¹ Laffranque, M., Poseidonios d′Apamée Essai de mise au point (Paris, 1964), ch. 2, tries to do full justice to the influences of Posidonius' social and historical context on his life and thought.Google Scholar

² For a review see K. Reinhardt, RE 22,1 col. 570.28–624.11; Laffranque, op. cit., pp. 1–44; Hoven, R., Stoïcisme et Stoïciens face auproblème de I′au-delà (Paris, 1971), pp. 95–102.Google Scholar

³ As in Laffranque (op. cit.) or Long, A., Hellenistic Philosophy (London, 1974), pp. 216–222. For the fragments I refer to the collection by Edelstein L. and Kidd I. (Cambridge, 1972).Google Scholar

⁴ As in Ph. Merlan, ‘Beitrage zur Geschichte des antiken Platonismus: II. Poseidonios fiber die Weltseele in Platons Timaios’, Philologus 89 (1934), 197–214; From Platonism to Neoplatonism (The Hague, 1953), who focuses on the mathematical definitions of soul and the ontological status of mathematics in general. See also Dillon, J., The Middle Platonists (New York, 1977), p. 107.Google Scholar

⁵ In the commentary on F85 (Sext. Emp. Math. 7, 93), which deals with a theory of sense perception based on the principle of ‘like known by like’, cf. Ti. 45B–46C; see also Posidonius F193; Etym. Magn. ‘sight’, F194, Aet. 4, 13, 3. The Sextus Empiricus fragment contains the following statement: …, ,… Lasserre, F., ‘Abrégé inédit du Commentaire de Posidonius au Timée de Platon (PGen inv. 203)’, Accademia Toscana di Scienze e Lettere, Studi 83 (Firenze, 1986), 71–127, claims to have found a summary of Posidonius' ‘commentary’ in the manuscript tradition.Google Scholar

⁶ For the Early Stoics' interest (primarily Zeno and Chrysippus) in the Timaeus, cf. my forthcoming book ‘Demiurge and Providence: Stoic and Platonist Readings of Plato's Timaeus′ (with Brepols Publishers).

⁷ De Placitis 4, 258.19–23. Zeno did study with Polemo

⁸ Along the lines of the Early Stoics and the Timaeus, Posidonius considers the universe a unity (Edelstein and Kidd F4; D.L. 7, 143) and a living, ensouled and rational being (Edelstein and Kidd F99a; D.L. 7,142–3).

⁹ At the heart of the controversy lies the fragment of Diogenes Laertius which I quote below as (1) (7,134; SVF2,299; 300): the mss. tell us the principles are corporeal (), the parallel Suda version says they are incorporeal (). The following are fragments which call the principles corporeal: SVF 1, 90 (ambiguous, however); 98 (vide infra); 153; 2, 305 (corporeal matter); 310 (see also 475); 313 (Plutarch's critique, vide infra); 314–315; 320 (Plotinus, vide infra); 323 (Galen); 326 (corporeal matter); 336; 340 (cause is body); 343; 394; 467; 469 (‘everything’ is corporeal); 533 (corporeal matter); 793; 1028 () 1029; 1032; 1034 (in which, curiously enough, Plato is aligned with the Stoics); 1046; 1051–3.

¹⁰ For a review see Lapidge, M., ‘Archai and Stoicheia: A Problem in Stoic Cosmology’, Phronesis 18 (1973), 240–278CrossRefGoogle Scholar. See also Graeser, A., Zenon von Kition: Positionen und Probleme (Berlin, 1975), pp. 89–118CrossRefGoogle Scholar; Sandbach, F., The Stoics (London, 1975), pp. 71–75Google Scholar; Hunt H., A Physical Interpretation of the Universe: The Doctrines of Zeno the Stoic (Melbourne, 1976) pp. 17–25; R. Todd, ‘Monism and Immanence: Foundations of Stoic Principles’, in The Stoics, ed. Rist, J.(Berkeley, 1978), pp. 137–160Google Scholar; Mansfeld, J., ‘Zeno of Citium’, Mnemosyne 31 (1978), 162,CrossRefGoogle Scholar 167ff.; Sorabji, R., Matter, Space and Motion. Theories in Antiquity and their Sequel (Ithaca, New York, 1988), ch. 6, esp. pp. 93–98.Google Scholar

¹¹ Plut. De Communibus Notitiis 1085b—c.

¹² Plotinus Enn. 6, 1,26, 12ff.; see also SVF2, 323 on the problem of a corporeal matter

¹³ Both kinds of regress are pointed out by Graeser, op. cit., p. 103.

¹⁴ This problem is hinted at in SVF2, 310 (cf. also 475). Note that the same problem applies to the notion of a corporeal soul permeating a body; cf. SVF 2, 467; 471 (which presents Chrysippus' attempt at solving the problem). See also Todd, R., ‘Alexander of Aphrodisias on Stoic Physics. A Study of the De Mixtione…’, PhilAnt 28 (1976).Google Scholar

¹⁵ See, for instance, Lapidge, op. cit., p. 264 and Sandbach, op. cit., p. 74.

¹⁶ See n. 9.

¹⁷ See SVF 1, 90; 2, 343; 363. Cf. Long and Sedley, pp. 273-^ (commentary on the fragments; 45A = SVF 1, 90; 45B = SVF2, 363).

¹⁸ See, for instance, SVF 1, 171; 2, 1027; 1037 (which contrasts with ); 1054 (); 1133; 1134.

¹⁹ See, for instance, SVF 2, 311; 604; 605; 1064; see also Long, A., ‘The Stoics on World-conflagration and Everlasting Recurrence’, Southern Journal of Philosophy Suppl. 23 (1985), 22–23.Google Scholar

²⁰ See, for instance, SVF 1, 134ff; 2, 773ff; 790ff

²¹ See, for instance, 2, 774; 776; 777; 787.

²² For a full discussion, see Kidd in his commentary on F92, 368–74.

²³ See n. 9. In a paper he graciously showed me before publication, ‘Asomatos: Nuances of Incorporeality in Philo’, John Dillon has tried to make the case again for the reading, but claiming it only for Posidonius, who is mentioned last. He bases his argumentation on (a) the symmetry of the passage, (b) the possibility that a later, ‘smart’ scribe corrected the passage in light of the well attested corporeality of the Stoic principles. But note that according to (a) (argument of form) would be the lectio difficilior, and that the insertion of could justify the break in the symmetry; according to (b) (argument of content) would be the lectio difficilior. Contra (b), however, I argue that is the result of a Platonist influence on a later source, and that one would have to find a reply to Kidd's combination of evidence from Plutarch (cf. my discussion of my fragment 5 below) and Diogenes Laertius, which indicates that Posidonius too considered matter to be corporeal. I would like to thank John Dillon for showing me his paper before publication, and for his willingness to participate in an exchange of ideas.

²⁴ I use Kidd's translations as a starting point, in order to have some common ground for the discussion. Only when I disagree on the points relevant for my discussion have I indicated alternatives between brackets.

²⁵

²⁶ SVF 1,86–87; 2, 316–318; 300; 323.

²⁷ 373.

²⁸ Although these fragments are nominally attributed to the Stoics, I am aware of the fact that they reveal doxographical patterns and traits, i.e. structuring features, which are typical of ‘collections of opinions' like Diogenes Laertius' or Arius Didymus’, and which might not have belonged to the original wording. Yet this does not imply that nothing of the original terminology survived, and the recurrence of a specific vocabulary, with subtle shifts, remains a valid tool to assess Stoic doctrine.

²⁹ Hicks in his Loeb translation takes (i) as the subject and not as the predicate and (ii) as two coinciding notions, i.e. as which is . I adhere to a different interpretation. I am gratefully indebted to J. van Winden for having engaged himself in a discussion about the issue and having strongly defended the Hicks-Waszink hypothesis, the one I am arguing against. The following are my motivations for preferring the alternative reading: (i) the no article-predicate rule, (ii) ‘first matter’, not matter tout court, is called ‘being’ here, and (iii) precisely because there is no article (cf. in contrast the Posidonius passage 2) it is hard to tell whether the two designations are coinciding or not; I take this to be anticipating the distinction made immediately following. Note that my hypothesis as a whole does not stand or fall with this reading: minimally there are two kinds of matter being distinguished. For a parallel to reading as ‘something is called by two names’, see Galen Meth. Med. 14; 10, 974,15 (where it is used as an equivalent of ).

³⁰ This does not have to impinge on the so-called ‘Growing Argument’ (cf. Long and Sedley, pp. 172–3): growth and diminishment in the case of particulars () require an identity of a subject, something remaining the same throughout the process, and that identity would be due entirely to the active principle, not to the substrate. The question is whether the notion of ‘becoming more or less’, on the other hand, does ask for such a strong identity requirement, and whether it is to be put on the same level as ‘growing and diminishing’, see Posidonius F96 = Long and Sedley 28D, (5): ‘The substance neither grows nor diminishes through addition or subtraction, but simply alters, just as in the case of numbers and measures.’ (. As with ‘stuff’, one can add to or subtract from numbers and measures; amounts and sizes can become more or less, but they cannot grow or diminish in the strict sense used here.) See also Long and Sedley 28q = SVF1, 87 = part F92 Posidonius, my passage 2, where Zeno, Chrysippus, and Posidonius are mentioned, with the claim that for Zeno the sum total of matter does not become more or less, and that for Chrysippus it does not grow and diminish (both of these being true). The latter passage is not so helpful with regard to particulars, because they are described in terms of division and fusion, which are due to the active principle, in the lines devoted to Zeno.

³¹ 370.

³² As Terence Irwin pointed out to me, in Posidonius F96 = Long and Sedley 28D, Posidonius himself uses ‘substance’ both for prime matter and matter underlying particulars. This could indicate that he did not always make the distinction, just as in part 1 of the fragment that concerns us here. Note that in fragment 96 Posidonius does not use the word ‘’ at all, not even when he talks about the ‘qualified individual’ being one of two receptive parts, next to substance, for a ‘peculiarly qualified thing’. Again, I think that in part 2 of the fragment 92 Posidonius aligns himself with and consciously refers to his Stoic predecessors, i.e. if the source has any accuracy.

³³ See also Alcinous, Didaskalikos 162.29–39.

³⁴ Cc. 275–320 (ed. Waszink).

³⁵ Cc. 279–301.

³⁶ Cc. 283–301.

³⁷ Cc. 310–11; 319–20.

³⁸ See, for instance, 296.11–297.6.

³⁹ It reappears, for instance, in 310.12–15; 311.5; and Waszink (1962), p. xxxiv, has used it for a subdivision in his outline of the structure of the text.

⁴⁰ As Kidd himself points out, 370, ‘without quality’ () appears in SVF2, 300 (D.L. 7, 134; quoted here as 2.1); 301 (Stoics in general); 309 (Stoics in general); 313 (Stoics in general); 315 (Stoics in general); 318 (Origen, apparently an exception to the claim made here; vide infra); 323 (Stoics in general); 326 (Stoics in general); ‘without shape’ (Kidd's translation, ) we find in SVF2,299 (again D.L. 7,134); 314 (Stoics in general); ‘without shape’ () cf. SVF2, 311 (Stoics in general); 318 (the Origen fragment). The Origen fragment seems to be an exception, since it does pair up the two labels. But Waszink points out that there are verbal similarities between the Origen passage and the one of Calcidius discussed here (a resemblance which he believes to go back to Numenius) and Kidd, 374, also notes the resemblance between this particular passage of Origen and Posidonius' fragment. The ps.-Timaeus Locrus (94A) describes matter as ‘… ’. Cf. Baltes' commentary, PhilAnt 21 (1972), 41. ‘Without quality/form, shape’ does also occur in Philo Fug. 8 and the Didaskalikos 162.29–39 (see n. 33).

⁴¹ See Duhot, J.-J., La conception stoïcienne de la causalité (Paris, 1989), pp. 139–152; Seneca Ep. 65,2ff.; 89,16; Stob. Eel. 1, 13,1; SVF1,89; 2, 336; Edelstein and Kidd F95; Theophrastus fr. 230 (ed. Fortenbaugh et al).Google Scholar

⁴² See Kidd 368–9; Stob. Eel. 1.11.5; Diels, H., Doxographi Graeci (Berlin/Leipzig, 1929), pp. 457–458.CrossRefGoogle Scholar

⁴³ …

⁴⁴ In the Timaeus itself (52B2) it is said that the receptacle ‘can be grasped in some kind of bastard reasoning, not accompanied by sense-perception’ (… …).

⁴⁵

⁴⁶ See H. Cherniss, in his edition of this text; Laffranque, op. cit., pp. 431ff; Taylor, A., A Commentary on Plato's Timaeus (Oxford, 1928), pp. 11 Iff; P. Thévenaz, L āme du monde, le devenir et la matière chez Plutarque, dissertation (Neuchatel, 1938), pp. 63–67; J. Rist, Stoic Philosophy (Cambridge, 969), pp. 205 flfGoogle Scholar

⁴⁷ … … Cf. Edelstein and Kidd F16; D.L. 7, 135

⁴⁸ Kidd (commentary) F16, 126 connects this with Plato Timaeus 53Cff. and Meno 76A.

⁴⁹ See Edelstein and Kidd F100, 101, 139.

⁵⁰ See also n. 11.

⁵¹ Note that Plutarch, 1012Dff., includes Xenocrates as equating ‘divisible being’ with the dyad(matter).

⁵² See also D.L. 7, 150, where prime matter…substance is called ‘limited’ ().

⁵³ See J. Brunschwig, ‘La théorie stoicienne du genre suprême et l′ontologie platonicienne’, Matter and Metaphysics, Fourth Symposium Hellenisticum, Elenchos 14 (ed. J. Barnes and M. Mignucci, Naples 1988), pp. 19–127. In the Stoic view of the world two kinds of things have reality value and fall under the general heading of ‘something’ (): bodies and the incorporeal space, void, time, and the (‘meaning’, ‘sayable’). On the one hand, the combination of both ‘in thought’ and ‘in reality’ seems particularly suited to describe the admittedly awkward category of incorporeals that, unlike mere constructs of the mind such as universals, do have objective reality value, but not in the strong sense of bodies. On the other hand, the ‘limits’ are granted ‘being’ in Plutarch's passage (), as Posidonius' interpretation of Plato's ‘divisible Being’. Strictly speaking the Stoics consider only bodies and not the incorporeals as entitled to being (). (Brunschwig, pp. 42–60, discusses and dismisses apparent exceptions, including Seneca's claim, Epistulae 58, 22, that the void and time are quasi being—quae quasi suni). All of this evidence once again points towards a corporeal matter, as is also indicated by the context of Plutarch's passage about Posidonius and his followers (, 1022F).

⁵⁴ At first glance Plutarch himself appears to be oscillating between a corporeal matter (cf. Cherniss' list of passages, in his Loeb edition, 185 n.c), and one that is strictly qualityless and hence derives its sensible features from a participation in the intelligible (1013C, 1014F). Plutarch's corporeal matter bears the ‘traces’ of the elements (1016D–E; Timaeus 52D–53B); matter in its basic sense is qualityless (51A; 52A), but the matter the Demiurge structures does already have some features (see also 30A). This tension is present in Plato's own text, though Plutarch takes it further. Cherniss notes that Plutarch borrows Stoic terminology to talk about matter, as in 1014B, 180 n.b., and 1014C, 183 n.d.; the Stoics are mentioned at 1015B as using . On p. 185 n.c. Cherniss also draws attention to a passage from the Placita (Doxographi Graeci 308 A4–9, B5–9) in which the Platonic receptacle is called , in the Stoic vein.

