¹ Eus, . PE 15.9.1–2Google Scholar = fr. 7 Des Places (Places, É. Des (ed.), Atticus: Fragments (Paris, 1977))Google Scholar. As Deuse, W., Untersuchungen zur mittel- undneuplatonischen Seelenlehre (Wiesbaden, 1983), pp. 9f.Google Scholar, observes, Atticus' words sound less like a description of an actually existing unity among Platonists than an exhortation to concentrate on this unifying doctrine, especially in the face of contradictory claims arising from Peripatetics. For Platonists themselves the nature and immortality of the soul had become problematical issues: ‘Man hat gelernt zu differenzieren, wenn von “der” Seele gesprochen wird, und man betrachtet es nicht mehr als nebensaählich, die Struktur der Seele auch unter dem Aspekt des irrationalen Seelenlebens zu untersuchen’ (ibid. p. 10). The extent to which already Xenocrates contributed to the problem of the irrational soul wiB be part of the present investigation.

² See the catalogue of Speusippus' writings in D.L. 4.4–5 = Tl (testimonia and fragment numbers refer to Tarán, L. (ed.), Speusippus of Athens (Leiden, 1981)CrossRefGoogle Scholar; henceforth ‘Tarán’); on the περì ψυχ⋯ς see Tarán's commentary, p. 194. For Xenocrates’ writings, see D.L. 4.11–14 (pp. 157–8, Heinze = Heinze, R. (ed.), Xenokrates (Leipzig, 1892)Google Scholar; henceforth ‘Heinze’ or ‘H’; fr. 2 IP = Parente, M. Isnardi (ed.), Senocrate – Ermodoro: Frammenti (Naples, 1982)Google Scholar; henceforth ‘Isnardi Parente’ or ‘IP’) and Dörrie's, H. overview by subject matter in RE IX A, 2 (1967), cols. 1515–16Google Scholar. Xenocrates' ψυχ⋯ς supposedly contained two books (D.L. 4.13).

³ For Speusippus the only explicit attestation to the immortality of the soul is fr. 55 Tarán (= fr. 75 H), which we will examine below; we do know, however, that he believed that soul was neither number nor magnitude, that nous, as the seat of knowledge in the soul, was divine, and that the cosmos was eternal and without an origin in time – all beliefs compatible with the immortality of the soul; see Tarán, pp. 47–8. On Xenocrates see, besides fr. 75, frr. 73–4 H = 209–10 IP.

⁴ In Phd. 177.1–7, pp. 107–9 Westerink ( = Westerink, L. G., The Greek Commentaries on Plato's Phaedo, volume II. Damascius (Amsterdam, Oxford, and New York, 1977)Google Scholar; Speusippus, fr. 55 Tarán; Xenocrates, fr. 75 H = 211 IP). The above translation is adapted from Westerink's. For comments see Dörrie, H., Platonica Minora (Munich, 1976), pp. 426–7Google Scholar and Westerink, op. cit., pp. 106–9; cf. Tarán, p. 372.

⁵ Pp. 371–4.

⁶ ‘Bipartition of the Soul in the Early Academy’, JHS 77 (1957), 118Google Scholar.

⁷ Although more can be said about this fragment, and Xenocrates in particular, than Tarán allows, I do think he is correct, as opposed to earlier commentators, that ⋯λοϒία indicates Speusippus' and Xenocrates' belief about human souls and not irrational animals. References to animals and plants may be implied, however, in the case of some of the other authors here mentioned; see Dörrie, op. cit. (n. 4), pp. 426ff.; cf. n. 89 below.

⁸ This paragraph, except for the reference to frr. 18 and 57 (nn. 9 and 12 below), is based upon fr. 15 H = 213 IP (Aët. 1.7.30; Dox. gr. 304). A precise analysis of Xenocrates' gods and their specific locales, examined against the background of Plato's Timaeus, is given by Baltes, M., ‘Zur Theologie des Xenokrates’ in Broek, R. van den, Baarda, T., and Mansfeld, J. (eds.), Knowledge of God in the Graeco-Roman World (Leiden, 1988), pp. 43–68CrossRefGoogle Scholar. On the tripartite cosmos, cf. Burkert, W., Lore and Science in Ancient Pythagoreanism, trans. Minar, E. L. Jr (Cambridge, Mass. 1972), p. 245 n. 36Google Scholar; Mansfeld, J., The Pseudo-Hippocratic Tract περì ⋯βδομάδων Ch. 1–11 and Greek Philosophy (Assen, 1971), p. 43 n. 34Google Scholar; cf. also the triad ὂλυμπος-κόσμος-οὐρανός ascribed to Philolaus in DK 44 A 16 (thereto, see Kerschensteiner, J., Kosmos (Munich, 1962), pp. 49f.Google Scholar; Burkert, , op. cit., pp. 243ff.Google Scholar; Mansfeld, , op. cit., pp. 42f.Google Scholar). Another way of understanding the Xenocratean universe, though not at variance with the three-tiered structure presented here, is in terms of an essential bipartition into an intelligible incorporeal order and a sensible cosmic order, the latter being subdivided into celestial and sublunary spheres; cf. Isnardi Parente, pp. 335, 405f., 407f. For a more detailed survey of Xenocrates' cosmology and metaphysical principles than can be given here, see Dörrie, art. cit. (n. 2), cols. 1520–1.

⁹ Fr. 18 H = 216–17 IP (, Plu.Plat. quaest. 1007fGoogle Scholar; see full annotation in Loeb, H. Cherniss's edn., Plutarch's Moralia, xiii/1 (Cambridge and London, 1976), n. a, pp. 92f.)Google Scholar. Cf. n. 15 below.

¹⁰ Cf. Dillon, J., The Middle Platonists (London, 1977), pp. 25f.Google Scholar; id., ‘Xenocrates’ Metaphysics. Fr. 15 (Heinze) Re-examined', AncPhil 5 (1985), 51 n. 2Google Scholar; Krämer, H. J., Platonismus undHellenistische Philosophie (Berlin and New York, 1972), p. 162 n. 42Google Scholar; Mansfeld, , op. cit. (n. 8), p. 121 n. 285Google Scholar; Bakes, , art. cit. (n. 8), 51Google Scholar. Although earlier in the Phaedrus, Zeus, driving his winged chariot, is called the great leader in heaven (μέϒας ⋯ϒεμὼν ⋯⋯ oὐρανὡ, 246e4), he ultimately leads the heavenly hosts beyond the sights within heaven (⋯ντòς οỦρανο⋯, 247a5) to the ‘summit of the arch that supports the heavens’ (Hackforth's translation of ἂκραν ⋯πì τ⋯⋯ ὑπουράνιον ⋯ψȋδα, 247a8–bl). Here, too, immortal souls actually pass beyond the summit to stand upon the surface of heaven (⋯πì τὡ τοῡ oὐρανο⋯ νωτψ, 247b7–cl) where they survey the sights outside the heavens (τ⋯ ἒξω το⋯ oὐρανο⋯, 247c2). The god of the ps.- Aristotelian, De mundo, 397^b25ff.Google Scholar, ‘has the highest and first dwelling and is called supreme (ὔπατος) because according to the poet he dwells “on the loftiest crest” (⋯κροτάτῃ κορυϕῇ II. 1.499) of the whole heaven’; see further Mansfeld, op. cit. (n. 8), pp. 121ff. Aristotle, , M.A. 699^b32ff.Google Scholar, adduces (not quite fittingly) the Homeric Zεὺς πατος (II. 8.22) as an illustration of an unmoved mover outside the universe (see Nussbaum, M. C., Aristotle's De Motu Animalium (Princeton, 1978), pp. 320f.Google Scholar).

¹¹ The text goes on to identify the Dyad with the world-soul. This, however, poses a difficulty, primarily because the world-soul is composed of both the Monad and Dyad (see fr. 68 H = 188 IP [, Plu.An. proc. 1012d–e]). For a concise discussion of the problem and two attempts towards its solution – either a textual corruption or a misunderstanding on the part of Aëtius – see Dillon, art. cit. (n. 10), 47–52. Baltes, art. cit. (n. 8), 50f, offers a third possibility, which commends itself by preserving the special sphere of the world-soul as well as its identity: Monad and Dyad are two aspects of a single divine being whose unity is not to be seen as a mixture but rather as a ‘differenzierte Einheit, in der der Nous als das mannliche Prinzip gegenūber der Seele als dem weiblichen Prinzip dominiert’Google Scholar.

¹² ενοκράτης κατ⋯ μις οἵεται ⋯πιϕανείας κεȋσθαι (κινεȋσθαι ps. Plu.) τοὺς ⋯στέρας (fr. 57 H=162 IP); thereto, see Heinze, p. 72; Döeeie, art. cit. (n. 2), 1524; on the variant κεȋσθαι/κινεȋσθαι, see Isnardi Parente, pp. 379f. If Xenocrates distinguished between the fixed stars and the planets, it is not necessary to see this as marking a major division of his cosmos. It seems rather that the upper hierarchy of heaven, from the fixed stars on its surface and the subordinate planets to the region of the moon, formed one partition; similarly, Baltes, art. cit. (n. 8), 56.

¹³ ⋯ρέσκει δ⋯ αὐτ «θείας τιν⋯ς δυνάμεις» καì ⋯⋯διήκειν τοȋς ὑλικοȋς στοιχείοις. Lac. suppl. Heinze: θε⋯ν δυνάμεις Zeller. According to Baltes, , art. cit. (n. 8), 61Google Scholar, the divine powers are emanations of the Olympian gods; on their number (twelve) and identity, see ibid. 61–3.

¹⁴ τούτων (i.e. θείων δυνάμεων) δ⋯ τ⋯ν μ⋯ν «δι⋯ το⋯ ⋯⋯ρος Ἅιδην ὡς» ⋯ειδ⋯ προσαϒορεύει. Heinze's restoration of the lacuna, inasmuch as it preserves ⋯ειδ⋯ iuxta lacunam, is preferable to Wachsmuth's «δι⋯ το⋯ ⋯⋯ρος προσϒείου Ἄιδην. But ⋯ειδής should be taken as itacism for ⋯ïδής (⋯ – Fιδεȋν; cf. variants at Plato, , Grg. 493b4Google Scholar and Dodds, E. R. (ed.), Plato: Gorgias (Oxford, 1959)Google Scholar, ad loc.) and translated ‘invisible’ (so also Isnardi Parente, p. 415), rather than ‘formless’ (so e.g. Dillon, , op. cit. (n. 10), 25Google Scholar). The invisible Hades is the proper domain of the ⋯⋯ρατοι δαίμονες. For a play on the invisible soul and Hades, cf. Plato, , Phd. 79a9f.Google Scholar, b7 and esp. 80d5–7 (thereto see Gallop, D. (trans.), Plato: Phaedo (Oxford, 1975), p. 143)Google Scholar. On further variations of the theme “ Αιδης – ⋯ειδής, see Heinze, p. 147 and n. 2; also Plot. VI. 4.16.37; Porph, . Sent. 29, p. 18, 12fGoogle Scholar. Lamberz (thereto see Dorrie, , op. cit (n. 4), p. 449Google Scholar).

¹⁵ Fr. 18 (n. 9 above). Given the triadic tendency of Xenocrates' philosophy, A. B. Krische supposed that Xenocrates also posited a Ζεὺς μέσος who animated the region above the moon (Die theologischen Lehren der griechischen Denker (Göttingen, 1840), p. 324)Google Scholar; this supposition has continued to Kramer, , op. cit. (n. 10), p. 124Google Scholar, and Baltes, , art. cit. (n. 8), 59 and n. 61Google Scholar, who adduces Plut, . Quaest. conviv. 745bGoogle Scholar. On the other hand, Cherniss, , op. cit. (n. 9), n. a, p. 93Google Scholar, argues that Plutarch would not have failed to mention a median Zeus in his discussion of the correspondence of Plato's tripartite soul to the lowest, middle, and highest notes of the musical scale, and that it may well have sufficed Xenocrates to have treated Zeus ‘only in his two commonly recognized aspects as ὒψιστος and χθόνιος …’ (cf. Dillon, , op. cit. (n. 4), 27Google Scholar; Schwabl, H., RE Suppl. XV col. 1345Google Scholar; Boyancé, P., ‘Xénocrate et les Orphiques’, REA 50 (1948), 226f.)CrossRefGoogle Scholar. This seems to me a still valid observation and may serve us as an instructive example when we come to examine Xenocrates' divisions of the soul: it is not necessary to see tripartition everywhere in Xenocrates.

¹⁶ From Plutarch, , Fac. lun. 943f–944a (fr. 56H = 161 IP), we get corroborating evidence that the physical domains of Xenocrates' universe were composed of specific elements (as we have already seen in the case of the moon) and also of different degrees of physical density (τò πυκνόν)Google Scholar.

¹⁷ Fr. 5 H = 83 IP (M. 7.147 (ii.36 Mutschmann)). Sextus' testimony concludes by listing the three Fates and the level of existence to which Xenocrates assigned each of them. On Xenocrates' use of the Μοȋραι as cosmological divisions, see Harder, R. (ed.), ‘Ocellus Lucanus’ (Berlin, 1926), p. 91Google Scholar.

¹⁸ See n. 10 above.

¹⁹ The opiniable existence of heaven is a composite (σύνθετον) of the sensible and the intelligible, since it is visible by sense perception but intelligible by astronomy (I am assuming that όρατ⋯ μ⋯ν ϒάρ ⋯στι τῇ αἰσθήσει, νοητ⋯ δ⋯ δι' ⋯στρολοϒίας in fr. 5 is not an explanatory clause by Sextus). The truth concerning the median heaven (though its celestial denizens – the sun, stars, and other planets – are no less objects of sense perception than the αἰσθητά of the sublunary world) is then ultimately a matter of intellection; it depends upon the calculations of mathematics, particularly astronomy. Cf. Theophrastus, , Met. 6b7–9Google Scholar; fr. 26 H = 100 IP; thereto, see Krāmer, H. J., Der Ursprung der Geistmetaphysik (Amsterdam, 1964), pp. 34f.Google Scholar, who adduces the often compared reference to Xenocrates in Arist, . Met. 1028^b24ffGoogle Scholar. (fr. 34 H = 103 IP) for the identification of ideas and mathematical; on this admittedly problematical issue in Xenocrates, see further, Heinze, pp. 49f.; Ross, W. D., Aristotle's Metaphysics (Oxford, 1924), lxxv, n. 2 and nnGoogle Scholar. ad 1028^b24, ^b26–7; ibid. and Fobes, F. H., Theophrastus Metaphysics (Oxford, 1929), pp. 56f.Google Scholar; Happ, H., Hyle (Berlin and New York, 1971), pp. 243, 245CrossRefGoogle Scholar; Cherniss, H., The Riddle of the Early Academy (Berkeley and Los Angeles, 1945), pp. 43f., 47f., 52, 58Google Scholar; ibid., Aristotle's Criticism of Plato and the Academy (Baltimore, 1944), pp. 208,196ff., 399,484; but cf. also the argument for mathematical as intermediates by Merlan, P., From Platonism to Neoplatonism ³ (The Hague, 1975), p. 42 nCrossRefGoogle Scholar. ** (sic), pp. 44ff., contra, e.g. Cherniss, , Aristotle's Criticism, p. 511Google Scholar; see further, Isnardi Parente, pp. 334f. The description of heaven as visible by sight but intelligible by astronomy implies in effect a distinction between observational and pure, i.e. mathematical astronomy, and brings to mind Plato's assertion that the true revolutions and inter-relations of celestial phenomena are ‘apprehended by reason and discursive thought, and not by sight’ (Rep. 529d; see Dicks, D. R., Early Greek Astronomy to Aristotle (London, 1970), pp. 104ff.Google Scholar, esp. p. 106).

²⁰ In the geocentric universe of the Greeks the earth occupies the centre of heaven (cf. e.g. Plato, , Phd. 108e5Google Scholar), which is tantamount to the centre of the cosmos, given the common identification of οὐρανός = κόσμος; cf. e.g. Plato, , Pit. 269d7–8Google Scholar, Ti. 28b 1–2, [PI, .] Epin. 977b2Google Scholar; see further, , Dicks, , op. cit. (n. 19), pp. 117f, 142Google Scholar; Mansfeld, , op. cit. (n. 8), p. 47Google Scholar. Cf. also [PI, .] Def. 411c5–6Google Scholar.

²¹ Plato's, pronouncement, Symp. 202eGoogle Scholar, on the intermediate state of the daemonic realm became virtually a theological dogma in subsequent doctrines of daemons. For Plato the median position of daemons is largely governed by their hermeneutic role (Symp. 202d–203a; cf. [PI, .] Epin. 984e–985bGoogle Scholar). On δαίμων and δαιμόνιον in Plato, see Friedländer, P., Platon i ² (Berlin, 1954), pp. 34–62, esp. 34–46Google Scholar. On Xenocrates' daemonology, see Heinze, pp. 78–123; cf. Burkert, W., Greek Religion, trans. Raffan, J. (Cambridge, Mass., 1985), pp. 329f., 332Google Scholar; Dome, , art. cit. (n. 2); 1524fGoogle Scholar; Dillon, , op. cit. (n. 10), pp. 31f.Google Scholar, Isnardi Parente, pp. 414–18. The main source for a reconstruction of Xenocrates' daemonology is Plutarch, from whose own doctrines Xenocrates' views must be carefully extricated; see nn. 54 and 56 below. On Plutarch, see Soury, G., La Démonologie de Plutarque (Paris, 1942)Google Scholar; Brenk, F. E., ‘“A Most Strange Doctrine”: Daimon in Plutarch’, CJ 69 (1973), l–llGoogle Scholar; ibid., In Mist Apparalled (Leiden, 1977), passim but esp. chs. 4 and 6; id., ‘In the Light of the Moon: Daemonology in the Early Imperial Period’, ANRW 16.3 (1986), 2117–30Google Scholar; ibid., ‘An Imperial Heritage: The Religious Spirit of Plutarch of Chaironea’, ANRW36A (1987), 275–94; Dillon, , op. cit. (n. 10), pp. 216–24Google Scholar.

²² , Plu.Def. or. 416c–dGoogle Scholar (fr. 23 H = 222 IP); cf. Soury, , op. cit. (n. 21), pp. 26, 35Google Scholar. The two-fold nature of Plutarch's daemons is bound up with the bipartition of the soul into rational and irrational parts – a conception, as we shall argue, that derives from Xenocrates. Cf. Babut, M. D. in Actes du VIII Congrès G. Budé (Paris, 1969), p. 531Google Scholar.

²³ , Plu.Def. or. 419aGoogle Scholar (fr. 24 H = 226 IP). According to Plutarch again (Is. et Os. 361b; fr. 25 H = 229 IP), Xenocrates characterized evil daemons as great and mighty natures in the atmosphere, but surly and sullen, who delighted in days of ill omen and those festivals that entailed beatings, lamentations, fastings, blasphemies, or obscenity (cf. Def. or. 417c and the discussion of both Plutarch passages by Soury, op. cit. (n. 21), pp. 5Iff.). Whatever the motivations at certain Greek festivals for the activities enumerated by Plutarch (on fasting, obscenity, and flagellation at e.g. the Thesmophoria, see Burkert, , op. cit. (n. 21), pp. 243–4Google Scholar; cf. pp. 104f.; Parke, H. W., Fetivals of the Athenians (London, 1977), p. 86Google Scholar; Soury, , op. cit. (n. 21), pp. 52f.Google Scholar), Xenocrates apparently viewed them as having an underlying apotropaic function: meeting with such activities, evil daemons ‘turn to nothing worse’ (in Def. or. 417c, eating raw flesh, tearing apart victims, fastings, and beating of the breast are likewise said to be practised δαιμόνων ϕαύλων ⋯ποrροπ⋯ς ἓνεκα; cf. n. 54 below). Daemons also rejoiced in ήμέραι ⋯ποϕράδες, most likely because of the impurity associated with these days: the annual Plynteria at which the garments of Athena were purified, and those days on which homicide cases were judged (see Mikalson, J. D., ‘Ήμέρα ⋯ποϕράς’, AJPh 96 (1975), 19–27Google Scholar; Soury, , op. cit. (n. 21), p. 52Google Scholar, should only be read in light of Mikalson's delimitation of ill-omened days). One should note further that Xenocrates distinguishes the practices favoured by evil daemons from the honours accorded the gods and ‘good daemons’ (δαιμόνων … χρηστ⋯ν, Is. et Os. 361b); here one may think of the libations to ‘good fortune’ (⋯ϒαθὡ δαίμονι), a custom that seems to have been invested with cultic dimensions (Burkert, , op. cit. (n. 21), p. 180 and n. 9Google Scholar; Rohde, E., Psyche i ³ [Tübingen and Leipzig, 1903], p. 254 n. 2Google Scholar), but other than this there is no evidence of cults relating specifically to daemons in the fourth century; cf. Mikalson, J. D., Athenian Popular Religion (Chapel Hill and London, 1983), p. 66Google Scholar. (A slightly different matter were the cults dedicated to the ‘averting gods’ – ⋯ποτρόπαιοι θεοί; see Herter, H., Rhein. Jb. f. Volkskunde 1 (1950), 136Google Scholar = Kl. Schr. (Munich, 1975), 68)Google Scholar. But leaving aside the question of cult practices associated with daemons, the distinction between good and bad daemons existed only as an inchoate notion before the fourth century, and Xenocrates is not unjustly credited with having brought into relief the evil daemonic powers. Although Plutarch, (Def. or. 419aGoogle Scholar) also cites Plato as a source for ϕαϋλοι δαίμονες, in the Platonic corpus daemons are consistently portrayed under a beneficent aspect; see e.g. Phd. 107d–e, Crat. 398b–c. The daemons of the Epinomis (984b ff.) possess a sympathetic knowledge of pain and pleasure, but have not yet attained the full participation in human affections as we see it in Xenocrates and Plutarch (see again Def. or. 416c–d (fr. 23 H = 222 IP) and n. 24 below; cf. Solmsen, F., ‘Antecedents of Aristotle's Psychology and Scale of Beings’, AJPh 76 (1955), 163Google Scholar; for a list of affinities between the author of the Epinomis and Xenocrates, in areas besides daemonology as well, see Kramer, , op. cit. (n. 10), p. 126 n. 88Google Scholar, but cf. also Tarán, L., Academical Plato, Philip of Opus, and the Pseudo-Platonic Epinomis (Philadelphia, 1975), p. 152)Google Scholar. Rep. 619c implies that τυχή and daemons are unjustly blamed for misfortunes. Xenocrates' rudimentary typology of daemons was of consequence for later refinements and schematizations in the doctrines of daemons, in which special emphasis and interest were given the sinister type. Cf. Burkert, , op. cit. (n. 21), pp. 179, 332Google Scholar, Herter, , op. cit., p. 141Google Scholar (= Kl. Schr. 73).

²⁴ …οὓς (i.e. δαίμονας) καì πλάτων καì πυθαϒόρας καì ὡενοκράτης καì Χρύσιππος ⋯πόμενοι τοȋς πάλαι θεολόϒοις ⋯ρρωμενεστέρους μ⋯ν ⋯νθρώπων ϒεϒονέναι καì πολὺ τῇ δυνάμει τ⋯ν ϕύσει ὑπερϕέροντας ⋯μ⋯ν, τò δ⋯ θεȋον οὐκ ⋯μιϒ⋯ς οὐὡ ἄκρατον ἓχοντας, ⋯λλ⋯ καì ψυχ⋯ς ϕύσει καì σώματος αἰσθήσει [⋯⋯] (⋯⋯ Ω: deest ap. Eus.: ἓ⋯ Griffiths) συνειληχòς ⋯δον⋯ν δεχομένους (–μένους Strijd: –μένην Ω: –μένῃ Griffiths, Sieveking corr. ex Eus.: –μενον Cherniss) καì πόνον καì ὃσα ταύταις ⋯πιϒενόμενα (ϒενόμενα Eus. corr. Pohlenz: ⋯ϒϒενόμενα Ω Griffiths) ταȋς μεταβολαȋς πάθη τοὺς μ⋯ν μᾱλλον τοῡς ὡ ⋯ττον ⋯πιταράττει ϒίνονται ϒ⋯ρ ώς ⋯⋯ ⋯νθρώποις καì ⋯ρετ⋯ς ⋯ρετ⋯ς διαϕοραì καì κακίας. Is. et Os. 360d–e (fr. 24 H = 225 IP); See Commentaries by Hopfner, T., Plutarch über Isis und Osiris, ii (Prague, 1941), pp. 112ff.Google Scholar, Griffiths, J. G., Plutarch's De Iside et Osiride (University of Wales Press, 1970, pp. 383ff.)Google Scholar. Cf. the similar description of daemons in Def. or. 416c: … ϕύσεις εἰσί τινες ὣσπερ ⋯⋯ μεθορίψ θε⋯ν καì ⋯νθρώπων δεχόμεναι πάθη καì μεταβολ⋯ς … For the meaning of the term μεταβολή, see n. 61 below. On Xenocrates as the dominant source for chs. 25 and 26 of Is. et Os. (360d ff.), see Heinze, p. 82 (Detienne, M., REA 60 (1958), pp. 272ffCrossRefGoogle Scholar, in contrast, argues that this and other passages assigned by Heinze to Xenocrates are largely derived from ancient Pythagorean daemonology; but cf. n. 55 below). Although the influence of Chrysippus should not be discounted, it remains uncertain whether he included among daemons the souls of the dead, whereas this was almost certainly the belief of Xenocrates (see Def. or. 417b, and n. 55 below); on Chrysippus see SVF. ii 1101–5; Pohlenz, M., Die Stoa, i (Göttingen, 1947), p. 96Google Scholar; ii (1949), p. 54.

²⁵ On the problems posed by the immaterial nature of daemons, cf. Isnardi Parente, pp. 417f. and see n. 100 below.

²⁶ See n. 23 above.

²⁷ Gr. aff. cur. 5. 19 p. 127 Raeder (Diels, , Dox. gr. 389f.Google Scholar; fr. 70 H = 206 IP).

²⁸ 35a, 41c–d, 42e–44d, 65a5, 69c–d, 72d. Cf. Graeser, A., Probleme der platonischen Seelenteilungslehre (Munich, 1969), p. 70Google Scholar; Robinson, T. M., Plato's Psychology (Toronto, 1970), pp. 71ffGoogle Scholar.

²⁹ Cf. 42a3–d2, 43a6–44bl, 69c6–d6, 70d7–71a7, 71d4, 77b (hence Philo of Alexandria, , Leg. all. 2.6Google Scholar, could simply state τ⋯ δ⋯ ἄλοϒον αἳσθησίς ⋯στι καì τ⋯ ταύτης ἔκϒονα πάθη).

³⁰ On the non-cognitive, non-propositional nature of αἵσθησις in the Timaeus, see Silverman, A., ‘Plato on Perception and “Commons”’, CQ 40 (1990), 149–58CrossRefGoogle Scholar.

³¹ Opinions and beliefs arise to the rational soul whenever it is concerned with the sensible (37b6–8, cf. Graeser, , op. tit. (n. 28), p. 69Google Scholar), but the rational soul never engages directly in sense perception; although it may form opinions and beliefs about what is sensed, the sensations themselves are simply material movements that terminate in the mortal, irrational soul (69c5–d6 in conjunction with 42e5–44c4).

³² See esp. 45b2–47e2, 61c3–68d7. The only comparable account of sense perception is found in the Theaetetus; thereto, Silverman, , art. cit. (n. 30), 158–75, esp. 158–63Google Scholar.

³³ The question of the lower soul's mortality, which we will take up below, was quite possibly a point of divergence.

³⁴ On the three parts see 44d5, 69d6–70d7, 73d, 87a2–3, 89e4–5, 90a4–5; cf. Graeser, , op. cit.(n. 28), p. 70Google Scholar; Robinson, , op. cit. (n. 28), pp. 105ff.Google Scholar, 119ff; Rees, , art. cit. (n. 6), 113Google Scholar; Dirlmeier, F., Aristoteles Nikomachische Ethik (Berlin, 1960), p. 279Google Scholar; on the ‘psycho-physical system’ in the Timaeus, see Solmsen, , art. cit. (n. 23), 153–7Google Scholar.

³⁵ Isnardi Parente, p. 398, points to the possible Pythagorean origins for the Platonic contrast of reason versus passion.

³⁶ Aristoteles, Magna Moralia (Berlin, 1958), pp. 163–5Google Scholar; op. cit. (n. 34), pp. 278–9, 292–3.

³⁷ Cf. e.g. Iamb, . Protr. 34, 12–15Google Scholar Pistelli = fr. 23 During; 41, 2 (0, p. 35 Ross = fr. 6, pp. 33f.) Walzer = fr. 60 During; see further Rees, D. A., ‘Theories of the Soul in the Early Aristotle’, in Düring, I. and Owen, G. E. L. (eds.), Aristotle and Plato in the Mid-fourth Century (Gothenburg, 1960), pp. 195–9Google Scholar (where at p. 195 n. 7, ‘Eud. fr. 6’ should read ‘Protr. fr. 6’) and Hardie, W. F. R., Aristotle's Ethical Theory ² (Oxford, 1980), pp. 218ffCrossRefGoogle Scholar.

³⁸ ‘The Peripatetic Interpretation of Plato's Tripartite Psychology’, GRBS 26 (1985), 283–302.Google Scholar On De an. 432^a24–b7, see ibid., ‘Aristotle's Criticism of Soul-Division’, AJPh 108 (1987), 627–43.

³⁹ See Graeser, , op. cit. (n. 28), pp. 102ff.Google Scholar; cf. Rees, , art. cit. (n. 6), 115fGoogle Scholar.

⁴⁰ On the consequent influence of Peripatetic repartition o n the doxographical tradition, see Waerdt, P. A. Vander, ‘Peripatetic Soul-Division, Posidonius, and Middle Platonic Moral Psychology’, GRBS 26 (1985), 373–94Google Scholar.

⁴¹ In this, I think, lies the lesson of Vander Waerdt's article for students of Plato. For a general understanding of the soul in Plato as a ‘differentiated unity’, see Hall, R. W., Plato and the Individual (The Hague, 1963), pp. 141–62Google Scholar.

⁴² Gr., aff. cur. 5.19, p. 127Google Scholar Raeder (Diels, , Dox. gr. 389Google Scholar): see p. 149 above. Theodoret's source for the Platonic bipartition and the further subdivision into spirited and appetitive parts is Aët. 4.4.1 (Diels, , Dox. gr. 389f. = psGoogle Scholar. Plut, . Plac. 898eGoogle Scholar = Eus, ., PE 15. 60. 1–4Google Scholar; cf. Plut, . Virt. mor. 442aGoogle Scholar); cf. Rees, , art. cit. (n. 6), 114Google Scholar; Dirlmeier, , op. cit. (n. 36), p. 165Google Scholar; Waerdt, Vander, ‘Peripatetic Soul-Division’ (n. 40), 375, 379fGoogle Scholar. (For his recapitulation of the strife –ἔρις– about the soul among pagan philosophers, Theodoretus explicitly lists Plutarch and Aëtius, along with Porphyry, as his sources: Gr. aff. cur. 5. 16, cf. 2.95, 4.31; unnamed for this section are Clement and Eusebius. On Theodoretus' use of source texts see Canivet, P., (ed.), Théodoret de Cyr, Thérapeutique des maladies helléniques (Paris, 1958), pp. 55–9Google Scholar; Places, É. Des, ‘Le Platon de Théodoret:les citations de Lois etde l'Epinomis’, REG 68 [1955], 171–84CrossRefGoogle Scholar = Etudes platoniciennes (Leiden, 1981), pp. 229–42)Google Scholar. On the appearance of Pythagoras with Plato, see Waerdt, Vander, ‘Peripatetic Soul-Division’ (n. 40), 376f., 387ff., 391ffGoogle Scholar.

⁴³ So Heinze, Burnet, and others; see Dirlmeier, , op. cit. (n. 36), p. 165Google Scholar; op. cit. (n. 34), pp. 274–5.

⁴⁴ See n. 37 above. This is no t to deny that Aristotle later, in the ‘new psychology’ of the De Anima, discounted the usefulness of speaking of the soul's rational an d irrational parts; see Rist, J. M., The Mind of Aristotle (Toronto, 1989), pp. 183–4Google Scholar. Even here in Bk. 1 of the Me. Ethics (1102^a28–32), Aristotle goes on to say that the question of the divisibility of the soul, whether understood as real or logical, makes no difference to his argument. On the ⋯ξωτερικοì λόϒοι see Jaeger, W., Aristotle, tr. Robinson, R. (Oxford, 1948 ²), pp. 246–58Google Scholar; Waerdt, Vander, ‘The Peripatetic Interpretation’ (n. 38), 283 n. 2; on E.N. 1102^a26–8Google Scholar in particular, see Hardie, , op. cit. (n. 37), pp. 68ff.Google Scholar; cf. Rees, , art. cit. (n. 6), 118Google Scholar; Graeser, , op. cit. (n. 28), p. 70 n. 6Google Scholar.

⁴⁵ Porph, . In harm. pp. 30ffGoogle Scholar. During (fr. 9 H = 87 IP) offers a minor instance illustrating the turmoil as well as the deceptiveness of sensations, especially in regard to the hearing of musical notes. The senses, being conditioned by movements (one thinks of Plato's kinetic theory of sensations), are ever ⋯⋯ ταράχῃ and consequently fail to achieve accuracy (τ⋯ ⋯κριβές) in their perceptions (p. 32, 23f. Düring, , cf. p. 31, 6Google Scholar; for Platonic parallels on ταραχή, see n. 46 below); cf. Isnardi Parente, p. 318; Heinze, pp. 5–10. Having adduced this section of the In Ptolemaei harmonica it is only fair to point out that its trustworthiness as a witness to Xenocratean doctrine is far from certain: see Gottschalk, H. B., ‘The De Audibilibus and Peripatetic Acoustics’, Hermes 96 (1968), 450–2Google Scholar, but also now Isnardi Parente, pp. 314–19. Cf. Gaiser, K., Platons mgeschriebene Lehre (Stuttgart, 1963), p. 556 n. ad 72Google Scholar.

⁴⁶ Plato frequently describes the turmoil (ταραχή) wrought by the sensations and desires that affect the soul as a result of its entanglement in matter. A few selections suffice to illustrate what is almost a commonplace in Plato (and in later philosophers). In the analogy of soul with political tyranny in Republic 9 (577d–578a), the soul that is tyrannized by desires and passions loses its power of choice and is filled with turmoil (ταραχ⋯ς, 577e2) and regret. In the treatment of sensory illusions in Bk. 10 (602cff.), the soul that is subject to errors of vision is full of turmoil of every kind (π⋯σά τις γαραχή, 602c 12) which must be corrected by the (calculating) function of reason (το⋯ … ἒρϒον, 602el; cf. earlier in Bk. 7 (524a–b), where the ⋯πορία of the soul about sensations is resolved by the aid of λοϒισμός and νόησις see further, Graeser, op. tit. (n. 28), pp. 102f.). In the Phaedo (66dff.), the contemplation of truth ultimately demands a more extreme remedy than the harmonizing of soul parts (which is the general prescription of the Republic, cf. e.g. 572a); pure knowledge depends upon the total separation of the soul from the body, for even the philosophical life may still be beset by tumult and turmoil (θόρυβον … καì ταραχήν, 66d6) arising from the body (and its sensations, cf. 79c). As a last example, one might consider the course of reincarnation described in the Timaeus (42b–dl). The return to the soul's original, unfallen state requires the domination by reason (λόϒψ– pace Taylor) of the tumultuous and irrational mass (ὂχλον … θορυβώδη καì ἃλοϒον) that clings to the soul from the four material elements (cf. 43b–c).

⁴⁷ Art. cit. (n. 6), 118. Yet as Rees' article itself makes clear, the division of the soul in the early Academy an d in the doxographical reports is a n essentially bipartite one, such as we reviewed above.

⁴⁸ 373f. Cf. Waerdt, Vander, ‘The Peripatetic Interpretation’ (n. 38), 285 n. 5Google Scholar.

⁴⁹ See Wilamowitz, , Der Glaube der Hellenen i (Berlin, 1931), pp. 368f.Google Scholar; Dodds, E. R., The Greeks and the Irrational (Berkeley, 1951), p. 42Google Scholar; Burkert, , op. cit. (n. 21), p. 181Google Scholar.

⁵⁰ Arist, . Top. 112^a37–8Google Scholar (fr. 81 H = 236 IP; Heinze, p. 144, takes the ϒάρ clause as Aristotle's addition, though I rather think it belongs t o Xenocrates; so apparently also Cherniss, , Aristotle's Criticism (n. 19), p. 14Google Scholar, and already Krische, op. cit. (n. 15), p. 321; cf. Isnardi Parente, p. 421); cf. fr. 83 H = 239 IP. This excerpt along with Top. 152²7–9 (fr. 82 H = 240 IP) most likely comes from one of Xenocrates' many ethical treatises (the obvious choice would be the two books of the περì εὐδαιμονίας, D.L. 4.12). On the critical context in which Aristotle cites Xenocrates in the Topics, see Cherniss, op. cit. pp. 13f. See further, Krämer, op. cit. (n. 10), p. 172 n. 278; Brenk, , In Mist Apparalled (n. 21), p. 93n. 9Google Scholar.

⁵¹ Cf. Herter, , op. cit. (n. 23), p. 139Google Scholar ( = Kl. Schr. 71) with literature in n. 100; Rohde, op. cit. (n. 23), ii. 316 n. 1; Daniélou, J., ‘Démon, démonologie platonicienne et néo-platonicienne’, Dictionnaire de spiritualité, iii (Paris, 1957), col. 153Google Scholar. In the background of Xenocrates' mind may lie no t only Empedocles DK 31 B 115 (see Barnes, J., The Presocratic Philosophers, ii (London, 1979), pp. 196–9)Google Scholar, but also Democritus DK 68 B 170 and 171 (on the context of B 171 in Democritus' ethics, see Kahn, C. H., ‘Democritus an d the Origins of Moral Psychology’, AJPh 106 (1985), 12ff.)Google Scholar. In Plato's, Tim. 90a3, c5Google Scholar, the God-given daemon is the divine part of the soul residing in the head (cf. Graeser, , op. cit. (n. 28), p. 86Google Scholar; Dodds, , op. cit. (n. 49), p. 213 n. 31Google Scholar). Xenocrates differs in so far as he seems to equate the daemon with the soul in toto.

⁵² The belief that exceptional mortals became δαίμονες after death crops up repeatedly in Greek literature, beginning with Hesiod, (Erga 122–6)Google Scholar, while the dead in general are referred to as daemons in sepulchral inscriptions of the Hellenistic period. See Burkert, , op. cit. (n. 21), p. 181Google Scholar; cf. Wilamowitz, , op. cit. (n. 49), i. 366Google Scholar; Roscher, , s.v. ‘Daimon’ (i. 938)Google Scholar; Herter, , op. cit. (n. 23), pp. 139fGoogle Scholar. (= Kl. Schr. 70f.); Horsley, G. H. R., New Documents Illustrating Early Christianity, ii (1982), p. 51Google Scholar. On cults of the dead seen as cults to daemons, see Rohde, , op. cit. (n. 23), i. 246ffGoogle Scholar. Damascius rejected as un-Platonic the notion that soul would pass into a daemon, angel, or god (In Phd. 175 [ad69e2], p. 107 Westerink (n. 4); see Westerink's note ad loc.). About Xenocrates' daemonology, however, one may still wonder why he retained, in addition to the daemon of departed souls, those traditional types of daemons who never were human souls; cf. Brenk, , In Mist Apparalled (n. 21), p. 93Google Scholar. The same question applies to Plutarch's daemons; see Dillon, op. cit. (n. 10), pp. 223–4. In the later daemonology of Proclus, daemons by nature (ϕύσει) are ontologically distinguished from daemons by relation (σχέσει); see O'Meara, D. J., Pythagoras Revived (Oxford, 1989), p. 151Google Scholar.

⁵³ Since πρίττωμα generally designates a product or residuum of the body, such as an excretion or a secretion, Plutarch's use of it here to describe the post-mortem soul makes for a particularly suggestive simile.

⁵⁴ Fr. 24 H = 227 IP. Heinze, pp. 81ff., has shown that chs. 13–15 (416d–418a) of the Def. or. derive from Xenocrates. Jones, R. M., The Platonism of Plutarch (diss. Chicago, 1913Google Scholar; repr. New York and London, 1980), p. 29, cautiously concurs with Heinze on this point. The salient factors seem to me to be these: 1. the explanation in Def. or. 417b of the differences in virtue that obtain among daemons as among men (εἰσì ϒάρ, ὡς ⋯⋯ ⋯νθρώποις, καì δαίμοσιν ⋯ρετ⋯ς διαϕοραί) in terms of the different degrees of intensity by which affective and irrational elements still cling to daemons, is a variation of the similar theory expounded in Is. et Os. 360e, which is part of a Xenocratean section (see n. 24 above): the daemons' participation in the sensory human soul and their resultant susceptibility to affections explain the different degrees of virtue and vice (ϒίνονται ϒ⋯ρ ὡς ⋯⋯ ⋯νθρώποις καì δαίμοσιν ⋯ρετ⋯ς διαϕοραì καì κακίας); 2. Plutarch's closing remark to this passage, ‘there are in many places sacrifices, rituals, and myths (μυθολοϒίαι) that preserve and carefully maintain the scattered traces and signs of these things’ (i.e. the involvement of daemons in sacred ceremonies and mysteries (417a) and the nature of daemons, just described), recalls the quasi-theodicy of Is. et Os. chs. 25 and 26 (360dff.); there it seems to be largely Xenocrates' interpretation that the stories about Typhon, Osiris, and Isis, about the Giants and Titans, the lawless deeds of Kronus, etc., which are told in myth (μνθολοϒούμενα), as well as those things veiled in mysteries and rites, pertain to the experiences (παθήματα) of neither gods nor men but daemons; 3. the connection between these sections of the Def. or. and Is. et Os. and their common source in Xenocrates finds as close a confirmation as one could demand in Def. or. 417c: the mysteries reveal the truth about daemons, and the festivals associated with the eating of raw flesh, the rending of victims, fasting, beating of the breast, and profanity are not rites performed for gods but for the turning away of evil daemons. The apotropaic function of practices such as these is expressly cited as part of Xenocrates', daemonology in Is. et Os. 361bGoogle Scholar; see n. 23 above; cf. Soury, , op. cit. (n. 21), pp. 51, 55Google Scholar.

⁵⁵ Cf. Heinze, , p. 83Google Scholar, on Def. or. 417b: ‘Daraus folgt doch wohl, daB die Dämonen einst Menschenseelen waren: denn ein λείψανον kann das Unvernünftige nur aus dem irdischen Leben der Seele sein’ (this, incidentally, seems also to have been Werner Jaeger's conviction; in his copy of Heinze's Xenokrates, which I am here using, he crossed out ‘doch wohl’ and inserted ‘sicherlich’ in the margin); cf. also ter Vrugt-Lentz, J. in Reallexikon für Antike und Christentum, ix, col. 614Google Scholar (s.v. ‘Geister [Dämonen]: B. II. Vorhellenistisches Griechenland’); similarly, Dillon, , op. cit. (n. 10), p. 223Google Scholar. For opposing views, see von Arnim, H., Plutarch über Dämonen und Mantik (Amsterdam, 1921), pp. 62f.Google Scholar; Eisele, T., AGPh 17 (1904), 43f.Google Scholar; Brenk, , ‘In the Light of the Moon’ (n. 21), 2126Google Scholar (though in ‘An Imperial Heritage’ (n. 21) Brenk says that Xenocrates may have viewed daemons as disembodied human souls). According to Isnardi Parente, p. 418, Xenocrates' daemon-soul represents a juxtaposition (rather than ‘una reale sutura’) of the cosmological daemonology of Philip of Opus and the psychological, religious daemonology of a more traditional Pythagorean–Platonic type. The Pythagorean influences upon both Plato's and Xenocrates' daemonology are neglected by Heinze, but in correcting this omission, Detienne, M. (La Notion de daimon dans le pythagorisme ancien (Paris, 1963)Google Scholar, cf. ibid. ‘Xénocrate et la demonologie pythagoricienne’, REA 60 (1958), 271–9CrossRefGoogle Scholar) has overemphasized the Pythagorean (and often pseudo-Pythagorean) origins of the belief that human souls are daemons and become lunary daemons after death; he consequently minimizes the important role of Xenocrates in systematizing and incorporating daemons within the hierarchy of Academic ontology; see Burkert, W., Gnomon 36 (1964), 563–7Google Scholar.

⁵⁶ See Heinze, pp. 123–47. Besides using the myth of the Fac. lun., Heinze, , pp. 102f., 128ff.Google Scholar, also draws on the Timarchus myth in Gen. Socr. ch. 22 (590bff.). In both texts Heinze believed he was able to separate neatly the Platonic, Xenocratean, and Poseidonian sources, but it has been shown, notably by Jones, op. cit. (n. 54), pp. 30–3, 48–63 (cf. ibid.), ‘Posidonius and Solar Eschatology’, CP 27 [1932], 113–35; reprinted in Platonism of Plutarch (n. 54), that although these two accounts evidence some Stoic influences, they are primarily based upon Plato, especially the Timaeus. See further Cherniss's, H. survey in his introduction to the Loeb edn. of the Fac. lun., Plutarch's Moralia, xii (Cambridge and London, 1957), pp. 22ff.Google Scholar; Görgemanns, H., Untersuchungen zu Plutarch's Dialog De facie in orbe lunae (Heidelberg, 1970), p. 80 n. 117Google Scholar. Cf. n. 57 below and Appendix.

⁵⁷ The body-soul-intellect schematization ultimately goes back to Plato; see Cherniss, , op. cit. (n. 56), n. c, p. 197Google Scholar; Pepin, J., Idées grecques sur l'homme et sur Dieu (Paris, 1971), p. 94 n. 1Google Scholar; Deuse, , op. cit. (n. 1), pp. 45fGoogle Scholar. Cf. the myth, in Gen. Socr. 591d–eGoogle Scholar (on problems of consistency in Plutarch's views of nous in Gen. Socr. and elsewhere, see Deuse, , op. cit. (n. 1), pp. 46f.Google Scholar; see also Brenk, , ‘An Imperial Heritage’ (n. 21), 530–2Google Scholar; Dillon, J. M., ‘“Orthodoxy” and “Eclecticism”: Middle Platonists and Neo-Pythagoreans’ in Dillon, J. M. and Long, A. A. (eds.), The Question of ‘Eclecticism’ (Berkeley, 1988), pp. 109–13)Google Scholar. See also Appendix.

⁵⁸ On ‘the fields of Hades’ as a purgatory, see Soury, , op. cit. (n. 21), pp. 187f.Google Scholar, Brenk, , ‘An Imperial Heritage’ (n. 21), 284Google Scholar. Plutarch does not specify what sort of punishment unjust and intemperate souls undergo; reincarnation seems to be the fate of the souls who reach the moon but fall again into the deep, while those who cling unsuccessfully to the moon are swept once more into Hades (943d, cf.Gen. Socr. 591c). Reincarnation appears again, as we shall observe shortly, as a punishment at the later judgment of souls after they have become daemons. Soury, , op. cit. (n. 21), pp. 190ffGoogle Scholar. has tried to categorize three different classes of souls and their respective fates.

⁵⁹ 943f–944a (fr. 56 H = 161 IP); see n. 16 above. Although Plutarch adduces Xenocrates, his account of the constituent elements of the physical bodies is not in all points identical with Xenocrates'; see Arnim, v., op. cit. (n. 55), pp. 53fGoogle Scholar.

⁶⁰ The judgment takes place in the ‘gulf of Hecate’, which Plutarch describes a little earlier as part of the topographical description of the moon (944c). Here Plutarch also makes clear that souls render or demand satisfaction for what they have suffered or done only after having become daemons (ἢδη ϒεϒενημέναι δαίμονες). Cf. Cherniss, , op. cit. (n. 56), n. a, p. 210Google Scholar.

⁶¹ Plutarch also uses the term μεταβολή for the transformation of daemons (or mortals) to gods; see Pelop. 16.8, Def. or. 415b; Aujoulat, N., Le Néo-platonisme Alexandrin, Hiéroclès d'Alexandrie (Leiden, 1986), p. 180CrossRefGoogle Scholar. Whether he also meant it to refer to the change of daemons to men, i.e. reincarnation (so Heinze, pp. 110, 138), is less certain. Notwithstanding Def. or. 438d, where Plutarch refers to some (unknown) thinkers who hold that the things above the moon undergo μεταβολαί and παλιϒϒενεσίαι, in the two passages where μεταβολαί might be thought to refer to the incarnation of daemons, Is. et Os. 360e and Def. or. 416c (both quoted above n. 24), the changes seem rather to indicate those wrought in the psyche by the vicissitudinous nature of affections such as pleasure and pain, to which the soul is subject as long as it is involved with matter. But once daemons achieve the final transformation to gods, they are relieved of any further changes because, as Plutarch says elsewhere, the divine, being immaterial, does not undergo constraints, chances, and changes (⋯νάϒκαι, τύχαι, μεταβολαί); see Ad princ. inerud. 781f; cf.E ap. Delph. 394a, Quaes. conviv. 717e, An. proc. 1015e.

⁶² The closing part of the myth (945c), describing the function of the three Fates and their relation to sun, moon, and earth/ may be an adaptation from Xenocrates (n. 17 above); see Dorrie, , op. cit. (n. 4), pp. 291–3Google Scholar; cf. Cherniss, , op. cit. (n. 56), n. b, p. 221Google Scholar.

⁶³ The idea of a celestial Hades owes its increasing prominence in the Hellenistic period to advances in astronomy coupled with beliefs in astral immortality. The acceptance of the sphericity of the earth and the divine nature of the heavenly bodies, which revolved in orderly fashion around the earth (the Pythagorean/Philolaic source of these ideas, as set forth by Dicks, , op. cit. (n. 16), pp. 72f.Google Scholar, is disputed; see Morrison, J. S., CR 21 (1971), 227Google Scholar; Burkert, , op. cit. (n. 8), pp. 337–50Google Scholar), gave scientific credence, as it were, to the notion that perfection resides ‘above’ and would have largely supplanted the old picture of a subterrestial Hades as the way station or destination of departed souls. See Burkert, ibid. pp. 357–68; cf. Buffière, F., Les Mythes d'Homère et la pensée grecque (Paris, 1956), pp. 496ffGoogle Scholar. But it might also be observed that to the extent that Xenocrates makes earth and water part of the sublunary realm and accessible to daemons, our world here below is already Hades; cf. Boyancé, P., ‘La Religion astrale de Platon a Cicéron’, REG 65 (1952), 334Google Scholar.

⁶⁴ Already Heraclides Ponticus spoke of a heavenly Hades, the Milky Way, in which discarnate souls tarry between incarnations or through which they pass in their ascent to the higher gods (Heraclides may also have posited a sublunary Hades for the punishment of the wicked); see frr. 96 and 97 Wehrli; Burkert, op. cit. (n. 8), p. 367; Gottschalk, H. B., Heraclides ofPontus (Oxford, 1980), pp. 100ffGoogle Scholar.

⁶⁵ Cf.Gen. Socr. 592b–c; Virt. mor. 445b.

⁶⁶ Heinze, p. 139 (cf. p. 95), maintained that Xenocrates, as a pupil of Plato, would not have let souls become daemons right upon death, for they would first have to be punished for their earthly sins in the aer between earth and moon. Their punishment could not consist of rebirth, since then the process of reincarnation would only come to an end with the complete purification of the soul, and the evil passions of daemons would be inexplicable. This line of reasoning, however, is not necessarily convincing and appears to me at variance with some of Heinze's own postulates derived from Plato. We do not know what motivation Xenocrates gave for the soul's first incarnation, but we may assume with Heinze (pp. 145f.) that embodiment was a matter of fate and necessity for every soul, as in the Timaeus (41e2–42a4), and not the result of some misfortune or sin incurred by certain souls in their pre-existent state, as e.g. in the Phaedrus (248c). Xenocrates could not have adopted the latter alternative, since all daemons, if our reconstruction of his daemonology is correct, retain some share of the nature of the human soul. What then should stand in the way of souls from becoming daemons upon death? The evil passions of certain daemons result from the stronger remnant of the sensitive and irrational element, which would in turn render them subject to reincarnation, though each sojourn upon earth would also provide an opportunity to bring the lower elements of the soul under control of reason and thus bring about an eventual release from the cycle of rebirth (cf. Tim. 42b2–dl). Good daemon-souls would never have to undergo reincarnation in the first place.

⁶⁷ See Appendix.

⁶⁸ A doxographical snippet carries the same import: πυθαϒόρας Ἀναξαϒόρας πλάτων ενοκράτης Κλεαθες θύραθεν εìσκρίνεσθαι τòν νοȗν Aët. 4.5.11 (Diels, , Dox. gr. 392Google Scholar); fr. 69 H = 205 IP. Cf. Arist, . G.A. 736^b27Google Scholar; thereto, Ross, D. W. (ed.), Aristotle, De Anima (Oxford, 1961), p. 42Google Scholar; Guthrie, W. K. C., ‘Plato's Views on the Nature of the Soul’, Recherches sur la tradition platonicienne (Entretiens sur l'antiquité classique, iii. Vandoevres-Geneva, 1955), 19Google Scholar.

⁶⁹ As Plutarch puts it (Fac. lun. 945a–b), the process of dissolution is quick for temperate souls who have loved a leisurely, quiet, and philosophical life, whereas ambitious and active souls are drawn by their restlessness and fondness for the body to another birth. The moon tries to restrain this latter class of souls in order to prevent them from being reborn as mindless monsters.

⁷⁰ See Gallop (n. 14), ad loc. (p. 143).

⁷¹ Plato often emphasizes these attributes of the soul in conjunction with its being μονοειδής (see e.g. Phd. 80b), which, according to Graeser, , op. cit. (n. 28), pp. 58fGoogle Scholar. (with further references), signifies the soul's affinity to the uniform world of Ideas.

⁷² Cf. Heinze, pp. 140–1; Merlan, P. in Armstrong, A. H. (ed.), The Cambridge History of Later Greek and Early Medieval Philosophy (Cambridge, 1970), p. 28Google Scholar; Robinson, , op. cit. (n. 28), p. 131Google Scholar.

⁷³ Σπεύσιππος … ενοκράτην ⋯⋯ αὐτο⋯ κατέστησε τ⋯ν πλατωνικ⋯⋯ δοϒμάτων ⋯ξηϒητήν. Galen, ps.-, Hist, philos. 3Google Scholar (Diels, , Dox. gr. 599, 14–17Google Scholar) = T 12 Taran (see also Taran's commt., p. 211); ⋯ ενοκράτης ὑπεραπολοϒούμενος το⋯ πλάτωνος … ps.-Alex. Aphr. In Arist. Met. N. 4 1091^a27, p. 819, 37 Hayduck (fr. 33 H = fr. 116 IP). But cf. Numenius' evaluation of the Early Academy and especially of Xenocrates in Eus, . PE 14. 4. 16ff. = frGoogle Scholar. 24 Des Places (Numénius, Fragments (Paris, 1973)Google Scholar), on which see Taran, p. 215. On Xenocrates' Timaeus exegesis and its possible influence upon Middleplatonic tradition, cf. Krämer, H. J., ‘Zur geschichtlichen Stellung der aristotelischen Metaphysik’, Kant Stud. 58 (1967), 328ff.Google Scholar, id., ‘Grundfragen der aristotelischen Theologie’, Th & Ph 44 (1969), 484ff.

⁷⁴ See Dörri, , op. cit. (n. 4), p. 173Google Scholar; Dillon, , op. cit. (n. 10), pp. 55ffGoogle Scholar.

⁷⁵ It is far from certain how much of the renewal of Platonism, save for the break from Academic Scepticism, can be traced directly to Antiochus. The valuation of Antiochus is controversial in three regards: 1. his status as a bona fide Platonist; 2. his own use of Plato's writings; 3. his influence upon subsequent developments in Platonism. See Dörrie, , op. cit. (n. 4), pp. 172–4Google Scholar; id., Von Platon zum Platonismus (Opladen, 1976), pp. 14–15,46; further literature in Deuse, , op. cit. (n. 1), p. 8 n. 2Google Scholar. See also now Barnes, J., ‘Antiochus of Ascalon’ in Griffin, M. and Barnes, J. (eds.), Philosophia Togata (Oxford, 1989), pp. 51–96Google Scholar.

⁷⁶ Dörrie, , op. cit. (n. 75), pp. 32ff.Google Scholar, explains the impulses that governed the revived interest in the Timaeus.

⁷⁷ So Proclus, (In Tim. i.75.30–76.2Google Scholar Diehl), who calls Crantor ⋯ξηϒητής (but cf. ps. Galen, n. 73 above).

⁷⁸ The evidence for Eudorus' comments on the Timaeus derives mainly from references in Plutarch's An. proc. On the gap in Timaeus commentaries between Crantor and Eudorus, see M. Baltes (comm.), Timaios Lokros, Über die Natur des Kosmos und der Seele (Leiden, 1972), p. 26 n. 1Google Scholar.

⁷⁹ Dörrie, , op. cit. (n. 4), pp. 174f.Google Scholar, and in detail, Baltes, M., Die Weltenstehung desplatonischen Timaios nach den antiken Interpreten i (Leiden, 1976), pp. 231ffCrossRefGoogle Scholar.

⁸⁰ Philo, , De spec. leg. 1. 333Google Scholar et al.; see further, Dillon, , op. cit. (n. 10), pp. 174ff.Google Scholar; Tim. Locr. De natura mundi et animae (ed. Marg, W. (Leiden, 1972), 218. 5fGoogle Scholar. (also see Baltes, op. cit. (n. 78), ad loc.), cf. 224.3f. Baltes, ibid. 20ff., esp. 25, considers the Timaeus Locrus a product of the school of Eudorus, but also admits the possibility of its dating to the first century A.D.; cf. Dillon, , op. cit. (n. 10), 131Google Scholar; Tobin, T. H. (text, trans., and notes), Timaios ofLocri, On the Nature of the World and the Soul (Chico, 1985), 7Google Scholar.

⁸¹ For Alcinoos' argument on the immortality of the soul see ch. 25 of the Didaskalikos (= 177–8 in the Bude edition of J. Whittaker (Paris, 1990), who for reference retains Hermann's pagination but uses his own line numbers). Alcinoos considers irrational souls mortal and perishable since they are altogether without a conception of the noetic nature and even of a different οὐσία than rational souls (178, 30–2 Whittaker). See further, Deuse, , op. cit. (n. 1), pp. 91–5Google Scholar. On the controversy of the name, Alcinoos or Albinus, I tentatively side with Whittaker for Alcinoos and a date for the Didaskalikos anywhere between the reign of Augustus and the middle of the second century (Whittaker, , op. cit., pp. vii–xiiiGoogle Scholar).

⁸² 178, 26 Whittaker.

⁸³ Eus, . PE 15. 9. 1–2Google Scholar = fr. 7 Des Places. Here Atticus also reveals that there had arisen a great deal of ambitious rivalry (πολλ⋯ … ⋯ ϕιλοτιμία) among Plato's disciples in the contest to defend his dogma. Cf. n. 1 above.

⁸⁴ See Proclus, , In Tim. iii. 234, 8–18 DiehlGoogle Scholar. On this important passage, which also mentions Albinus, see n. 95 below. Cf. Festugière, A. -J. (trans, and notes), Proclus: commentaire sur le Timée (Paris, 1968)Google Scholar, v. 99 n. 1.

⁸⁵ This is from the oft-cited reference in Hermeias, , In Plat. Phdr. p. 102, 13–14 CouvreurGoogle Scholar; thereto see Dillon, J., ‘Harpocration's Commentary on Plato; Fragments of a Middle Platonic Commentary’, CSCA 4 (1971)Google Scholar, 139ff.

⁸⁶ See Dillon, , op. cit. (n. 10), pp. 258ffGoogle Scholar.

⁸⁷ Fr. 52, 64–74 Des Places; see Deuse, , op. cit. (n. 1), pp. 67fGoogle Scholar.

⁸⁸ Fr. 43 in conjunction with frr. 49 and 52 Des Places; see Deuse, , op. cit. (n. 1), pp. 77ffGoogle Scholar. Numenius' theory of elements that accrue to the descending soul from matter (προσϕυομένων … ⋯πò … τ⋯ς ὓλης, fr. 43,8–9) harbingers the idea of the vehicle of the soul which was developed into an elaborate doctrine in Neoplatonism; cf. Des Places, op. cit. (n. 73), p. 122 n. 3 (ad fr. 43); Dillon, , op. cit. (n. 10), p. 376Google Scholar; see further, n. 92 below.

⁸⁹ Plants and animals would in any case be included. Dörrie, , op. cit. (n. 4), p. 426Google Scholar adduces as parallel fr. 53 Des Places in which Numenius says that Sarapis (as a type of universal deity) shares in the being of all the animals and plants cared for by nature. See further Deuse, , op. cit.(n. 1), pp. 77–9Google Scholar, with a survey of the various interpretations given ⋯μμψύχου ἓξις at p. 77 n. 49.

⁹⁰ As Dörrie, , op. cit. (n. 4), p. 426Google Scholar, says about the doxographical list in Damascius: ‘Dabei ist die Anordnung so getroffen, daB diejenigen, die der Unsterblichkeit den weitesten Raum geben, zuerst genannt werden; dieser Rahmen verengt sich dann von Position zu Position.’

⁹¹ Fr. 35, 21–3 Des Places. For ⋯ ⋯⋯ ⋯νδράσι ζωή = ἄλοϒος ψυχή Des Places, op. cit. (n. 73), p. 119 n. 7 (ad. fr. 35), refers to Leemans, E.-A., Studie over den wijsgeer Numenius van Apamea mil uitgave der fragmenten (Brussels, 1937), p. 101 n. 4Google Scholar, who in turn deduces this meaning from Porphyry's summary (fr. 44 Des Places) of the contrast between the two souls in Numenius. Cf. also Proclus, , In Tim. iii. 234, lOffGoogle Scholar. Diehl. Concerning the fate of the irrational soul, Dillon, , op. cit. (n. 10), p. 260Google Scholar, suggests ‘presumably it survived indefinitely in the sublunar realm, constantly subjected to recycling in connexion with rebirth’. Deuse, , op. cit. (n. 1), p. 79Google Scholar, however, in light of his preceding discussion of the world-souls in Numenius, argues more specifically that after the rational soul ascends to its divine source, the irrational soul, having lost its locus of existence, becomes one with the irrational psychic forces that originate from the interaction of the world-soul and the evil world-soul.

⁹² For a comprehensive survey, Dodds, E. R. (ed.), Proclus, The Elements of Theology ² (Oxford, 1963), pp. 313–21Google Scholar, is still standard; see also Finamore, J. F., Iamblichus and the Theory of the Vehicle of the Soul (Chico, 1985), pp. 1–6Google Scholar, with further literature at p. 7, n. 1; Deuse, , op. cit. (n. 1), p. 218 n. 331Google Scholar; cf. Geudtner, O., Die Seelenlehre der chaldäischen Orakel (Meisenheim, 1971), p. 18 n. 83Google Scholar; Majercik, R. (Text, trans., and comm.), The Chaldean Oracles (Leiden, 1989), pp. 31–3CrossRefGoogle Scholar.

⁹³ See further, Dodds, , op. cit. (n. 92), pp. 315f.Google Scholar; Finamore, , op. cit. (n. 92), pp. 1fGoogle Scholar.

⁹⁴ See Kissling, R. C., ‘The ὂχημα–πνε⋯μα of the Neo-Platonists and the De Insomniis of Synesius of Cyrene’, AJPh 43 (1922), 320 n. 21Google Scholar.

⁹⁴ See Dodds, , op. cit. (n. 92), pp. 316fGoogle Scholar. Dillon, J., Iamblichi Chalcidensis in Platonis dialogos commentariorumfragmenta (Leiden, 1973), pp. 371–2CrossRefGoogle Scholar, argues on the basis of Proclus, , In Tim. iii. 234, 8–18Google Scholar Diehl that the πνευματικòν ὂχημα may already have been known to Atticus and Albinus in the second century A.D. According to Proclus, both denied the immortality of the irrational soul (called here ἃλοϒος ζωή) and the pneumatic vehicle, though there is no evidence apart from this passage that either Atticus or Albinus ever held to the notion of the vehicle as it was developed in Neoplatonism. The problem is compounded by the question of the identity of Albinus (see n. 81 above). That this Albinus is not the author of the Didaskalikos is, of course, Whittaker's, J. view: see ‘Platonic Philosophy in the Early Empire’, ANRW 36. 1 (1987), 88–9Google Scholar. Alcinoos, at any rate, gives no evidence of the doctrine of the vehicle, except in the Platonic sense of the body as the ὂχημα of the head and the rational soul (176, 16–17 Whittaker). On Poseidonius as a possible catalyst for the Neoplatonic linkage of the pneuma and the luminous body, see Aujoulat, N., ‘De la phantasia et du pneuma stoiciens d'après Sextus Empiricus, au corps lumineux neo-platonicien’, Pallas 34 (1980), 123–46, esp. 132ffCrossRefGoogle Scholar.

⁹⁶ See e.g. Sent. 29, p. 18, 6–7 Lamberz, Antro nymph. 11Google Scholar; ap. Procl, . In Tim. iii. 234, 18ffGoogle Scholar. Diehl; ap. Iambi. De an. (in Stob. i. 384, 19–28 Wachsmuth) and comments by Deuse, , op. cit. (n. 1), pp. 213–30Google Scholar; as Deuse argues emphatically (pp. 218ff.), although the vehicle and the irrational soul are intimately bound up with each other, they are not to be equated.

⁹⁷ Procl, . In Tim. iii. 234, 9–30Google Scholar Diehl. On Albinus here, see n. 95 above.

⁹⁸ See In Tim. fr. 81 Dillon (n. 95) and commentary ad loc. (pp. 371–7); Finamore, op. cit. (n. 92), pp. 17ff. On the creation of the vehicle by the cosmic gods (and not by the demiurge, contra Finamore, ibid. pp. 1 Iff.), see Deuse, W., Gnomon 59 (1987), 409fGoogle Scholar. Though the vehicle and the irrational soul can become increasingly purified (through theurgy), they do not enter the noetic realm; see Finamore, , op. cit. pp. 145ffGoogle Scholar.

⁹⁹ In. Tim. iii. 236, 3Iff. Diehl (on ⋯κρότητες, see Dillon, , op. cit. (n. 95), p. 374Google Scholar), El. th., prop. 209 and Dodds, ad loc. (pp. 306–8Google Scholar); for a comparison of Proclus and Iamblichus, see Finamore, , op. cit. (n. 92), pp. 86ffGoogle Scholar.

¹⁰⁰ Another may have been his daemonology (cf. Heinze, pp. 110–23). Thus Porphyry held that those sublunary souls who subjugated their pneumatic bodies by force of reason were to be considered good daemons, whereas those who were controlled by the pneuma and its impulses and desires became κακοερϒοί (Abst. 2.38 p. 167, 8–11; p. 167, 26–p. 168, 5 Nauck; cf. Deuse, , op. cit. (n. 1), p. 216Google Scholar; Heinze, p. 120). The explanation of evil daemons in terms of overpowering affections recalls Xenocrates' belief, which we derived from Plutarch, that evil daemons were the souls of men in whom the affective and irrational element remained strong. (Origen the Platonist was undoubtedly an important influence on Porphyry's De abstinentia, but not, I think, to the exclusion of Xenocrates, Numenius, and Plotinus; I cannot follow Lewy, H., Chaldean Oracles and Theurgy (Paris, 1978), pp. 497–508Google Scholar, that Origen was Porphyry's ‘only literary source’.) A significant difference between Xenocrates' daemons and later daemonology is the pneumatic body with which Neoplatonists invested the daemons. For both daemons and human souls one of the functions of the πε⋯μα was as the organ of sense perception and the imagination (αἰσθητικόν and ϕανταστικόν – cf. e.g. Simpl, . De an. ad 403*8 p. 17, 16–17Google Scholar Hayduck); it thus provided an intermediary between sensations and the incorporeal soul (and also furnished a body that could be brought into Hades where souls might be punished by the ϕαντασία of their former sins; see Porph, . Sent. 29Google Scholar; id.ap. Stob. i. 428, 8ff. Wachsmuth). The impulse for much of this comes from Aristotle's ‘inherent breath’ (πνε⋯μα σύμϕυτον see e.g. Somn. vig. 456^a12 and Ross, W. D. (ed.) Aristotle's Parva Naturalia (Oxford, 1955), pp. 40–3Google Scholar; see also n. 93 above, and Kissling, , op. cit. (n. 94), 319Google Scholar); moreover, to some extent the ὂχημα–πνε⋯μα may be considered the Neoplatonic response to Aristotle's premise (esp. of De anima ii) that sensations, being affections and movements, are only possible in the composite substance of body and soul (as a further step, Iamblichus' bestowal of immortality upon the irrational soul and its vehicle has been thought to meet ‘the Aristotelian objection that a soul must be the) ⋯νετεχέχεια of some body’; Dodds, , op. cit. (n. 92) ad prop. 209 (p. 307Google Scholar; cf. p. 317). As for Xenocrates, his daemons (which we saw to be the souls of the departed) were invisible and incorporeal; their sensitive element existed by participation in embodied souls (hence we spoke of a ‘psychic bond’) and enabled them to experience human affections only vicariously; it further explained the overwhelming tendency in some daemons towards the material and their consequent reincarnation. But their immateriality is problematical in at least two respects. First, in what did their sensitive, irrational component subsist? The second question arises out of the first: when the irrational part of the soul becomes weak and dissolves into the moon, and the rational part is freed to return to the intelligible realm, what is it exactly that is dissolved? In fr. 15 (above, pp. 145–6) Xenocrates speaks of certain divine powers that pervaded the material elements. Presumably those powers that pervaded the aer, the element characterizing Hades, are to be identified with the daemons. It is then conceivable that the irrational part of the daemon-soul derived from its element of aer a material substance, a sort of airy body (cf. the aerial creatures of Epin. 984el; thereto, see Tarán, op. cit. (n. 16), pp. 283f.), that eventually dissolved into the moon, whose substance was composed of aer and the second density (fr. 56 H). But to this conjecture, based in part on the restoration of a lacuna (see n. 14 above), it is preferable to suppose that these were problems bequeathed by Xenocrates to our later Platonists. It is notably Porphyry's application of the pneuma body to daemons that went some way to answering the questions left by Xenocrates: the πνε⋯μα of daemons was called ⋯ερ⋯δες its hylic nature rendered it visible on occasion; and it could be dissolved into its element(s) (⋯ναστοιχειο⋯σθαι – used of human souls in Porph. In Tim. iii. 234, 21 Diehl, but also applicable to the pneuma of certain daemons which Porphyry calls παθητικόν and ϕθαρτόν, Abst. ii, 39, p. 168, 12–13 Nauck); see Dodds, , op. cit. (n. 92), p. 319Google Scholar. While Porphyry's intricate daemonology is a product of a variety of sources and influences, many of its formulations appear to be Xenocratean in origin. See the accounts by T. Hopfher, Griechisch-Ägyptischer Offenbarungszauber i (Leipzig, 1921; repr. Amsterdam, 1974), pp. 21–8; Zeller, E., Die Philosophie der Grkchen in ihrer geschichtlichen Entwicklung iii/2 (Leipzig, 1903), pp. 727ff.Google Scholar; Bidez, J., Vie de Porphyre (Ghent, 1913), pp. 88–97Google Scholarpassim; Lewy, , op. cit., pp. 497–502Google Scholar.

¹⁰¹ In connection with Porphyry, Dodds, , op. cit. (n. 92), p. 306 n. 4Google Scholar, compares ‘the perishable ψυχή of the curious myth in Plut, . de facie, 28Google Scholar, 943 A ff., which Reinhardt, , Kosmos und Sympathie 318ffGoogle Scholar. refers to Poseidonius'. But Jones, , op. cit. (n. 54), pp. 116–31Google Scholar (see esp. 128ff), has in the interim shown that the myth of the Fac. lun. is not that of Poseidonius but Plato (cf. n. 56 above). Jones's argument, though, should not be taken to rule out the intervening influence of Xenocrates on both Plutarch and Porphyry, especially in regard to the perishable irrational soul.