¹ For criticism of the Presocratics in Epicurean writings, seeBailey's, Commentary, ii. 709Google Scholar; Rösier, W., ‘Lukrez und die Vorsokratiker’, Hermes 101 (1973), 62–3Google Scholar; Kleve, K., ‘The philosophical polemics in Lucretius’, in ‘Lucrèce’, Entretiens Hardt 24 (Geneva, 1978), 64–5Google Scholar. Lucretius has selected chief representatives of monism and pluralism, together with the ultra-pluralism of Anaxagoras, which most nearly approached atomism (for the order, see Rösier, , loc. cit. 55 n. 3)Google Scholar. The broad scope of his criticisms is indicated by the use of plurals referring to followers or like-minded thinkers (e.g. 641 f., 655 f., 665 f., 782 f., 861) and the allusion to various forms of monism and pluralism (705 f., 712 f., 734 f.).

² The organic connection between this ‘digression’ and the Presocratic section is demonstrated byLenaghan, L., ‘Lucretius 1. 921–50’, TAPA 98 (1967), 221–51, esp. 227–36Google Scholar; for the way in which Lucretius prepares for 921–50 see alsoSchrijvers, P. H., Horror ac Divina Voluptas (Amsterdam, 1970), pp. 41–7, 85, 158–60Google Scholar.

³ For an interesting and useful study of these passages seeKollmann, E. D., ‘Lucretius' criticism of the early Greek philosophers’, Studii Clasice 13 (1971), 79–93Google Scholar.

⁴ The contrast is remarked upon by many critics, e.g. Giussani on 635–704; Ernout-Robin on 716 (and Text, i. 28 n. 1); Bailey, ii. 723, 725, 728;Boyancé, P., Lucrèce et l'Épicurisme (Paris, 1963), p. 101Google Scholar.

⁵ Kollmann, loc. cit. 82, who also mentions Heraclitus' statement about ‘war, the father of all and king of all’ (22 B 53 DK), to which M. F. Smith (in the Loeb edition) believes there is an ironic allusion, along with the contention that ‘strife is right’ (B80); less convincingly, Rozelaar, M. believes the expression is a sign of Lucretius', not Heraclitus' aggressiveness: Lukrez: Versuch einer Deutung (Amsterdam, 1943), p. 70Google Scholar. Primus too has satirical point if it not only refers to the usual order of treatment (cf. D.O. fr. 5 Chilton, which exhibits a similar vehemence) but ironically recalls the primacy of Epicurus' generalship in ridding the earth of superstition (1. 66 f., cf. 3. 2, 5. 9, 6. 4 f.); see Lenaghan, loc. cit. 227–8. The ‘primus motif’ has also been used of Iphigenia (1. 94) and Ennius (1. 117), and will be applied by Lucretius to himself in 1. 926 f. and 5. 336 f.

⁶ See Bailey's note and the references in Rösier, , loc. cit. 52 n. 3Google Scholar; with obscurant linguam cf. the descriptions obscure (Cic. Fin. 2. 15) and obscuritas (Sen. Ep. 12. 7). Lucidity, by contrast, was required by Epicurus (D.L. 10. 13, Cic. Fin. 2. 15) and claimed as his own special virtue by Lucretius (1. 136–45, 921–2, 933–4).

⁷ Munro compares 1. 475 (clara accendisset saevi certamina belli). An additional pun on Ήρακλείτος/clarus is suggested bySnyder, J. M., ‘The significant name in Lucretius’, CW 72 (1978–1979), 228–9Google Scholar, and Puns and Poetry in Lucretius' De Rerum Natura (Amsterdam, 1980), pp. 117–18Google Scholar. For Heraclitus' own use of oxymoron, antithesis and other word plays, see Kollmann, , loc. cit. 79–80Google Scholar, Snyder, , op. cit. 54–5Google Scholar.

⁸ Kollmann, , loc. cit. 84Google Scholar.

⁹ See, e.g., the notes of Munro, Giussani, Duff, Merrill, Ernout-Robin, Leonard-Smith, Bailey and M. F. Smith. Munro, followed by Duff, states that Lucretius ‘retorts upon them [the Stoics] their own term of reproach’, though he could equally well be echoing the aggressiveness of Heraclitus, as Kollmann thinks (loc. cit. 83 n. 22, also p. 80), or indeed of Empedocles (cf., e.g., 31 B11. 1 DK). It was generally assumed that the criticism of Heraclitus is a covert attack on the Stoics until D. J. Furley's comprehensive denial of specific Stoic criticism by Lucretius, in ‘Lucretius and the Stoics’, BICS 13 (1966), 13–33, esp. 15–16Google Scholar. This thesis has been challenged by Schmidt, J., ‘Lukrez und die Stoiker. Quellenuntersuchungen zu De Rerum Natura’ (Diss. Marburg/Lahn, 1975)Google Scholar, which I have not seen; see also Kleve, , loc. cit. 39–40, 63–70Google Scholar. The presence of stolidus in two ‘Stoic’ locations is highly suggestive (Furley does not address this point) and I.would agree with M. F. Smith that, while the Stoics may not be Lucretius' only (or even main) target in the Heraclitus section, it is most unlikely that he wasn't thinking ot them at all.

