^{page 256 note 1} In writing this paper, a version of which was delivered to the London Classical Society in October 1964,1 have gained much from discussion with Professor F. Solmsen and Mr. J. Longrigg.

^{page 256 note 2} Paideia, Engl. tr. (2nd ed., New York, 1945) i. p. 295.Google Scholar

^{page 256 note 3} See for instance Kirk, G. S. and Raven, J. E., The Presocratic Philosophers (Cambridge, 1957), pp. 343 ff.;Google ScholarSolmsen, F., ‘Greek Philosophy and die Discovery of the Nerves’, Museum Helveticum xviii. 3 (1961), 157 ff.,Google Scholar id., ‘Tissues and the Soul’, Ph. R. lix. 4 (1950) 436–442;Google ScholarVlastos, G., ‘Presocratic Theology and Philosophy’, Ph. Q. ii. 7 (1952), 119–21;Google ScholarKahn, C. H., ‘Religion and Natural Philosophy in Empedocles’ Doctrine of the Soul’, A.G.P. xlii (1960), 1–35;Google ScholarBollack, J., ‘Die Metaphysik des Empedocles als Ent-faltung des Seins’, Philologus ci (1957), 30–54.Google Scholar

^{page 256 note 4} Empedocle (Turin, 1916), p. 260.Google Scholar

^{page 256 note 5} From Religion to Philosophy (London, 1912), p. 225.Google Scholar

^{page 256 note 6} Aristotle's Criticism of the Presocratics (Baltimore, 1936), p. 294.Google Scholar

^{page 256 note 7} Op. cit., p. 121 n. 109.Google Scholar Cf. also Schwabl, H., ‘Empedokles fr. B 110’, Wiener Studien lxix (1956), 55, ‘Empedokles hat das Geistige offenbar völlig mit den Elementen verbunden’.Google Scholar

^{page 257 note 1} Kern, O., ‘Empedokles und die Orphiker’, A.G.P. i (1888), 498 ff.Google Scholar finds ‘Die Lehre von der Seelenwanderung’ the only Pythagorean doctrine in Empedocles and suggests Orphic sources for this. It is well to distrust the doxographical account concern ing Empedocles' relationship with Pythago ras (cf. the conflicting evidence in Diogenes Laertius 8. 54 ff.) but the points of similarity between the Pythagorean system and Em pedocles, whoever influenced whom, are beyond doubt, e.g. Empedocles’ explanation of compounds in terms of numerical propor tion (B 96; B 98), the common emphasis on vegetarianism and the need for purification, the conception of the universe as a kosmos; see Guthrie, W. K. C., A History of Greek Philosophy (Cambridge, 1962), i. 208 ff.,Google Scholar and Burkert, W., Weisheit und Wissenschaft (Nuremberg, 1962), pp. 113 ff.Google Scholar

^{page 257 note 2} occurs once in the B 138, where it is generally taken to mean ‘life’ as in Homer. For cf. B 115, B 126, and Plutarch's introduction of B 122. At B 59 in the has no meaning of ‘soul principle’ but refers either to Love and Strife (so Diels-Kranz) or to the four elements. Whether every man has been, possesses, or can become a is not at all clear. But the principle has nothing to do with cognition. and are Empedocles' words to denote ‘intellect’, cf. B 2. 8; B 17. 14; B 23. 9; B 136. 2. Détienne, M., ‘La démono-logie d'Empédocle’, R.É.G. lxxii (1959), 1–17,Google Scholar distinguishes two categories of spiritual daimon, ‘fallen beings’ (B 115. 13) and DK, p. 360, 1 (from Porphyry). But the evidence is too scanty to permit any firm conclusion; see further p. 274.

^{page 257 note 3} Edition of Aristotle, , De Anima (Cambridge, 1907), p. 221:Google Scholar ‘Empedocles did not hold that the soul is composed of the elements: but what we call the activity of the soul he explained by the elementary composition of the body; a soul distinct from the body he did not assume.’

