^* An earlier version of this article was delivered as lectures between February and May 1986 at the Universities of Michigan, California at Irvine, Lille, Innsbruck, Milan and Rome; I wish to thank my hosts and audiences at those institutions, and particularly L. Koenen and R. Scodel, as well as several friends who read the article in manuscript, especially R. Kannicht and T. Rosenmeyer, for their extremely helpful suggestions and criticisms.

¹ Lobel, E.et al., The Oxyrhynchus Papyri XXIV(London, 1957), pp. 52–5.Google Scholar

² Calame, C., ed., Alcman(Rome, 1983), pp.104–7.Google Scholar In Page, D. L., ed., Poetae Melici Graeci (Oxford,1962), this is Fr. 5.2,1.229, II. 129. Hereafter, fragments of Alcman are cited by their number in Page's edition, followed by ‘ = ’ and their number in Calame's.Google Scholar

³ Lobel, op. cit. (n. 1), p. 55.

⁴ Barrett, W. S.,Gnomon 33(1961),682–92Google Scholar, here 689; Page, D. L.,CR 9(1959),15–23, here 21.Google Scholar

⁵ Lloyd-Jones, H.,apudC. M. Bowra, Greek Lyric Poetry from Alcman to Simonides2(Oxford, 1961),26Google Scholarn. 1. Cf.Penwill, J. L., ‘Alkman's Cosmogony’, Apeiron 8(1974),13–39,CrossRefGoogle Scholar here 26–7; Treu, M., ‘Licht und Leuchtendes in der archaischen griechischen Poesie’, Studium Generate 18 (1965),83–97,Google Scholar here 86; West, M. L.,‘ThreePresocraticCosmologies’,CQ 13(1963),154–76, here 154f.CrossRefGoogle Scholar

⁶ Burkert, W., rev. H. Frankel,Dichtung und Philosophie desfriihen Griechentums, Gnomon 35 (1963),827–Google Scholar8, here 827. Cf. Garzya, A., Studi sulla lirica greca da Alcmane al primo Impero (Messina-Florence,1963), here23Google Scholar; Apicella, G., ‘La cosmogonia di Alcmane’, QUCC 32(1979),7–27Google Scholar, here 13ff.; Vernant, J.-P.,‘Thetis et le poeme cosmogoniqued’Alcman‘, pp. 38–69 in Hommages a Marie Delcourt(Brussels,1970),Google Scholar here 41ff. (reprinted in Detienne, M. etJ.-P. Vernant, Les Ruses de Vintelligence. La metis des Grecs[Paris,1974],134–Google Scholar64);West, M. L., ’Alcman and Pythagoras‘, CQ 17(1967),1–15, here 3, 5–6.CrossRefGoogle Scholar

⁷ Besides the works cited in notes 5, 6, and 11, cf. especially Frankel, H., Dichlung und Philosophie desfriihen Griechentums. Eine Geschichte der griechischen Epik, Lyrik und Prosa bis zur Mine desfunften Jahrhunderts³(Munich,1969), pp.183Google Scholarf., 290; and West, M. L.Early Greek Philosophy and the Orient(Oxford,1971), pp.206–8Google Scholar; also Y. Hirokawa, ‘Alcman as One of the Forerunners of Philosophical Cosmogonists’, Journal of Classical Studies [Kyoto] 20 (1972), 40–8 and Jelnickij, L. A., ‘The Origins of the Ancient Etruscan Cosmology’, Veslnik Drevnej Istorii 140(1977),121–8.Google Scholar

⁸ Vernant, op. cit. (n. 6), 41

⁹ Cf. Voelke, A. J., ‘Aux origines de la philosophie grecque: La cosmogonie d'alcman’, pp. 13–24 in Metaphysique. Histoire de la philosophie. Festschrift F. Brunner(Neuchatel,1981), here 13–14.Google Scholar

¹⁰ Kirk, G. S., J. E. Raven and M. Schofield, The Presocratic Philosophers2(Cambridge, 1983), pp.47–9.Google Scholar

¹¹ Segal, C., in P. E. Easterling and B. M. W. Knox, ed., The Cambridge History of Classical Literature, i: Greek Literature(Cambridge,1985), p.179.Google Scholar

¹² Calame, op. cit. (n. 2), pp. 437–8.

¹³ FGrHist 328 T 1 (Testim. 41 Calame).Google Scholar

¹⁴ FGrHist 595 F 6 (Testim. 42 Calame).Google Scholar

¹⁵ FGrHist 273 F 95, 96 (Testim. 43 Calame).

¹⁶ Testim. 4, 31, 34 Calame. Evidently, the many expressions of homoerotic sentiments on the part of the choruses of maidens in his poems (e.g. 1 = 3 passim, 3.3 = 26.61f., 37[a] = 151, 58 = 147, 59[a] = 148, 59[b] = 149, 81[b] = 150) were misconstrued as statements of personal involvement on the part of the poet himself: so Calame, op. cit. (n. 2), p. xx; of little if any value are the speculations ofSirna, F. G., ‘Alcmane ’,Aegyptus 53 (1973),28–70. This is of course a typical strategy of ancient literary criticism: cf. below n. 22.Google Scholar

¹⁷ Met. A.3.983b27–33,4.984b23–31,8.989a9–l 1. Aristotle's silence is pointed out by Penwill, op. cit. (n. 5), 13, and Ricciardelli, op. cit. (n. 6), 7f.

¹⁸ As suggested by Penwill, op. cit. (n. 5), 13.

¹⁹ Hist. An. 5.31.557a2.

²⁰ P. Oxy. 2389, Fr. 9, 1.12 = PMG 13 Page.

²¹ So e.g. Calame, op. cit. (n. 2), p. 357; and Pfeiffer, R.,History of Classical Scholarship from the Beginnings to the End of the Hellenistic Age(Oxford, 1968), p.242.Google Scholar

²² Cf. in general Mary Lefkowitz, The Lives of the Greek Poets(Baltimore,1981).Google Scholar

²³ Stesichorus: Hist. An.5.9.542b25; Rhet.2.20.1393b9, 21.1395al, 3.11.1412a22. Sappho: Rhet.1.9.1367a8, 2.23.1398b2, 27. Alcaeus: Pol.3.14.1285a7, Rhet.1.9.1367a9, Fr. 75 Rose3. Simonides: 16 references. Pindar: Rhet.1.7.1364a28, 2.24.1401a7, Fr. 75 Rose3. Aristotle seems never to refer to Ibycus, Anacreon, or Bacchylides.

²⁴ Poet. 1.1447b 13–20.

²⁵ PMG 842.

²⁶ Diels, H., Doxographi Graeci(Berlin,1879), 386 test. b6; cf. e.g. 590.22ff., 610.1 Iff. In contrast, Pindar did write a cosmogonic hymn in which he recounted the creation of the world and the birth of the Muses (Hymn 1, cf.Google ScholarSnell, B., Die Entdeckung des Geistes. Studien zur Entstehung des europdischen Denkens bei den Griechen1[Gdttingen, 1975], pp.83ff.): and Theophrastus cited a fragment of it in his (Pind. Fr. 33eSn.-M. = Theophr. Phys opin. Fr. 12, Dox. Gr. 486.27–487.5 Diels.Google Scholar

²⁷ Cf. in general Schwabl, H., ‘Weltschopfung’, RE Suppl.9(Stuttgart,1962)Google Scholar; and Guthrie, W. K. C., A History of Greek Philosophy, i: The Earlier Presocratics and the Pythagoreans (Cambridge,1962), pp. 142ff.; also Penwill, op. cit. (n. 5), 13 and 31 n. 4.Google Scholar

²⁸ Metaph. A.2.984a8ff.

²⁹ 59 B12 D-K.

³⁰ 64 B3 D-K.

³¹ Mem. 1.4.5, 7; 4.3.13.

³² Cf. Guthrie, op. cit. (n. 27), v:The Later Plato and the Academy(Cambridge, 1978), pp.Google Scholar

³³ Tim. 28aff.

³⁴ SVF 1.85ff., 2.299ff., 579ff.

³⁵ It is symptomatic that most of the mysteries of the Louvre Partheneion are due, not to any obscurity in the poet's language, but rather to our ignorance of the ritual which that poem accompanied; this is quite misunderstood by Penwill, op. cit. (n. 5), 14.

³⁶ Calame, C.,Les Choeurs de jeunes filles en Grece archaique. i: Morphologie, fonction religieuse et sociale. n: Alcman(Rome, 1977)Google Scholar;idem ed., Rito e poesia corale in Grecia. Guida storica e critica (Ban, 1977), e.g. p. 112; cf. idem, op. cit. (n. 2), pp. xvi-xxi.Google Scholar

³⁷ Thus, in the Louvre Partheneion, the best preserved of Alcman's poems, the highly fragmentary first twelve lines of the papyrus recounted the story of the combat between the Tyndarids and the Hippocoontids (1.1–12 = 3.1–12), and the scholarly consensus is that the animosities were almost certainly due to the erotic rivalry between these heroes reported by other sources (cf. D. L. Page, Alcman. The Partheneion [Oxford, 1951], pp. 31–3; