprehend and the non-apprehensibility of things of this world (392 a—e), concluding ψεὑδετας ⋯ αισθησις ⋯γνοíα το⋯ ὃντος ειναι τò φαινóμενον. Ammonius', caution is seen at Mor. 391 efGoogle Scholar (cf. 744 b), and his willingness to judge by the criteria of TO CIKOS (435 C) and πιθανóτης (427 f.; cf. Lamprias' reply at 428 b) is significant.

⁹⁷ λóγον τιν⋯ ποθο⋯ντα δι⋯ τí καì διδασκλíαν τ⋯ς αíας. cf. Meno 98 a, K 3. 1—7.

⁹⁸ Philo's Biblical text is LXX Exod. 3: 14; the title ó⋯ν occurs widely in his writings. For Ammonius see Mor. 393 a ff.

⁹⁹ Philo's flux-doctrine and this-world scepticism occur most obviously atJos. 125–47. Human life is compared with a dream (126), then Philo dwells on man's changes from birth to old age (127–9), recalling Ammonius', remarks at Mor. 392deGoogle Scholar; cf. A's comments on the growth-process, cols. 69–70, probably on 152el. At 142 we read of TO τò ⋯ν τοῖς καταλψεσιν ⋯β⋯βαιον, and at 147 the flux/non-apprehension doctrine is applied specifically to earthly things, as opposed to heavenly things where both stability and cognition are to be found. Similar links between flux and non-apprehension are found in the Aenesideman passage at Ebr. 170, 172–4, 178–80.

¹⁰⁰ e.g. Dillon, MP, p. 144 n. 1Google Scholar, where he assumes that Philo must, as a dogmatist, be using Aenesidemus for his own dogmatic purpose.

¹⁰¹ Note the infallibility (in his field) of one who is wise, temperate, or philosophical (142), which is inspired by the Stoic concept of the sage (cf. Cic, . Ac. 2. 23–4, etc.)Google Scholar. We have an Antiochian view of the relationship of mind to senses at 143: it uses them as servants, and through them perceives more than they do alone (cf. Cic, . Ac. 2. 30Google Scholar; E, S.. Math. 7. 144, 226Google Scholar; Ph, . L.A. i. 29)Google Scholar. At 144 we learn that just as mind is a sensation of the sensations (cf. Cic, . Ac. 2. 30)Google Scholar so ⋯πιστήμηis a τέχνη τεχν⋯ν (cf. Antiochus, ' ‘ars quae vera a falsis possit distinguere’, Ac. 2. 57)Google Scholar.

¹⁰² Aristo Alexandrinus, fr. 5 (Mariotti): ‘ars … collectio est ex perceptionibus et exercitationibus ad aliquem vitae finem pertinens, etc.’. Mariotti compares Olymp, . In Gor. 63. 11Google Scholar = SVF i. 22. 3 ff. Aristo's, defection is recorded by Index Acad. 35. 14Google Scholar; Strabo (17. 1. 5) calls him a Peripatetic.

¹⁰³ e.g. Conf. 125–7, Fug. 188–93, Jos. 125–41.

¹⁰⁴ e.g. Ebr. 169, Conf. 127, Her. 246.

¹⁰⁵ The adjective καταλπτικός occurs only at Her. 32 in the course of a totally unremarkable division of presentations. In the same passage and later a t 209 he uses the pair ⋯κατ⋯ληπτος and καταληπτός; the latter term occurs only 13 times in other treatises, often with a negative, never in such a way a s t o suggest strong Stoic influence.

¹⁰⁶ Philo of Larissa seems to have recognized criteria (Stob, . Eel. 2. 40. 19 Wachsmuth)Google Scholar, probably intellect and senses;K2. 23–32 may follow, as does Plut, . Mor. 1024eGoogle Scholar. Antiochus insists on the retention of the Stoic criterion.

¹⁰⁷ Ebr. 169, 192, 200, 205.

¹⁰⁸ See E, S.. PH. 1. 13Google Scholar; cf. how Plato and Protagoras are convicted of dogmatism at 1. 223 and 1. 219 respectively.

¹⁰⁹ Clem, . Strom. 8. 16. 2Google Scholar = SVF 2. 37. 10 ff.; Ac. 2. 53.

¹¹⁰ Note Cic, . Ac. 2. 19–20, 30Google Scholar, making much of the cognitive power of the senses.

¹¹¹ It is clear that flux-doctrine is most naturally at home in Platonism. The language of images and dreams suits Platonism (⋯ναζωϒράϕειν, ⋯ειδωλοποιεȋν, 126), as does the eternal wakefulness of the heavenly world (147). The phrase εἰκ⋯σι καì εὐλόϒοις πιθανότησι(143) combines Platonic and Academic terminology. Pyrrhonism did not recognize the concept of πιθανότης, nor the distinction between ὂντα and μ⋯ ὂντα(126). At 142 we seem to meet the notion that κατάληψις can fall short of certainty, as in Philo of Larissa and [Galen] 14. 685 (Kūhn). For a reply to H. von Arnim's theory that the Heraclitism of Jos. comes from Aenes, . (Quellenstudien zu Philon von Alexandria (Berlin, 1888), 79 ff.)Google Scholar see Burkhard, op. cit. (above, n. 66), pp. 175–82.

¹¹² A striking case occurs early in Somn. i, where Philo argues that the fourth items in various series of four components are ⋯κατάληπτον while the rest are καταληπτά. But note that κατάληψις need not imply certainty in Philo, n. 111 above.

¹¹³ Usually σκεπτικοί are just ‘inquirers’ in Philo (see Janacek, K. in Listy Filologicke 102 (1979), 65–8)Google Scholar, and for this meaning see Her. 247, 279, Ebr. 98, 202. The last-mentioned case of the non-sceptic sense is notable, as it comes at the end of the Aenesideman passage (preceded by σκέψεις ⋯μύθητοι).

¹¹⁴ Against this see Burkhard, op. cit. (n. 66), pp. 182–94.

¹¹⁵ See Burkhard, op. cit. (n. 66), p. 193.

¹¹⁶ I judge from Phot. Cod. 212, 170a 2–3, 5–6, 9–11, where Aenes. seems to be criticizing the dogmatic terminology of Academic Heraclitism: ‘at one time such, at another such, for one man such, for another such’, etc. At 17–22 he complains openly about dogma in the Academy, attacking such distinctions as virtue/vice, good/bad, true/false, probable/improbable, real/not-real.

¹¹⁷ Perhaps a sceptic's attack was directed against ⋯πόδειξις and σημεȋα rather than πίστεις generally, but D.L. 9. 78, following mention of Aenes., says that sceptics abolish πίστεις concerning πράϒματα. Moreover Philo's πίστεις imply τò πιθανόν, which Aenes. rejects (Photius 170a 7, 20), just as A's account of Pyrrho involves rejection of the ⋯ναρϒής/⋯μυδρός distinction (61. 35).

¹¹⁸ Aenesidemus in Phot. cod. 212, 169b42: οὐδεìς αὐτ⋯ν τò παράπαν οὒτε ⋯κατάληπτα πάντα εἲρηκεν … and cf. K 61. 30–7: μηδεμίαν … διαϕορ⋯⋯ κατ⋯ τò … καταληπτòν ἣ ⋯κατάληπτον, which is intended to reject both terms.

¹¹⁹ See Cic, . Ac. 2. 56, 84Google Scholar.

¹²⁰ Aenes, . in books 2 and 5 of his Pyrrhonian Logoi attacked both αἰτία and αἰτιολοϒία (Photius170b 3–8, 17–22)Google Scholar.

¹²¹ Perhaps όπωσο⋯⋯ at 166 and 198, found only 5 times in Philo but traced in Aenes. (D.L. 9. 78). Possibly κατατρίβω at 195, found only once elsewhere in Philo, (L.A. 2. 98)Google Scholar, but used similarly in Phot. Cod. 212, 169b24. Following Janacek, K. (Eirene 14 (1976) 99)Google Scholar, one is forced to consider also τò παράπαν (186), but this term is common in Philo.

¹²² Since no attempt has been made to preserve the numbering of ‘tropes’ this is not surprising, Also missing is the ninth mode (tenth in D.L. 9. 87), but then there were only nine modes in Aenes.’ Hypotyposis (Aristocles, in Eus. P.E. 14. 760b)Google Scholar, so that the tenth must have been a late development.

¹²³ See Burkhard op. cit. (above, n. 66), pp. 180–2 on the non-Aenesideman nature of the Heraclitism here.

¹²⁴ See Whittaker, J., ‘Ammonius on the Delphic E’ CQ n.s. 19 (1969), 185–92CrossRefGoogle Scholar, on the connections.

¹²⁵ cf. Smyrnaeus, Theon, Expos. 14. 17–16. 2 (Hiller, )Google Scholar; Albinus, Prol. 6 (150. 15 ff. H)Google Scholar. Perhaps also Cic, . Ac. 2. 60Google Scholar, Plut, . Mor. 391Google Scholar de (an odd reference to the mysteries in a Platonic context).

¹²⁶ See the note of F. H. Colson in the Loeb. The length of the quotation and the naming of the source of the earlier passage is remarkable in an author who usually only alludes to his Platonic sources.

¹²⁷ My own collection of Platonic allusions agrees with Dillon's, findings (MP, p. 140)Google Scholar, gathered from the Cohn-Wendland's index, as to the important dialogues for Philo. Tim. far outstrips others, and Crat. might also have been mentioned.

¹²⁸ See Jones, R. M., ‘Posidonius and the flight of the Mind’, CP 21 (1926), 97–113Google Scholar.

¹²⁹ For an odd collection of Academic placita on the senses, of a basically Fourth Academic character, see Diels, , Dox. 396 b 5–7, 17–19, 398 b 24, 403 b 8–11Google Scholar.

¹³⁰ It is difficult to make out a case for όμοίωσις having been the New Academic telos, yet A seems to be committed to the defence of such a claim, since the telos is one of two supreme issues (Ac. 2. 29). Glucker, has shown (Am. pp. 55–60)Google Scholar that Cic. was aware of a telos espoused by Plato, (Ac. 2. 129)Google Scholar,for his position was said to beclose to that of Erillus(the good = knowledge) and to have something in common with the Megarics (the good is one/same/eternal). Eudorus (Stob, . Eel. 2. 55. 15–17Google Scholar, on which see Dörrie, H., Platonica Minora p. 303)Google Scholar was aware that some saw wisdom + pleasure (allegedly the second good of Phil. 66 b 1–3) as the telos of human life, suggesting that he was seen to propose some more divine telos even before Eudorus. Indeed both Posidonius (cf. F187 Kidd) and Antiochus (in Clem, . Strom. 2. 131. 2–133. 3Google Scholar, see Hoyer, R., De Antiocho Ascalonita, Diss. (Bonn, 1883), 26 ff.)Google Scholar seem aware that όμοίωσις θε served as a kind of telos in Plato, though Antiochus recognized that the problem was extremely complex. Ac. 2. 129 demonstrates that Antiochus' Academic opponents also recognized the complexity of the issue, but saw that wisdom was crucial; Eudorus saw ϕρόνησις as the key to όμοίωσις (Stob, . Eel. 2. 49. 10)Google Scholar, placing emphasis on Tht. 176c4 and b2 (μετ⋯ ϕρονήσεως). Note in particular that the wisdom of Tht. 176c4 is not merely a system of factual knowledge, and thus not entirely incompatible with Academic hopes.

¹³¹ One might also mention Theomnestus, see below, p. 182.

¹³² Thrasyllus wrote on the first Principles of Pythagoreanism and Plato (Porph, . Vit. Plot. 20)Google Scholar, and believed that Democritus had been a follower of Pythagoras (D.L. 9. 38).

¹³³ He is a major source of Theon's Expositio on Platonic mathematics; his astronomy features twice in Achilles, Intr. ad Aratum (43. 9, 46. 30 Maass)Google Scholar; musical work is mentioned by Porphyry, (Ptol. Harm. 266)Google Scholar; he became Tiberius' astrologer.

¹³⁴ Pythagoras is alleged to have introduced the problem of growth (70. 5–9).

¹³⁵ Albinus, Prol. 4 (149. 13 H), D.L. 3. 56–61Google Scholar.

¹³⁶ See Zeller, E., Eclectics (London, 1883), trans. Alleyne, , pp. 104 n. 1, 105 n.l = Ph. Gr. iii⁵ pp.634n.2,635n. 1Google Scholar;Dōrrie, H., Platonica Minora (Mūnchen, 1971),pp. 159–60, 303Google Scholar; Theiler, W., ‘Philo von Alexandria und der Beginn des kaiserzeitlichen Platonismus’ in Parusia: Festschrift für J. Hirschberger (Frankfurt, 1965), pp. 214Google Scholar, 217, etc. = Untersuchungen zur antiken Literatur (Berlin, 1970), pp. 498, 501, etc.Google Scholar; Dillon, , MP, p. 116Google Scholar: ‘After his summary of the whole subject-matter of ethics, Arius, it seems, goes on to give us a number of these problêmata’. Dillon appears to mean Eudorus' problemata, which had made up his work on the divisions of philosophy (Eel. 2. 42. 9), interpreting 2. 45. 7–10 to mean that Arius (?) presents some of these, but in revised order. He regards the material up to 55. 21 as Eudoran (p. 125), and 55.22 ff. as Arius’ summary. Though the arguments hitherto used for Eudoran influence after 2.45. 7 are less than compelling, I am in broad agreement with the conclusion. Ecl. 2. 45. 11–57. 12 gives a Platonic-Academic view of ethics to balance the later Stoic (2. 57. 13 ff.) and Peripatetic (2. 116. 19 ff.) accounts, but the material is handled differently in so far as the emphasis falls on the problêmata themselves, not on a single set of doctrines; Plato's answers are given prominence, but interest is shown in related thinkers, as might be expected of an Academic. Interpretation of Plato is rather detailed to have originated with Arius, a Stoic, and Eudorus would be Arius' natural source for Academic material when he has just been consulting Eudorus’ book of problêmata. Even if I am mistaken here, it must be noticed that the previous view of Eudorus has owed much to the same assumption.

¹³⁷ Glucker, (Ant. p. 97, n. 268)Google Scholar rightly questions the usual view (e.g. Dörrie, , Hermes 79 (1944), 26)Google Scholar = Platonica Minora p. 298) that Eudorus must have studied with Antiochus. Dillon, (MP, p. 115)Google Scholar suggests indirect influence via Dio. Nothing in extant fragments had to be learnt from Antiochus' school; indeed Eudorus' basic study may have been with that of Posidonius, for (a) he is the source of Posidonian and Diodoran material in Achilles, (b he may be the source of Posidonian interpretation of Tim. at Plut, . Mor. 1023Google Scholar be, (c) his view of the Platonic goal, deeply indebted to Tim. 90a-d (see n. 177), may derive from Posidonius' similarly inspired view that one must follow the daemon within one, as it is akin to that which governs the universe (Galen, , Plac. 5. 469Google Scholar = F187 Kidd), and (d) Posidonian emphasis on controlling the πάθη may account for the prominence which Eudorus gives to the philosophy of impulse as one of the three parts of ethics (Eel. 2. 42. 13–24 and particularly 2. 44. 3–6). Just as Aristo and Cratippus could later describe themselves as ‘Peripatetics’, so a Posidonian who developed (a) doubts and (b) love of Plato might come to describe his philosophy as ‘Academic’ once the school itself had ceased to exist.

¹³⁸ As H. Chemiss (Loeb) points out, Plut. commits himself to proceeding on the basis of τò εἰκός at 1014a; τò εἰκός is again coupled with τò πιθανόν at 728 f.; and we have an example of Plut. insisting on the application of τò εἰκός to Plato's own opinion at 430b: πρòς τ⋯⋯ ⋯κείνου διάνοιαν ⋯πάϒειν τò εἰκός. I find no clear example of the application of these criteria in the remains of Eudorus.

¹³⁹ Stob, . Eel. 2. 42. 7Google Scholar (Wachsmuth), Simpl, . In Categ. p. 187Google Scholar. 10 (Kalbfleisch), anon, , i Intr. ad Aratum (p. 97 Maass)Google Scholar. The adjective is used, not the phrase ‘from [the] Academy’.

¹⁴⁰ Philo, QG 3. 33Google Scholar, Sen, . Ep. 88. 44Google Scholar, Tac, . Dial. 31Google Scholar; all point to the fact that Academics are expected to argue in utramque partem, and are most easily compared with Pyrrhonists.

¹⁴¹ Glucker, , Ant. p. 114Google Scholar does not appear keen to accept this idea of Zumpt, K. G., Über den Bestand der philosophischen Schulen in Athen und die Succession der Scholarchen (Berlin, 1844), p. 69Google Scholar, followed by Zeller, Ph. Gr. iii.⁵ p. 630 n. 4Google Scholar.

¹⁴² Glucker, , Ant. p. 115Google Scholar, observes that Cicero does not send his son to hear Theomnestus in 45 B.C., even though that Academic was available there in 44 B.C.

¹⁴³ QG 3. 33, Cong. 52, Fug. 209; note also the use of Academic terminology when Philo speaks of the ‘Egyptian sophists’, Migr. 76, Somn. 1. 220.

¹⁴⁴ An Academic rather than a Pyrrhonist; see Glucker, , Ant. p. 281Google Scholar.

¹⁴⁵ Plut, . Luc. 28. 7Google Scholar: ‘Antiochus the philosopher’, 42. 2–3: ‘ not the allegedly “new” Academy’; no title is afforded to Antiochus or Aristus at Brut. 2. 1–2, Cic. 4. 1. Strabo talks similarly of ‘Antiochus the philosopher’ (16. 2. 29), while Carneades is still the archetypal Academic (17. 3. 22).

¹⁴⁶ Eus, . Praep. Evang. 14. 763d (from Aristocles)Google Scholar.

¹⁴⁷ See above, n. 129. Note that in contrast to Plato and other early thinkers, who are together credited with the view that the senses are ‘false’ in the preceding lines, the Academics are said to accept that they can be healthy, but not that they have sufficient accuracy. The ‘Academic’ position is thus depicted as being more reasonable than Plato's (though well short of Antiochus').

¹⁴⁸ Plato in Stob, . Eel. 2. 49. 8 ff., 2. 53. 1 ff., 2. 54. 10 ff. (Wachsmuth)Google Scholar; also Plut, . Mor. 1013b, 1019e, 1020c; possibly some doxographical entries in Achilles, Intr. ad Aratum. Pythagoras inGoogle ScholarStob, . Eel. 2. 49. 8–9, 16–21Google Scholar; Simpl, . In Phys. p. 181. 10 ff. (Diels)Google Scholar.

¹⁴⁹ Achilles, , Intr. ad Aratum 30. 20 (Maass)Google Scholar; probably other views of Posidonius and Diodorus reported in this work.

¹⁵⁰ plut, Mor. 1013bGoogle Scholar: no doubt he is approving of (a) Xenocrates' use of the principles of number, and (b) Crantor's having constructed the soul out of the materials which it is meant to cognize. Posidonius' interpretation at 1023 be appears to preserve the element of number, probably by using the Same and the Different as vehicles for the One and the Dyad (as in Plut. at 1024d, probably from Eudorus); it certainly postulated that the soul was composed of an intelligible stuff and a sensible stuff, in conformity with the requirement that it should apprehend intelligibles and sensibles, and we know from E, S.. Math. 7. 93Google Scholar that Posidonius adhered to the like-by-like theory of cognition when interpreting Tim.

¹⁵¹ Simpl, . In Categ. 159. 32, 174. 14, 187. 10, 206. 10, 236. 28, 246. 22, 256. 16, 263. 27, 268. 13 (Kalbfleisch)Google Scholar.

¹⁵² Stob, . Eel. 2. 46. 5–10Google Scholar; Achilles, Intr. ad Aratum 30. 20 (Maass)Google Scholar.

¹⁵³ The material in Stob. comes from his διαίρεσις το⋯ κατ⋯ ϕιλοσοϕίαν λόϒου (Eel. 2. 42. 8 Wachsmuth).

¹⁵⁴ For the concept of διάρθρωσις ⋯ννοι⋯ν in K see 46. 44, 47. 45, 53. 46, 56. 36. For διάρθρωσις in Eudorus see Stob, . Eel. 2. 49. 9, 2. 51Google Scholar. 8 (Wachsmuth); in neither case does the term appear with cwoia, though in the former case the object of the διάρθρωσις certainly was an ἓννοια of the τέλος.

¹⁵⁵ e.g. Cic, . Ac. 1. 18, 43, 2. 15Google Scholar.

¹⁵⁶ If Cic. follows Philo, at Ac. 2. 116–34Google Scholar.

¹⁵⁷ Stob, . Eel. 2. 49. 18–23Google Scholar: this shows various ways, including the Pythagorean one, in which Plato has presented his doctrine of the telos. At 2. 50. 2–4 (Wachsmuth) the unity of the doctrine is stressed.

¹⁵⁸ This may be typical of his age, for so do Posidonius (T95 Kidd on the tripartite soul; perhaps he too saw much of Tim. as Pythagorean, for see E, S.. Math. 7. 93–4)Google Scholar and, one suspects, Thrasyllus. See also AT 70. 5–9 and Achilles, , Intr. adAratum 37. 29–38. 2 (Maass)Google Scholar, both of which may be the work of Eudorus. Moderatus and Numenius continue the trend. Known examples in Eudorus are Stob, . Eel. 2. 49. 8, 18–21 (Wachsmuth) andGoogle ScholarSimpl, . In Phys. 181. 10 ffGoogle Scholar. (Diels). The trend stems from a search for increasingly ancient authorities.

¹⁵⁹ For cosmology see n. 149; to expect to find distinctive Posidonian ethics in Eudorus would be highly optimistic, though he certainly follows Posidonius in regarding the distinction between rational and non-rational parts of the soul as fundamental.

¹⁶⁰ Seen in his ‘Correction’ to the text of Arist, . Meta. 988 a 7Google Scholar, recorded by Alex, . Aphr. In Meta. 59. 1 ffGoogle Scholar.

¹⁶¹ Simpl, . In Phys. 181. 10 ff. (Diels)Google Scholar.

¹⁶² He usually gives the impression of being firmly in agreement with those doctrines which he attributes to Plato, and he speaks on his own authority when criticizing the younger Peripatetics (Eel. 2. 46. 13–17 Wachsmuth).

¹⁶³ See above, n. 151.

¹⁶⁴ A similar tendency is found in Nicostratus, Simpl, . In Categ. 73. 15 ff. (Kalbfleisch)Google Scholar.

¹⁶⁵ He is quite capable of arguing that the universe is πρός τι at Simpl, . In Categ. 188. 31 ff. (Kalbfleisch)Google Scholar.

¹⁶⁶ ibid. 206. 10 ff.

¹⁶⁷ ibid. I cannot here discuss the relationship of Ps.-Archytas to Eudorus, but tend to believe that ‘Archytas’ follows later.

¹⁶⁸ Perhaps their general futility is a reaction against the solemnity with which Categ. was treated by Peripatetics.

¹⁶⁹ Eudorus, in Ecl. 2. 48. 1, 2. 43. 19Google Scholar; K 4. 31–5. 3, 9. 30–11. 40.

¹⁷⁰ Eudorus, in Ecl. 2. 47. 12 ff, etc.Google Scholar; K 5. 14–8. 6.

¹⁷¹ Eudorus, in Eel. 2. 43. 19–44. 1Google Scholar; K 8. 23–7. Note that Dio of Alexandria (Plut, . Mor. 612e)Google Scholar also shares the interest of both these passages in symposia.

¹⁷² For Eudorus see n. 158; for Posidonius see T 91, T 95 (Kidd).

¹⁷³ See K 70. 5–9.

¹⁷⁴ K 7. 14 and Stob, . Eel. 2. 49. 8 ffGoogle Scholar.

¹⁷⁵ K has δεíξομεν (preceded by a brief lacuna), cf.Eel. 2. 50. 6 ff.

¹⁷⁶ As Didasc. 28 (181.36H); hence at 2 (153.4–7H) we meet, όμοίωσις πρòς τò θεȋον, a better, description of the process described at Tim. 90a–d, cf. Albinus, Prol. 5 (150. 8–12H)Google Scholar. Note that Eudorus' comments on Pythagoras' directive ‘follow God’ do make it clear what kind of God \ he envisages man following: the intelligible principle of cosmic harmony (Stob, . Eel. 2. 49. 16–18 Wachsmuth)Google Scholar, not an anthropomorphic one. Hence he has the same God in mind as Albinus and [Alcinous], but as yet there is no need to distinguish this God from a higher one. Presumably the mistake against which [Alcinous] reacts had not yet been made.

¹⁷⁷ See Stob, . Eel. 2. 47. 6, 49Google Scholar. 19–21, 50. 7, 53. 1–4; Tim. also determines what kind of God. we must follow, 49. 16–18.

¹⁷⁸ Of Eudorus' source-passages this alone has the phrase which he sees as Plato's clarification of the concept: κατ⋯ τ⋯ δυνατόν (Stob, . Eel. 2. 49. 10 Wachsmuth)Google Scholar.

¹⁷⁹ See above, nn. 154, 178.

¹⁸⁰ Mor. 1000e; Prol. 6 (150. 21 H); cf. Didasc. 5 (158. 3 H).

¹⁸¹ Stob, . Eel. 2. 53. 7–17Google Scholar (Wachsmuth), cf. 2. 42. 20.

¹⁸² K9. 37–10. 3; Eudorus can talk of the virtues separately, Ecl. 2. 43. 12, but elsewhere tends to think of it as a single entity (e.g. Ecl. 2. 50. 4–6, 51. 1–2 Wachsmuth).

¹⁸³ K 9. 37–10. 3 allows the εὐϕυίαι to oppose one another, unlike the perfected virtues. Eudorus introduces his conception of the ὑποτελίς, which compensates for the remoteness of his τέλος (Stob, . Eel. 2. 47. 12 ff. Wachsmuth)Google Scholar.

¹⁸⁴ Eudorus on the ὑποτελίς (Stob, . Eel. 2. 48. 3–5 Wachsmuth)Google Scholar, and K 7. 20–5 on οἰκείωσις in Socrates and the Sophists. Note that Eudorus associates οἰκείωσις with the ὑποτελίς at 2. 47. 12–18, so that their claims are almost identical.

¹⁸⁶ Particular thanks are due to Professor John Dillon for his comments on my case during the early stages of its development, and to an anonymous referee for helpful comments.