⁵⁵ The intriguing point is that Iamblichus ascribes this definition to Speusippus (Tarán F54a–b; Lang fr. 40). Tarán (365–71; contra Merlan et al.) defends the hypothesis that Speusippus used this definition as an interpretation of the Timaeus (35A7, where ‘does not mean “idea” at all’; 36E2) and as a defence of Plato against Aristotle, without necessarily adhering to it himself

⁵⁶ Although I do not consider as sufficient evidence the fact that Macrobius attributes the term to Posidonius, in his Commentarii in Ciceronis Somnium Scipionis 1, 14, 19; Edelstein and Kidd F140

⁵⁷ And Alcinous, cf. n. 33.

⁵⁸ For the entire mathematical definition and its Old Academy heritage, see Merlan (1934) and Kidd in his commentary on F141, 533–5.

⁵⁹ A physical variant of this tenet for the human soul makes it extend even through the bones of the human body and connects this with the Timaeus 73B, where it is said that the soul's bonds are in the roots of the bone. Cf. Edelstein and Kidd (commentary 150–1) F 28a, b; Scholia in Homerum; Eust. Il. 12, 386.

⁶⁰ De Placitis, mainly books 3–5 (up to 336.15) in the edition of Ph. De, Lacey, Corpus Medicorum Graecorum 5, 4, 1, 2 (Berlin, 1978).Google Scholar

⁶¹ See the commentary of Kidd. For a thorough analysis of Galen's methodology and a defence of Chrysippus, see also Tieleman, T., ‘Galen and Chrysippus. Argument and Refutation in the De Placitis Books ii-iii’, dissertation, Quaestiones Infinitae 3 (Utrecht, 1992).Google Scholar

⁶² A point overlooked by Pohlenz, op. cit., 1, 227; 2, 114 and Laffranque, op. cit., 429: ‘L'ame comporte done trois facultés rattachées a trois endroits différents du corps humain (?) une substance unique issue fun seul lieu, le coeur, et enfin, des parties en nombre sensiblement plus élevé que celui des facultés, mais leur correspondant au moins partiellement.’ The last part of this statement is based on a very problematic account by Tert. De Anim. 14, 2; Edelstein and Kidd F147. Laffranque also plays the importance of the Timaeus for Posidonius down to the bare minimum: see, for instance, op. cit., pp. 373–4

⁶³ See also F142; Galen De Placitis 5, 312.29–34.

⁶⁴ I am gratefully indebted to David Blank for drawing my attention to the debate about the Stoic Diogenes of Babylon; see Nussbaum, M., ‘Poetry and the Passions: Two Stoic Views’, Passions and Perceptions. Studies in Hellenistic Philosophy of Mind; Proceedings of the Fifth Symposium Hellenisticwn (ed. Brunschwig, J. and Nussbaum, M.) (Cambridge, 1993), pp. 97–149Google Scholar. If, as Nussbaum claims, Diogenes' interest in music and poetry foreshadows Posidonius' doctrine, and if, indeed, Diogenes prepares the ground for Posidonius' revision of Stoic psychology, then we have yet another reason for breaking through Galen's ideological marshalling of Posidonius as a radical reformer. Two remarks are appropriate here: (a) I believe that Posidonius' position of allocating three faculties to one and the same soul substance is considerably weaker than a full-fledged return to Plato's doctrine of a tripartite soul; (b.l) Diogenes' musical theory might be ‘non-cognitive’ or irrational in the sense that it focuses on the senses, which are instrumental to the , but this position is still essentially different from a non-cognitive approach which assigns irrational faculties to the soul's ruling component itself, and it also has to be noted that perception is cognitive for the Stoics; (b.2) this point is related to the question whether any model of the impact of musical harmony on the soul necessarily presupposes this soul to be composite, in Plato's or Posidonius' sense. For a theory of musical harmony which is compatible with a unitary soul, see A. Long, ‘The Harmonics of Stoic Virtue’, Oxford Studies in Ancient Philosophy, Suppl. (1991), 97–116. For a more extended argument concerning the points of similarity between Chrysippus and Posidonius, see Fillion-Lahille, J., ‘Le De Ira de Seneque et la théorie stoïcienne des passions'. Etudes et Commentaires 94 (Paris 1984), 119–162.Google Scholar

⁶⁵ Cf. J. Mansfeld, ‘The Idea of the Will in Chrysippus, Posidonius and Galen’, Proceedings of the Boston Area Colloquium in Ancient Philosophy 1 (ed. Cleary, 1991, Lanham, 1993), 124.

⁶⁶ Galen himself makes this slip, see for instance De Placitis 5, 318.13–14. See also, for instance, infra, Edelstein and Kidd F186; Clem. Al. Strom. 2,21, 129.1–5 (text 9).

⁶⁷ Op. cit., pp. 33–4.

⁶⁸ Inwood (op. cit., pp. 30–3) refers to Iamblichus, Stob. Eel. 1, 369; Simpl., SVF3,203; Philo, SVF2, 844; Arms Didymus, SVF3,112

⁶⁹ De Placitis 5, 305.18–27.

⁷⁰ In a recent article, ‘Posidonian Polemic and Academic Dialectic: The Impact of Carneades upon Posidonius’ , Greek, Roman and Byzantine Studies 34 (1993), 229–323, John A. Stevens goes an important step further. He claims (p. 322) that Posidonius himself ‘attributes to Chrysippus subtly altered representations of Carneades' views’. But this apparent concession to the skeptics' criticisms is in fact motivated by Posidonius' attempt to salvage core Stoic doctrine by drawing on common sense experience about the nature of the soul and irrational behaviour. If Stevens is right, we have yet another, deeper-level manipulation of Chrysippus' views, one by Posidonius, that is embedded within Galen's own purposes.

⁷¹ See Edelstein and Kidd F31; Galen De Placitis 5, 322.28–326.8; F33, Galen De Placitis 5, 332.31–334.10. Posidonius uses for his description of the lowest category of animals a terminology which is reminiscent of Plato's account of plants (cf. Ti 77B).

⁷² In the Timaeus (44B), too, reason, in connection with its cognitive circular motions, is said to start out weak in children; cf. Edelstein and Kidd (commentary) F31, 155.162.

⁷³ For the case of animals, see the excellent analysis by Sorabji, R., ‘Animal Minds and Human orals’, Cornell Studies in Classical Philology 54 (Ithaca, NY, 1993).Google Scholar

⁷⁴ 398–9; 478, see also Reinhardt, op. cit., col. 746

⁷⁵ See Edelstein and Kidd (commentary) F 31, 156; 158–160, which also indicates that even in his use of the Phaedrus image, Posidonius borrowed terminology from the Timaeus.

⁷⁶ De Placitis 5,460–3.

⁷⁷ See Rist (1969), p. 212; I do disagree with Rist, however, that this needs to lead to a strong dualistic view of the human soul. See also part iv of this paper.

⁷⁸ Quocirca primum mihi videtur, ut Posidonius facit, a deo, de quo satis dictum est, deinde a fato, deinde a natura, vis omnis divinandi ratioque repetenda.

⁷⁹

⁸⁰ Two points fully explored by Graeser, A., Plotinus and the Stoics (Leiden, 1972), p. 110 andGoogle ScholarDragona-Monachou, M., ‘Posidonius' “hierarchy” between God, Fate and Nature and Cicero's De Divinatione’, Philosophia 4 (1974), 286–301. Kidd (commentary) thinks that Posidonius' (F108; Cic. Div. 1, 64) view of our minds being capable of divination because of reason, i.e. our kinship with the divine, stands in conscious opposition to the Timaeus 71D–72B, where divination is an essentially irrational process.Google Scholar

⁸¹ See, for instance, Jaeger, W, Nemesios von Emessa, Quellenforschung zum Neuplatonismus und seine Anfangen bei Posidonios (Berlin, 1914), pp. 97–133.Google Scholar

⁸² Edelstein, L., ‘The Philosophical System of Posidonius’, A JP 57 (1936), 292f., 301–5.Google Scholar

⁸³ Kidd in his commentary points out (p. 416) conflicts between this hypothesis and other views ascribed to Posidonius on cause (F95), soul (F141), punctual existence (F98, F16), matter and substance (F92) and on god (F101). I focus on the general role of the active and passive principle as in F5, vide infra

⁸⁴ F5; SVF1, 85; 2,299 (not 229 as in Kidd 104); D.L. 7, 134.

⁸⁵ Rist (1969), p. 211

⁸⁶ Op. cit., pp. 340,351.

⁸⁷ Op. cit., p. 292, and throughout her argumentation.

⁸⁸ See C i c. Div. 1, 1 3 1: ‘…, cum domus sit omnium una, eaque communis, …’

⁸⁹ She does write, however (op. cit., p. 299): ‘Nor is God unconnected with the corroboration of divination by Fate, since God is Fate, having decided, once and for all, all that is decreed to happen; but God's primary aspect is to be providential.’

⁹⁰ The latter view is defended by Brehier, E., History of Philosophy. The Hellenistic and Roman Age (trans. W. Baskin, Chicago, 1965), p. 137Google Scholar, and Verbeke G., Lévolution de la doctrine du pneuma du stoïcisme à S. Augustin (Louvain/Paris, 1945), pp. 130–131.

⁹¹ Op. cit., pp. 300–1.

⁹² cf. SVF1, 179ff.

⁹³ SVF3,4 = D. L. 7. 87–8 = Long and Sedley 63C.

⁹⁴

⁹⁵ I agree with Reinhardt (op. cit., col. 747) and Kidd (commentary on F187; 677), against for instance Edelstein (1936, 314) and Rist (1969, 212), that this passage does not necessarily imply we also have an evil demon, in the irrational part of our soul, a theory which would turn Posidonius into more of a dualist.

⁹⁶ See also Edelstein and Kidd F85.

⁹⁷ See also Tht. 176B, for the theme of .

⁹⁸ See, for instance, Pohlenz, op. cit., p. 229 and in reply Reinhardt, op. cit., pp. 747–8.29; Kidd (commentary 669–78).

⁹⁹ Compare and contrast also Xenocrates fr. 81 (ed. Heinze; Aristotle Topics 2, 6, 112a32): ; ‘as Xenocrates says, happy is the man who has his soul in good condition. For that (the soul) is each man's daimon.’ Notice that here, however, the soul in its entirety is our daimon.

¹⁰⁰ …

¹⁰¹ ‘’instead of‘’.See Kidd, pp. 672–3, for a discussion of the problems with this reading.

¹⁰² See Laffiranque, op. cit., pp. 450,478; Rist (1969), pp. 214–15.

¹⁰³ For a discussion of the Cicero passage (62C = SVF 2.974) in connection with Alexander of Aphrodisias On Fate 181.13–182.20 (62G = SVF2.979), see Long and Sedley, commentary, pp. 392–4.

¹⁰⁴ I presented a version of this paper at the University of Leiden, The Netherlands, in June of 1994, and at Cornell University, on February 2, 1996. The paper has greatly benefited from both audiences, and from the comments of the people I thanked in the footnotes, among whom Anthony Long and John Dillon, who saw me through the entire process and let the disagreements stand. Many essential final touches were added during my stay at the ‘Fondation Hardt’ in Geneva, May 20-June 8, 1996, made possible by a grant from the University of Notre Dame Institute for Scholarship in the Liberal Arts. I also would like to thank Kenneth Sayre and the reader of the journal