^{page 258 note 1} Frs. 131–4 form a single group, with statements about the gods as their subject- matter. According to Tzetzes, , Chiliad, vii, 522,Google Scholar B 134 should be assigned to the ‘third’ book of the Bignone, , op. cit., pp. 631–49,Google Scholar defends Tzetzes' attribution against Diels, , Sitzb. Berl. (1898), 399 ff.Google Scholar, who denied the division into books (pace Simplicius, Phys. 157, 25, etc.) and argued that the materialistic doctrine set out in the was unlikely to have included any theology. But the of B 29 is a material entity and the repetition of these lines in B 134 suggests that the two passages may belong together. Yet even if the and the are closely connected this does not provide a link between human thought and the thought of the ‘holy mind which darts all over the world’; see Kahn, , op. cit., p. 28 n. 68.Google Scholar

^{page 258 note 2} Verdenius, W.J., Parmenides (Groningen, 1942), pp. 71–72,Google Scholar explains away this contrast between the poems on the grounds that Empedocles excepts himself from the restrictions on knowledge set out in B 2 and B 3. This involves a petitio principii since Verdenius argues entirely from statements made in the

^{page 258 note 3} Alcmaeon DK 24, B 1. Alcmaeon's date is notoriously difficult to fix (see Guthrie, , op. cit., pp. 341 ff.Google Scholar) but he was almost certainly active by the mid-fifth century, and earlier if Aristotle Met. A, 986^a27 is genuine. Alcmaeon, like Empedocles, makes use of the principle of symmetry (B 5), a theory of pores (A 5), and an explanation of vision which may have influenced Empedocles; see p. 263. But Alcmaeon distinguishes thinking from perception and makes perception depend on the action of unlike qualities; in both respects Empedocles differs from him.

^{page 259 note 1} ‘Le Fragment III d'Empédocle’ Classka et Mediaevalia xvii (1956), 47–61.Google Scholar Van Groningen takes 11. 4–5 of B 112 to mean ‘Empédocle a l'air d'une divinité, on le considère comme tel, il semble être ce qu' on croit de lui’, p. 51. This makes far too much of line 5 (if is the righi reading) and requires humour in Empedocle: which would be not only unique, but alsc quite inappropriate to the highly serious torn of the This interpretation is forcec upon van Groningen because he wishes to find B in spurious, and all claims by Empedocles to superior powers must therefore be explained away. B III should bt interpreted with an eye to B 129, Whether or not this refers to Pythagoras, it reveals an awareness in Empedocles that some men possess extraordinary powers, cf Snell, B., ‘Die Ausdrücke für den Begriff des Wissens’, Philologische Untersuchungen (Berlin 1928), 67.Google Scholar

^{page 259 note 2}

^{page 259 note 3} DK. 31 A 86, p. 302. 20 ff.:

^{page 259 note 4} All the elements are necessary for flesh and blood (B 98), and for bones too (B 96), if Simplicius is right in saying that here equals air as well as water. Galen (A 34) says that the of compound bodies is composed from all four elements, but Aetius (A 78) omits air from the formation of sinews and claims that nails are produced when ‘chilled sinews’ encounter air. Em pedocles' position comes close to that of Anaxagoras, but he probably did not hold that ‘everything has a share of everything’; cf. Millerd, C. E., On the Interpretation of Empedocles (Chicago, 1908), p. 41.Google Scholar

^{page 260 note 1} Cf. Plato, Meno 76 c,

^{page 260 note 2} Gf. DK 23 A 86, p. 303, 8 ff., A 88, A 8g and in particular A 86, p. 303, 10 ff.,

^{page 260 note 3}

^{page 261 note 1} Cf. Theophrastus (DK A 86, 2) On its origins and history see Guthrie, , Greek Philosophy i. 209,Google Scholar and Verdenius, W. J., ‘Empedocles' Doctrine of Sight’, Studia Varia Carolo Gulielmo Vollgrqff Oblata (Amsterdam, 1948), p. 155 n. 2.Google Scholar

^{page 261 note 2} Op. cit., p. 155 n. 3.Google Scholar

^{page 261 note 3} There is no clear evidence for Verdenius's claim in the passages which he cites (DK 31 A 86. 7; A 87; A 92; 28 A 47), op. cit., p. 155 n. 1.Google Scholar

^{page 261 note 4} DK 31 A 86, p. 303, 9 ff.,

^{page 262 note 1} DK 31 A 86, p. 305, 5 ff. Theophrastus criticizes Empedocles' theory that dvanvciv is necessary to smelling since, he argues, this does not account for the fact that creatures which do not inhale possess the sense.

^{page 262 note 2} Diels's supplement is almost certainly right since we have Empedocles' own words for the water in the eye, B 84. 10.

^{page 262 note 3} The text printed here makes one alteration from that in Diels-Kranz: I have preferred Förster's to the two MSS. readings, and see Hermes lxxiv (1939), 102–4.Google Scholar

^{page 263 note 1} For details of the literature on this problem see Verdenius, , op. cit. (p. 261, n. 1), pp. 155 ff.Google Scholar

^{page 263 note 2} Although this must be the meaning, unless Theophrastus has misunderstood Empedocles, there is a difficulty because not all aspects of the comparison are precise. The water (and other elements ?) in the eye must correspond both (1) with the winds which the lantern plates prevent from extinguishing the flame and also (2), along with the mem branes, the plates themselves (the circum ference of the eye) through which the eye's fire (in the centre) penetrates. In B 84 Empedocles emphasizes (1); Theophrastus (ad loc.) concentrates on (2). The incom pleteness of both accounts may be partly responsible for the alleged inconsistency in Empedocles' theory of vision.

^{page 263 note 3} Greek Theories of Elementary Cognition, p. 97.Google Scholar

^{page 263 note 4} Aristotle's Criticism of the Presocratics, p. 318.Google Scholar

^{page 263 note 5} Tim. 45 b ff. Taylor, A. E., A Commentary on Plato's Timaeus (Oxford, 1928), pp. 278 ff.,Google Scholar argues that Empedocles explained sight in general by fire issuing from the eye but brought in pores and effluences to account for colour vision, cf. DK A 86, p. 301. 34 ff. But it is hard to see why Empedocles should have distinguished vision in general from perception of colour, see Verdenius, , op. cit., p. 158. Beare, however, pp. 15 ff., goes too far in making the pores of fire and water alone responsible for vision. We cannot as sume that there were no pores of earth and air, and Beare's account fails to explain the primacy of function assigned to fire.Google Scholar

^{page 263 note 6} DK 24 A 5. 20 ff. Verdenius, , op. cit., pp. 159 ff.Google Scholar

^{page 264 note 1} A statement by Aetius (DK A 92) might lend support to the view that there are four sets of pores for the eye: ‘Empedocles de fined colour as that which fits into the pores of the eye (cf. DK A 86, p. 301. 34 ff.; Plato, Meno 76 c); white, black, red, and yellow are the primary colours, equal in number to the elements.’

^{page 264 note 2} DK 31 A 86, p. 301. 38 ff. Whether or not Empedocles abstracted light from fire it seems most probable that and in Theophrastus mean the eye's fire and fire (light) outside the eye. In B 85 when describing the formation of the eye Empedocles talks of meeting earth, and is used in B 84. 11. In B 84. 5, and are both attested for the lantern light.

^{page 264 note 3} Onians, R. B., The Origins of European Thought (Cambridge, 1951), pp. 76–77, quotes pre-philosophic evidence for the belief ‘that the eyes not only receive light or what is “breathed” from objects but are themselves stressed as sources of the same’.Google Scholar

^{page 265 note 1} in 1. 11 lacks a subject, but added by Diels, is certain, cf. 12–13. I retain in 1. 10 with Beare, instead of reading (Karsten) adopted by Diels, see Beare, , op. cit., p. 97.Google Scholar

^{page 265 note 2} I give Millerd's paraphrase for op. cit., p. 86.Google ScholarBeare translates on the basis of Diels's emendation but questions this in a note, p. 96. Bignone, , p. 370, translates ‘come un sonaglio che riproduce all'unisono suoni (esterni)’.Google Scholar

^{page 265 note 3} Stratton, G. M., in his edition of Theophrastus, De Sensu (London, 1917), p. 167, argues in favour of ‘trumpet-bell’, which would presumably be a vibrating device. Much depends on whether one agrees with Beare (p. 96) that means something in the ear discovered by dissection. Yet whatever Empedocles referred to by cannot now be determined; and at this date philosophers did not feel obliged to modify theories if empirical evidence failed to confirm them. It cannot simply be assumed that Aetius's source is drawing his own conclusions from the wordGoogle Scholar

^{page 265 note 4} Theophrastus (DK A 86, p. 301. 26 ff.),

^{page 266 note 1} The fact that the other senses are ex plained by a form of touch means, as Solmsen puts it, ‘that touch itself is no longer a sense function of distinct individuality’, Aristotle's System of the Physical World (Cornell, 1960), p. 349.Google Scholar

^{page 266 note 2} With this cf. Aristotle's view that ‘touch’ is not direct touch or contact but perceived through the medium of die flesh, De An. 423^b2–27; see Solmsen, , ‘Antecedents of Aristotle's Psychology’, A.J.P. lxxvi (1955), 159–60.Google Scholar

^{page 266 note 3} The suggestion that each sense requires a condition or medium of perception should not be taken to imply diat Empedocles anticipates Aristotle's theory of perception that there is a medium (e.g. for sight, De An. 418^a7 ff.) which when actualized provides a communication between object and organ. Aristotle's theory, as Solmsen has shown, ‘Antecedents …’, A.J.P. lxxvi (1955),Google Scholar breaks new ground. In Empedocles the act directly on the Empedocles seems to have welded together a new scientific theory of perception based upon his general postulates about the natural world and the common-sense fact that we appear to see when there is light and to hear when there are sounds to be heard. See in general Kirk, G. S., ‘Sense and Common-sense in Greek Philosophy’, J.H.S. lxxxi (1961), 111–12.Google Scholar

^{page 267 note 1}

^{page 267 note 2} A.G.P. xlii (1960), 14.Google Scholar

^{page 267 note 3} Cf. Homer Il. 22. 59; Aesch. Choeph. 517; Plato, Soph. 249 a.

^{page 267 note 4} The Aristotelian commentators (cf. Diels-Kranz ad loc.) take dreaming to be the subject of the passage. Verdenius, , Parmenides, p. 20, argues that this is improbable.Google Scholar

^{page 268 note 1} DK 28 B 16. See Vlastos, G., ‘Parmenides' Theory of Knowledge’, T.A.P.A. lxxvii (1946), 66–77.Google Scholar The objects of Parmenides’ ‘mixed’ thinking have of course a quite different ontological status from those of Empedocles.

^{page 268 note 2} DK 31 A 86, p. 302, 25 ff.

^{page 268 note 3} ‘Greek Philosophy and the Discovery of the Nerves’, Museum Helveticum xviii (1961), 157. On Alcmaeon's theory cf. Theophrastus (DK25 A 5, p. 212. 6 ff.).Google Scholar

^{page 268 note 4} The are clearly blood-vessels and probably have nothing to do with transmitting

^{page 268 note 5} ‘Empedokles fr. B 110’, Wiener Studien lxix (1956), 49–56.Google Scholar

^{page 269 note 1} Bollack, J., ‘Die Metaphysik des Empedokles als Entfaltung des Seins’, Philologus ci (1957), 52, understands as ‘das Seiende’. His attempts to establish a syste matic concept of being in Empedocles are not wholly convincing.Google Scholar

^{page 269 note 2} Gf. B 17. 27, ibid. 34; B 21. 13; B 26. 3; see Munding, H., ‘Zur Beweisführung des Empedokles’, Hermes lxxxii (1954). 134–6.Google Scholar

^{page 269 note 3} Op. cit., p. 49.Google Scholar

^{page 269 note 4} Reduplication and repetition are common in the Presocratics, cf. ibid., line 10; and B 84. 10–11; and especially Heraclitus B 1 see Kirk, , Heraclitus (Cambridge, 1954), p. 242;Google Scholarand Parmenides B 1. 28, etc. This is natural enough in poetry. Kahn, , op. cit., p. 15,Google Scholar takes and to mean ‘our own nature and character’. This is perhaps possible and simpler, but it seems to me more appropriate in the context to refer the words to ‘elements’. At B 17. 28 it is each element which has its own character. I follow Schwabl's interpretation, op. cit., p. 52,Google Scholar ‘so wie eines jeden Natur ist’, on which see the detailed note in Bollack, , op. cit., pp. 50–51.Google Scholar

^{page 270 note 1} I take in a transitive sense, following Wilamowitz, Kranz, and Schwabl against Diels, Burnet, and Bignone. In the active elsewhere in Empedocles is transitive, cf. B 17. 14, and B 37. Even if we translate with Burnet, ‘these things will each grow into the heart’, the process must still work in the manner outlined above. But the context naturally requires a distinction between external and internal elements, cf. B 106, ‘human understanding (internal elements) grows in relation to what is present (external elements)’.

^{page 270 note 2} DK A 86, p. 303. 9 ff.

^{page 270 note 3} Growth in Empedocles is explained by Aristotle in terms of addition, De Gen. et Con. B 6. 333^a35, on the basis of Em pedocles' own words in B 37,

^{page 270 note 4} Cf. B 3. 12, where may be more than a metaphor.

^{page 270 note 5} This explains why an understanding of Empedocles' teaching results in a growth of the B 17. 14, a line which sums up in less technical language the essentials of the process set out in B 110.

^{page 271 note 1} Empedocles' application of laovoyiia to the universe is noted by Vlastos, , ‘Equality and Justice in Early Greek Cosmologies’, C.P. xlii (1947), 159. At any given moment one element may possess more power than another, but in the long run the power is shared equally, so that a true account of the universe would have to conform to a pattern ofGoogle Scholar

^{page 271 note 2} Op. cit., p. 52.Google Scholar

^{page 271 note 3} ‘So ist … ein Abstumpfen der Denkkraft ein ganz realer “elementarer” Verlust, der schlieβilich zur Auflösung führen muβ’, op. cit., p. 54.Google Scholar

^{page 272 note 1} Op. cit., p. 51.Google Scholar

^{page 272 note 2} Op. cit., p. 147.Google Scholar

^{page 272 note 3} ‘Le Fragment III d'Empédocle’, Classica et Mediaevalia xvii (1956).Google Scholar Van Groningen's reasons for condemning these lines are: (1) Empedocles does not claim these powers elsewhere; (2) the contents of B 111 reproduce the doxographical account of Empedocles as a wonder-worker, but exaggerate this, if anything; (3) the style is conventional and lacks the ‘images nouvelles et tournures audacieuses’ found elsewhere. With regard to (1) see above, p. 259, n. 1; it involves an improbable interpretation of B 112. (2) entails a petitio principii: doxo-graphers may embellish their sources but they require some evidence to start from. (3) is not very unconvincing: to establish it would require an exhaustive survey of Empedocles’ language and style, but van Gronin-gen makes no reference to Traglia, A., Studi sulla lingua di Empedocle (Bari, 1952).Google Scholar Within nine lines we find one hapax legomenon and and are Homericisms not attested elsewhere.

^{page 272 note 4} Op. cit., p. 55.Google Scholar

^{page 273 note 1} B 20, cf. B 17. 22–23.

^{page 273 note 2} Cornford, , From Religion to Philosophy, pp. 234–42;Google ScholarKahn, , op. cit., pp. 19 ff.;Google ScholarRaven, in The Presocratic Philosophers, pp. 348–61;Google Scholar cf. also Long, H. S., ‘The Unity of Empedocles' Thought’, A.J.P. lxx (1949), 142–58.Google Scholar

^{page 273 note 3} (sc. the other three elements) Line 4 of this fragment, though corrupt, repeats the emphasis on approximate equality.

^{page 273 note 4} Kahn, , op. cit., p. 22,Google ScholarBollack, , op. cit., pp. 49–50.Google Scholar

^{page 273 note 5} As Vlastos says, ‘Equality and Justice in Early Greek Cosmologies’, C.P. xlii (1947), 160, ‘Were not Harmony matched with its perfect equal in Strife, there would be no created world, only the nondescript mixture of the Sphairos’; cf. Aristotle Met. B 4, 1000^b1.Google Scholar

^{page 274 note 1} Cf. De Gen. et Corr. B 6. 334^a6 and De Caelo Γ 2. 301^a14. The very fact that elements are more evenly mixed in some compounds than others should be suffi cient reason for the uneven distribution of Love.

^{page 274 note 2} Met. B 4, 1000^b3. This is a penetrating criticism by Aristotle, which cannot be simply explained away as Bignone suggests, p. 140, as ‘un'obiezione che non ha luogo per la coscienza mistica di Empedocle, in cui la coscienza del male e un'imperfezione’. Perhaps Empedocles would have argued that at the period of Love's dominance Strife's activity counts for nothing and is therefore irrelevant to any knowledge of the universe at that particular time.

^{page 274 note 3} Op. cit., pp. 19 ff.Google Scholar

^{page 274 note 4} De An. 404^b11, 408^a16, see earlier, p. 257.Google Scholar

^{page 275 note 1} De Exilio 17. 607 (sc. Empedocles). Kirk, and Raven, , pp. 359 ff.,Google Scholar make a good deal of this passage and suggest that Plutarch's comment was based upon a lost passage of Empedocles. But Plutarch could equally well have fathered the notion on Empedocles since it was so much a feature of the Academic tradition. It is, moreover, strange if Plutarch (in spite of the work in ten books on Empedocles which the Lamprias catalogue [43] attributes to him) had access to passages of Empedocles unknown to Aristotle. Yet if Aristotle knew of such a passage his vagueness and speculation about m Empedocles become very surprising.

^{page 275 note 2} Cf. Wilamowitz, , ‘Die des Empedokles’, Sitzb. Berl. (1929), 28, ‘In den Katharmen ist alles Offenbarung oder Gebot’.Google Scholar

^{page 275 note 3} B 110. 1; cf. B 3. 2. This is a point well made by Schwabl, , op. cit., p. 55, who argues that the of the is ‘nicht Prinzip der Geistigkeit als solcher, sondern in erster Linie Prinzip des selbstverschuldeten oder selbstverdienten Schicksals’.Google Scholar

^{page 276 note 1} Cf. Wilamowitz, , op. cit., p. 32.Google Scholar

^{page 276 note 2} Cf. Friedlälnder, , Plato i, Engl. tr. (London, 1955), p. 189, ‘Plato's great myths of the soul presuppose conceptual analysis and carry it beyond the limits set for human existence and human knowledge’. But the distinction is more clear-cut in Empedocles.Google Scholar