¹⁰ This is the ingenious theory of Holtsmark, E. B., ‘Lucretius and the fools’, CJ 63 (1968), 260–1Google Scholar, who remarks upon the many earlier uses of inanis and solidus in the same metrical position and concludes: ‘Lucretius, it seems, here allows himself a droll critique of those fools, the inanis and stolidi, who posit fundamental principles of the universe different from his own. It is as though the poet, calling them what he does, involves them inextricably and helplessly in acceptance of his own basic premise on the existence of inane and solidum.’ Snyder, , op. cit. 118–19Google Scholar, makes some cogent points in favour of retaining the traditional interpretation, but perhaps Lucretius is capable of making more than one joke at the same time.

¹¹ For the opposition of the words in a context pertaining to void and matter, cf. 1. 364–7 (ergo quod magnumst aeque leviusque videtur, nimirum plus esse sibi declarat inanis; at contra gravius plus in se corporis esse dedicat et multo vacui minus intus habere).

¹² Several instances are mentioned by Kollmann, , loc. cit. 81–5Google Scholar, who compares such techniques with the rhythmical and expressive language of Heraclitus.

¹³ For the suggestion, or rather parody, of Heraclitus' figurative manner see Munro on 635–44 and 644; Duff, , Leonard-Smith, , Bailey, , Smith, M. F., and West, D., The Imagery and Poetry of Lucretius (Edinburgh, 1969), p. 26Google Scholar; Kollmann, , loc. cit. 84Google Scholar. The phrase lepido fucata sonore looks like a deliberate distortion of Lucretius' own stylistic ideal of lepos; cf. 1. 933–4 (deinde quod obscura de re tarn lucida pango | carmina, musaeo contingens cuncta lepore), Lenaghan, , loc. cit. 229Google Scholar.

¹⁴ loc. cit. 85.

¹⁵ D.L. 8. 57. The idea that Lucretius' style is influenced here by Empedocles occurred to Ernout-Robin (on 1. 716), Lenaghan, , loc. cit. 230Google Scholar and Kollmann, , loc. cit. 87–8Google Scholar.

¹⁶ For vasta Charybdis (722) cf. , Hom.Od. 12. 113, 428Google Scholar όλοήν…Xáρυβδιν), also Cat. 64. 156, , Virg.Aen. 7. 302Google Scholar, Prop. 2. 26. 54; magni magno cecidere ibi casu (741) recalls Il. 16. 776 (κεȋτο μέγας μεγαλωστί), and cf. , Virg.Aen. 5. 447–8Google Scholar. In addition, 720 seems to echo , Enn.Ann. 302, and 718–19 and 738–9Google Scholar may be influenced by Callimachus, , as I have tried to show in ‘Lucretius and Callimachus’, ICS 7 (1982), 85–6Google Scholar.

¹⁷ The suggestion that Lucretius has portrayed obliquely the four Empedoclean elements interacting with one another was made by MacKay, L., ‘De rerum natura 1.717 sqq.’, Latinitas 3 (1955), 210Google Scholar. Snyder, J. M., in ‘Lucretius' Empedoclean Sicily’, CW 65 (1972), 217–18Google Scholar, argues that Lucretius has paired the heavy and light elements (water and earth in 718–21, fire and air in 722–5) in a typically Presocratic manner, and has arranged them in a sequence which reflects Empedocles' ascending series of transmigrations through the elements (31 B 115 DK), thus conferring immortality on his poetic spirit ‘which rises like the bright thunderbolts of Aetna toward the realms of aether’.

¹⁸ MacKay, , loc. cit. 210Google Scholar, hints at this in his suggestion that Lucretius may mean to imply that Empedocles at least framed his theory in accordance with the evidence of the senses, unlike Heraclitus. See also Kollmann, , loc. cit. 87–8Google Scholar, who contrasts Heraclitus and Empedocles in terms of their closeness to their communities, and also suggests that ‘by his description of Sicily Lucretius seems to explain how Empedocles came to this theory of the four elements’. It is also tempting to connect Lucretius' emphasis on Etna with the legend that Empedocles died by leaping into the crater to prove his divinity (e.g. , Hor.A.P. 463–6Google Scholar, with Brink, , D.L. 8. 69)Google Scholar, a story which may go back to Heraclides Ponticus, according to Guthrie, W. K. C., A History of Greek Philosophy (Cambridge, 1965), ii. 131 n. 1Google Scholar. Incidentally, in view of the way Lucretius links Empedocles with the landscape, it is interesting to read Guthrie's remark that ‘anyone who goes to Acragas… must feel pleasure that so vivid and dramatic a character was born and lived in so appropriate a setting’ (op. cit. p. 129).

¹⁹ B112 DK. Many critics have referred to this fragment, including Munro, Giussani, Merrill, Ernout-Robin, Leonard-Smith, Bailey, and M. F. Smith; see also Kollmann, , loc. cit. 90 n. 51Google Scholar and Rösler, , loc. cit. 54 n. 2Google Scholar, who argues that the Lucretian and other references to the divinity of Empedocles stem from a misunderstanding of the original passage.

²⁰ cf. Il. 16. 776 (n. 16 above), and for the word play see Snyder, , op. cit. 82, 124Google Scholar. I do not mean to imply that the use of puns is Empedoclean, but that, as a satirical device, punning j provides a small link between the Lucretian attacks (for Anaxagoras, see below and n. 38).

^2l For the influence of Empedocles on Lucretius, see Jobst, F., ‘Über das Verhältnis zwischen Lukretius und Empedokles’ (Diss. Erlangen, 1907)Google Scholar, Kranz, W., ‘Lukrez und Empedokles’, Philologus 96 (1944), 68–107CrossRefGoogle Scholar, reprinted in Studien zur antiken Literatur und ihrem Fortwirken, Schriften, Kleine, ed. Vogt, E. (Heidelberg, 1967), 352–79Google Scholar, Niccolini, O. B., ‘De T. Lucretio Caro’, Lalinitas 3 (1955), 280–6Google Scholar, Bollack, J., ‘Lukrez und Empedokles, Die Neue Rundschau 70 (1959), 656–86Google Scholar, Boyancé, , op. cit. p. 102 n. 3Google Scholar. The religious imagery ties in with language used elsewhere of Epicurus and others, including Lucretius himself; for the oracular metaphor cf. 3. 14, 6. 6 (Epicurus), 5. 111–12, a repetition of 1. 738–9 (Lucretius); for the epithet sanctus (730) cf. 3. 371, 5. 622 (Democritus), and for divini pectoris (731) and bene ac divinitus (736) cf. 5. 8, 49–54 (including the same phrase bene ac divinitus) and 6. 7 (all three passages in reference to Epicurus); see further Giancotti, F., II preludio di Lucrezio (Messina and Florence, 1959), p. 79Google Scholar. To continue with other interconnections, Empedocles is said to be in the forefront (cumprimis 716, inferiores 734), like Epicurus, Ennius, Heraclitus (n. 5 above) and Lucretius; like Lucretius again, he composes carmina (731, cf. 1. 934, 3.420, singular in 1. 143, 946, 5. 1, 6. 937), which, if praeclara reperta (732) hints at clarity as well as distinction (contrast obscura reperta, 1. 136), are written in a lucid style. See further Lenaghan, , loc. cit. 230 f.Google Scholar, and for the view that Empedocles’ selfportrait influenced Lucretius' portrayal of Epicurus see Gigon, O. in ‘Lucrece’, Entretiens Hardt (n. 1 above), 170Google Scholar.

²² loc. cit. 91.

²³ Others have attempted to identify specific gibes at Heraclitus in the conjecture Musae (657), interpreted as an ironic reference to Heraclitus or the title of his work (a suggestion developed by Bignone, E. in ‘Le Muse eraclitee in Lucrezio’, Miscellanea di studi critici in onore di E. Stampini [Turin, 1921], 229–31Google Scholar, and accepted by some, e.g. Ernout-Robin, but not by Bailey, Büchner, K., ‘Lukrez und Vorklassik’, Studienzur Römischen Literatur [Wiesbaden, 1964], 142Google Scholar, West, , op. cit. pp. 73–4 and others)Google Scholar; also in the road image (659), as applied to someone ‘who himself expressed doubt about the possibility of finding the path’ (Lenaghan, , loc. cit 228 n. 26Google Scholar, referring to B 45, 108), in the use of cernunt (660), taken to be an ironic repetition of 642 and 657 (Leonard-Smith) and in nilo clara minus (697), which could be another allusion to Heraclitus' obscurity (Lenaghan, , loc. cit. 230)Google Scholar. One wonders whether ex uno si sunt ignipuroque creatae (646) could be a pun upon πūρ (cf. 658).

²⁴ loc. cit. 48–64, esp. 61–4. Various other critics who have doubted that Lucretius' knowledge was acquired first-hand are listed on p. 49 n.

²⁵ Rösier, , loc. cit. 54–7Google Scholar.

²⁶ Lucretius advocates Epicureanism in the proselytizing spirit of a missionary, not like an objective scholar, so it is only natural to expect unfairness in his polemics. The satirical streak in his attacks on opposing thinkers is firmly rooted in a tradition begun by Epicurus and carried on by his later followers: see Kleve, , loc. cit. (n. 1 above) 39–71 passim, esp. 49, 60 f.Google Scholar; however, some of the harsher abuse attributed to Epicurus is probably due to hostile propaganda: see Sedley, D., ‘Epicurus and his professional rivals’, Cahiers de Philologie 1 (1976)Google Scholar, also entitled Etudes sur l'Epicurisme antique, 119–59.

²⁷ Boyancé, , op. cit. p.104Google Scholar, Kollman, , loc. cit. 92Google Scholar.

²⁸ D.L. 10. 12. Anaxagorean theories about celestial phenomena are apparently alluded to in a friendly spirit in Pyth. 92–6, 101, 105–6. The only direct reference to Anaxagoras by Epicurus, in a fragment whose authorship is questionable, describes someone, possibly Nausiphanes, as reading the works of Anaxagoras and Empedocles and quibbling about them (Fr. 104 Arr.). On Anaxagoras and Epicurus, see further Bailey, ii. 744 and Sedley, , loc. cit. 135–6, 155 n. 59Google Scholar, who suggests that Epicurus might have set out to ‘assume the mantle of Anaxagoras and Archelaus’ at Lampsacus.

²⁹ Possibly Lucretius knew nothing more, or less, than the little we know about Anaxagoras' life, for which see Kirk, G. S. and Raven, J. E., The Presocratic Philosophers (Cambridge, 1957), pp.362fGoogle Scholar. and Guthrie, , op. cit. pp.266fGoogle Scholar. His reputation for lack of interest in public affairs, his prosecution for impiety, and his exile in Lampsacus (like Epicurus) are things which might have endeared him to Lucretius, if he knew about them.

³⁰ His style is ‘attractive and dignified’ (D.L. 2. 6), but flat in comparison with the other two.

³¹ On 1. 830–74.

³² Paladini, V., Lucrezio: il poema delta natura (Rome, 1946), p.59Google Scholar.

³³ Masson, J., Lucretius: Epicurean and Poet (London, 1907), p.108Google Scholar.

³⁴ loc. cit. 92.

³⁵ Bailey, C., The Greek Atomists and Epicurus (Oxford, 1928), p.35Google Scholar. It has been argued that a major flaw in Lucretius' definition is his inclusion of the Empedoclean elements (earth 840, fire and water 841, air 853) among substances like bone and flesh, whereas Aristotle implies that Anaxagoras placed them in a different category of complexity: see, e.g., Cornford, F. M., ‘Anaxagoras’ theory of matter’, Studies in Presocratic Philosophy 2Google Scholar ed. R. E. Allen and D. J. Furley (Atlantic Highlands, 1975), p. 279, who attributes the error to a misunderstanding of Arist. Metaphys. 984a 14 by later writers (e.g. Theophrastus); also Mathewson, R., ‘Aristotle and Anaxagoras’, CQ n.s.8 (1958), 77–8CrossRefGoogle Scholar, Guthrie, , op. cit. p.293 and n. 2Google Scholar, Rösier, , loc. cit. 59–60Google Scholar. But the Aristotelian evidence contrasting the elements of Empedocles with those of Anaxagoras is highly problematical; for further discussionsee, e.g., Vlastos, G., ‘The physical theory of Anaxagoras', Studies in Presocratic Philosophy 2 (see above) 339–40, 352 n. 84Google Scholar, McDiarmid, J. B., ‘Theophrastus on the Presocratic causes’, H.S.C.P. 61 (1953), 111f.Google Scholar, Kerferd, G. B., ‘Anaxagoras and the concept of matter before Aristotle’, Bulletin of the John Rylands Library 52 (1969)CrossRefGoogle Scholar, reprinted in The Pre-Socratics: a Collection of Critical Essays, ed. A. P. D. Mourelatos (New York, 1974), pp. 494 f., M. Schofield , An Essay on Anaxagoras (Cambridge, 1980), pp. 125 f., who arrives at the conclusion that Anaxagoras counted among basic ingredients the ‘opposites … air, aither, earth…and the seeds of animals and plants and (probably) their tissues’ (p. 132), to which he also adds water.

³⁶ Munro on 1. 875–9; cf. Bailey, op. cit. pp. 41 f. The technique of first presenting an opponent's view crudely and then in a more sophisticated form, as if designed to defend the initial crudity, is a common polemical tactic: cf., e.g., the progression of arguments against the Epicurean doctrine that ‘all sensations are true’ in Plut. adv. Col. 1109 f. Also traditional is the technique of creating a dilemma and driving the opponent to resort to further improbabilities, from which ridiculous consequences are inferred; see Kleve, loc. cit. 58–63, a general discussion of Lucretius' polemical tactics. Although Lucretius clearly distorts Anaxagoras, this is not the place to tackle the thorny problem of what exactly was meant by ‘everything in everything’ for recent accounts of his theory of matter see Guthrie, , op. cit. pp. 279–94Google Scholar (with bibliography on 280 n. 1) and Schofield, , op. cit. pp. 100–44Google Scholar.

³⁷ See, e.g., Bailey, op. cit. pp. 36–7 and Commentary, ii. 743, 752; Mathewson, loc. cit. 81; Boyancé, op. cit. p. 105; Rankin, H. D., ‘Lucretius on “Part of Everything is in Everything”’, AC 38 (1969), 158–61Google Scholar.

³⁸ There is also an aural play upon latices and lactis (886–7), and alliteration of fr in 881–2, 888–9, 892. For discussions of the ignis/lignum pun see Friedländer, P., ‘Pattern of sound and atomistic theory in Lucretius’, AJP 62 (1941), 17Google Scholar; West, , op. cit. p. 97Google Scholar; Snyder, , op. cit. 41–2, 130–3Google Scholar.

³⁹ Rankin, , loc. cit. 160–1Google Scholar.

⁴⁰ West, , op. cit. pp. 24–5Google Scholar. Leonard-Smith seem to suggest that the example of spontaneous forest fires goes back through Thucydides (2. 77) to Anaxagoras.

⁴¹ These lines are very similar to 2. 976–7 and were condemned by Forbiger, who has found a few supporters (see Merrill's note), of whom K. Müller is the most recent. But it is quite in Lucretius' manner to conclude with a reductio ad absurdum, and the grotesque parody of homoeomeria makes a perfect climax to the whole critique. Of course, Lucretius also wishes to convey a serious epistemological point about the limitations of arguing from the visible to the invisible (Lenaghan, , loc. cit. 235Google Scholar).

⁴² Lucretius' polemical tendency to abandon the plane of logic for the realm of the extravagant and grotesque has been excellently observed by L. Perelli, with reference to the Anaxagorean passage: see Lucrezio, : poeta dell' angoscia (Florence, 1969), pp.330–1Google Scholar, where the Lucretian technique is likened to surrealism.

⁴³ See Meillet, A. Ernout-A., Dictionnaire étymologique de la langue latine (Paris, 1959), s.v. scrutaGoogle Scholar.

⁴⁴ cf. Hor, . Sat. 2. 3. 276, Ep. 1. 18. 37Google Scholar, Ov, . Met. 15. 137Google Scholar; in the first century A.D. it became more popular, as the following sample of numbers shows: Manil. 7, Sen, . Epigr. and Tr. 7Google Scholar, Luc. 6, Stat. 14, Sil. 1, Juv. 2.

⁴⁵ op. cit. 125. For uses with a comparable force cf., e.g., Cic, . N.D. 3. 42Google Scholar (ii qui interiores scrutantur et reconditas litteras), de Or. 2. 146 (scrutari locos, ex quibus argumenta eruamus), Quint. 1. 4. 25, 5. 14. 28, Tac, . Ann. 16. 5Google Scholar. From a somewhat different point of view, Lenaghan describes it as ‘a businesslike and appropriate word’ in the light of the close relation of Anaxagoras to atomism (loc. cit. 233).

⁴⁶ Munro (on 834), Bailey (on 830, and op. cit. pp. 551–6) and others judge it to be authentic, and G. Vlastos does not reject the possibility (loc. cit. 332, 349 nn. 65–6); most modern scholars, however, are opposed to this view: see Kirk-Raven, , op. cit. pp. 386–8Google Scholar, Mathewson, , loc. cit. 77–81Google Scholar, Guthrie, , op. cit. pp. 325–6Google Scholar, Rösier, , loc. cit. 58–9Google Scholar, Schofield, M., ‘Doxographica Anaxagorea’, Hermes 103 (1975), 4–6Google Scholar, op. cit. 153 n. 39. Most probably, the adjective ⋯μοιομερ⋯ς was coined and first applied to Anaxagoras by Aristotle. The noun is found first in Epicurus (fr. 24. 33. 2, 29. 27. 7, 30. 7. 5, 12. 5, 28. 4 Arrighetti, and cf. Diels, , Dox. 307 ab 3Google Scholar, discussed by Bignone, E., Boll. di Fil. Class. 17 [1910–11], 135–8; 26 [1919–20], 60–3)Google Scholar, used apparently in a technically Epicurean way and without reference to Anaxagoras, but in the same abstract sense of the singular – ‘the attribute of being homoeomerous’ – as in Lucretius (the plural is used in D.O. fr. 5, Col. 2. 6 Chilton and elsewhere to describe the physical elements of Anaxagoras). It is possible, therefore, that the noun was an Epicurean term, which Lucretius may have found applied to Anaxagoras in his Epicurean source; see further, Lanza, D., PP 18 (1963), 281–9Google Scholar and Schofield, , loc. cit. 5–6Google Scholar. The specialized Epicurean usage seems to have escaped Lucretius, who treats the word as a technical term belonging to Anaxagoras alone.

⁴⁷ 3. 131–5 (redde harmoniai | nomen, ad organicos alto delatum Heliconi; | sive aliunde ipsi porro traxere et in illam | transtulerunt, proprio quae tum res nomine egebat. | quidquid 〈id〉 est, habeant), where see Kenney's notes.

⁴⁸ cf. also 3. 260, where Lucretius again stresses the difficulty (261); for other references to Greek terminology cf. 3. 100, 6. 424, 6. 908. For another translation into Latin of the concept of homoeomeria, cf. Cic, . Ac. 2. 118Google Scholar.

⁴⁹ His practice extends from the repetition of one word in the same line to the complicated interweaving of several words over a number of lines, and ultimately to the quasi-formulaic use of repeated phrases, lines and passages throughout the poem. For general discussion see Bailey's, Commentary, i. 155–8, 161–5Google Scholar; Deutsch, R. E., ‘The Pattern of Sound in Lucretius’ (Diss. Bryn Mawrt, 1939)Google Scholar; Ingalls, W. B., ‘Repetition in Lucretius’, Phoenix 25 (1971), 227–36CrossRefGoogle Scholar; Minyard, J. D., ‘Mode and value in the De Rerum Natura’, Hermes Einzelschriften 39 (Wiesbaden, 1978)Google Scholar. For his use of close or directly juxtaposed repetitions, such as abound in the present passage, see Merrill on 3. 71; Breazeale, E., ‘Polyptoton in the hexameters of Ovid, Lucretius and Vergil’, Studies in Philology 14 (1917), 306–18Google Scholar; Deutsch, , op. cit. ch. 3, esp. pp. 22 f., 25 f., 29Google Scholar; Rozelaar, , op. cit. p. 37Google Scholar; West, , op. cit. pp. 119–20Google Scholar, and note the parody of Lucretius in Pers. 3. 83–4 (aegroti veteris meditantes somnia, gigni | de nihilo nihilum, in nihilum nil posse reverti, cf. Lucr. 1. 150, 155–6, 237), which is mentioned by Alfonsi, L. in ‘Lucrèce’, Entretiens Hardt (n. 1 above), p. 293Google Scholar. Some very similar expressions in Book 2 seem to hark back to the Anaxagorean passage: cf. 2. 731 f. (ne forte haec albis ex alba rearis | principiis esse, ante oculos quae Candida cernis, | aut ea quae nigrant nigro de semine nata), 790–1, 823–4, 902–3, 930, 932, and see Schrijvers, , op. cit. p. 225Google Scholar.

⁵⁰ Note also the careful organization, whereby three examples drawn from the body (835–7) and three ‘Empedoclean’ elements (840–1) are deployed around the mineral gold (839–40).

⁵¹ Deutsch, , op. cit. pp. 76–7, also 135–6Google Scholar, where she shows that the whole passage 830–74 abounds in rhyme at the beginning and end of the verse.

⁵² Deutsch, , op. cit. p. 76Google Scholar.

⁵³ Note on 835, along with a comment on the effective placement of the repetitions at the verse-beginning or after a caesura; cf. also Paratore, E. and Pizzani, U., Lucreti De Rerum Natura (Rome, 1960), pp. 182–3Google Scholar, who note the use of repetition throughout the Anaxagorean section to aid the ‘slow and cautious’ progress of the argument.

⁵⁴ op. cit. pp. 236–7 and n. 2; also Paratore-Pizzani, , op. cit. p. 182Google Scholar, who comment on Lucretius' constant reference to sense-data and his fascination with the infinitely small.

⁵⁵ ‘Sometimes the poet has said well and accurately just what he wants to say, and when he wishes to say it again he does not hesitate to use the same phrase again’ (Commentary, i. 161, cf. 157).

⁵⁶ op. cit. p. 118.

⁵⁷ n. 35 above.

⁵⁸ cf. also 1. 894–6.

⁵⁹ See Giussani on 814, Bailey on 814–16, West, , op. cit. p. 119Google Scholar. The unusual extent and variety of the repetition is shown by Deutsch, , op. cit. p. 45Google Scholar.

⁶⁰ cf. the repetition of putarunt at the end of 1. 705 and 708, where Lucretius is reporting the views of monists, and of putabant at the end of adjacent lines in 5. 1178–9, where the subject is superstitious belief. For further uses of the word to distance the author from the unscientific beliefs of others, cf. especially the formulae si (forte) putas, ne (forte) putes, and, e.g., 1. 635, 3. 801, 4. 1236, 5. 22, 159, 1041, 6. 851.

⁶¹ ‘Connotes falsity’, according to Leonard-Smith (on 1. 104), and is ‘one of Lucretius' favorite verbs for denoting futile or deliberately deceitful rationalization’, in the words of Lenaghan, , loc. cit. 233 n. 41Google Scholar, who comments on the accumulation here of deprecatory vocabulary: ‘putat (839, 842, 877), fingit (842, 847), nec concedit (843–44) and sumit (876) describe Anaxagoras' somewhat arbitrary thought process, which takes him far from the truth (846, 880)’ (loc. cit. 233).

⁶² On the repetitive and polemical nature of these objections to Anaxagoras see Lenaghan, , loc. cit. 233 f.Google Scholar, Rösler, , loc. cit. 60Google Scholar.

⁶³ cf. Juv. 3. 288–9 (miserae cognosce prohoemia rixae, | si rixa est, ubi tu pulsas, ego vapulo tantum).

⁶⁴ The commentators quote no parallel except for Dante, , Purg. 7. 32Google Scholar, but cf. also Sen, . H.F. 555Google Scholar (mors avidis pallida dentibus). Duff seems right to identify the image as one of ‘a devouring animal’, and cf.Leonard-Smith, and West, , op. cit. p. 130 n. 11Google Scholar; O. Regenbogen uses the line rather extravagantly to exemplify Lucretius' primitive horror of death, and refers to archaic representations of demons: ‘Lukrez. Seine Gestalt in seinem Gedicht’, Neue Wege Zur Antike, 2. 1 (Leipzig and Berlin, 1932), 52Google Scholar. One suspects that Lucretius has tailored the image ironically to suit the examples of blood, bones and flesh (835–8, 853), which we are to imagine being torn up by ‘the teeth of death’.

⁶⁵ For this technique cf., e.g., 4. 118 f., 5. 43 f., 220 f., 6. 390 f.

⁶⁶ i.e. in 1. 149–214. Cf. the parallel passages in the criticisms of Heraclitus (672–4) and Empedocles (756–8).

⁶⁷ 59 B 10 DK, Arist, . G.A. 723 a 11Google Scholar, Simpl, . Phys. 460. 15–19 (59 A 45 DK)Google Scholar, Aet. 1. 3. 5 (59 A 46 DK). See further Bailey, , op. cit. pp. 35 f.Google Scholar, Cornford, , loc. cit. 277 f., 282 f.Google Scholar, Vlastos, , loc. cit. 325Google Scholar, Guthrie, , op. cit. pp. 287 fGoogle Scholar. But Schofield disputes the emphasis of Aristotle and the doxographical tradition on biological considerations as a foundation for the theory of ‘everything in everything’: op. cit. pp. 106, 121 f., 129 f., 133 f., 136–43. On the authenticity of B 10 see further n. 83.

⁶⁸ Adequate sense is restored by Lambinus' stopgap, et nervos alienigenis epartibus esse: see Bailey's note.

⁶⁹ The text is that of Bailey's Commentary, in which he adopted Diels' transposition and lacuna – a solution also accepted by, e.g., Leonard-Smith, , Pizzani, U., Il problema del testo e della composizione del De rerum natura di Lucrezio (Rome, 1959), pp. 61–3Google Scholar, and M. F. Smith; the deletion of 873 (Lambinus) produces the same repetition of ex alienigenis in successive lines, which Diels compares to other such emphatic repetitions in Lucretius: cf., e.g., 2. 955–6, 3. 12–13, 4. 789–90, 5. 298–9, 950–1, 6. 528–9, 1168–9. Other solutions include the deletion of 873 and 874 (Marullus), emendation of 874 (Lachmann) and the placing of a lacuna between 873 and 874 (Munro).

⁷⁰ Deutsch, , op. cit. p. 35Google Scholar.

⁷¹ ‘The recurrence of ex alienigenis in four lines so close together…is most emphatic’ (Deutsch, , op. cit. p. 136Google Scholar). The adjective is the Lucretian antithesis to ⋯μοιομερ⋯ς, as Munro and Bailey observe (on 865).

⁷² Such as cibus (859, 861, 864), corpus (859, 861, 862, 867), auget, alit (859, 874), (com)mixto (861, 866), constare (865, 868), necessest (868, 872); in 863 Lucretius might have used sanguis (cf. 860) instead of cruor had it been convenient. Moreover, if there is a lacuna after 873 or 874, it may well be because ‘the verses lost were so similar in form to some of the lines in 867–872 that the copyist overlooked them’ (Merrill on 873) – an explanation which could also be applied to the lacuna after 860.

⁷³ For Greek see Kühner, R.–Gerth, B., Ausführliche Grammatik der griechischen Sprache: Satzlehre (Hannover an d Leipzig, 1904), ii. 602Google Scholar, and especially Gygli-Wyss, B., ‘Das nominale Polyptoton im älteren Griechisch’, Zeitschrift für vergleichende Sprachforschung auf dem Gebiet der indogermaniscken Sprachen, Ergänzungshefte 18 (Göttingen, 1966)Google Scholar, who demonstrates its popularity with the tragedians and other writers influenced by rhetoric.

For Latin see Landgraf, G., ALL 5 (1888), 161 f.Google Scholar; Kühner, R.–Stegmann, C., Ausführliche Grammatik der lateinischen Sprache: Satzlehre (Hannover, 1914), ii. 617–18Google Scholar; Leumann, M.–Hofmann, J. B.–Szantyr, A., Lateinische Syntax und Stilistik (Munich, 1965), ii. 707–8Google Scholar. On Lucretius see n. 49 above, especially Breazeale's article, which includes Virgil and Ovid; for Virgil see also Pease, on Aen. 4. 83Google Scholar, and for Ovid, , Bömer, on Met. 5. 300, 8. 27Google Scholar.

⁷⁴ cf. Il. 13. 130–1, 16. 215, and Latin imitations such as Enn, . Ann. 572Google Scholar, with Vahlen, Furius, fr. 10 M, Virg, . Aen. 10. 361Google Scholar, Ov, . Met. 9. 44–5, withGoogle ScholarBomer, , Sen, . Ag. 498, Luc. 1. 6–7Google Scholar, Stat, . Th. 8. 398–9Google Scholar, Sil. 4. 352–3, 9. 322–5, also Liv. 33. 8. 14, 38. 17. 8, etc. As used by Hesiod, the device has a proverbial ring: cf., e.g., Erg. 23, with West, 25–6.

⁷⁵ ‘The obvious supplement’ (Bailey).

⁷⁶ Deutsch, , op. cit. p. 29Google Scholar (though the same claim could be made for 1. 835 f. with equal justification).

⁷⁷ See, e.g., the notes of Munro, , Ernout-Robin, , Leonard-Smith, , Bailey, , Kranz, , loc. cit. 368, 375Google Scholar and Bollack, . loc. cit. 683Google Scholar.

⁷⁸ See Arist, . G.C. 333 a 35 f.Google Scholar, and, for this and other Empedoclean polyptota, Gygli-Wyss, , op. cit. p. 106Google Scholar.

⁷⁹ For discussion of the extensive polyptoton in this fragment see Gygli-Wyss, , op. cit. pp. 57–8Google Scholar, and for other polyptota in Democritus, 131–2. The many philosophical variations upon the like-to-like principle naturally lent themselves to the use of this figure: cf. the Empedoclean fragments quoted above, and see further Gygli-Wyss, , op. cit. pp. 58–9Google Scholar, with the works referred to on 58 n. 2.

⁸⁰ For Presocratic use of polyptoton see further Gygli-Wyss, , op. cit. pp. 105 f., 130 fGoogle Scholar.

⁸¹ For Epicharmus see Berk, L., Epicharmus (Groningen, 1964), esp. pp. 93–5 (B 3) and 98–9 (B5)Google Scholar; the former ‘must be a forgery’, while the latter ‘may be genuine’ (p. 158); also Pinto, M. L. Silvestre, ‘Note sul pensiero filosofico di Epicarmo’, AAN 88 (1977), 237–59Google Scholar. For his use of polyptoton see Gygli-Wyss, , op. cit. pp. 106–7Google Scholar.

For Philolaus see, e.g., Kirk-Raven, , op. cit. pp. 308 f.Google Scholar, who, unconvinced by Mondolfo's defence, reject the fragments on the basis of research by Bywater and Frank that reveals a suspicious similarity to Aristotle's accounts of Pythagoreanism; also Guthrie, , op. cit. (Cambridge, 1962), i. 330–3Google Scholar, who takes a neutral position, and von Fritz, K., RE Suppl. 13 (Munich, 1973), 456–62, who is selectiveGoogle Scholar.

⁸² cf. similar turns of phrase, all referring to Anaxagoras, in Arist. Phys. 203 a 24 (59 A45 DK), Simpl, . Phys. 164. 20–1 (see B3 DK), 460. 12–13, 17, 461Google Scholar. 7–8 (A45 DK); note also Lucr. 1. 876–7 (id quod Anaxagoras sibi sumit, ut omnibus omnis | res putet immixtas rebus latitare). Polyptota involving π⋯ς are characteristic of the Presocratics, who were the first to exploit them: see Gygli-Wyss, , op. cit. pp. 44–6Google Scholar.

⁸³ See Gygli-Wyss, , op. cit. p. 131Google Scholar. Ernout-Robin (on 834 ff.) point out a similarity to Lucr. 1. 837 (visceribus viscus gigni etc.). Rösler, however, casts doubt upon whether ‘Fragment’ 10 contains the original wording of Anaxagoras, (loc. cit. 60)Google Scholar. A full discussion of the question is conducted by Schofield, , loc. cit. pp. 14 f.Google Scholar, op. cit. pp. 136 f., who concludes that, although the doctrine of growth may derive from Anaxagoras, we cannot be sure whether the scholium preserves his ipsissima verba; see also Sider, D., ‘The fragments of Anaxagoras’, Beiträge zur klassischen Philoiogie 118 (Meisenheim am Glan, 1981), 89–90Google Scholar, who argues strongly in favour of genuineness.

⁸⁴ cf. Plat, . Phaed. 96dGoogle Scholar (⋯πειδ⋯ν γ⋯ρ ⋯κ τ⋯ν σιτ⋯ων ταῖς μ⋯ν προσγ⋯νωνται, τοῖς δ⋯ κα⋯ οὕτω κατ⋯ τ⋯ν αὐτ⋯ν λ⋯γον κα⋯ τοῖς ἄλλοις τ⋯ αὐτ⋯ν οἰκεῖα ⋯κ⋯στοις προσγ⋯νηται…), which is ‘a clear reference to Anaxagoras’ (Guthrie, , op. cit. ii. 287 n. 1Google Scholar), , Arist.G. A. 723A 11Google Scholar (Ἀναξαγ⋯ρας μ⋯ν γ⋯ρ εὐλ⋯γως ϕησ⋯ ⋯κ τ⋯ς τροφ⋯ς προσι⋯ναι ταῖς Theophr, . ap. Simpl, . Phys. 27. 12–15Google Scholar (59 A41 DK) (⋯κεῖνος γ⋯ρ ϕησιν ⋯ν τῇδιακρ⋯σει το⋯ ⋯πε⋯ρου τŰ συγγεν⋯ ϕ⋯ρεσθαι πρ⋯ς ἄλληλα, κα⋯ ὅτι μ⋯ν ⋯ν τῷ παντ⋯ ἦν, γ⋯νεσθαι , ὅτι δ⋯ ⋯μο⋯ως δ⋯ κα⋯ τ⋯ν ἄλλων ἒκστον, ὡς οὐ γινομ⋯νων ⋯λλ' ⋯νυπαρχ⋯ντων πρ⋯τερον).

⁸⁵ Vlastos seems to imply direct use of Anaxagoras when he refers to concrescere (840) as a ‘Latin rendering for Anaxagoras' συμπηγν⋯ναι (B16)’, comparing Cic, . Ac. 2. 100Google Scholar (loc. cit. 349 n. 69, and cf. n. 70). Contrarily, Cornford believes the source to be ultimately Theophrastus, (loc. cit. 294)Google Scholar, while Rösler suggests an Epicurean work based on Theophrastus, (loc. cit. 59–60, 62)Google Scholar.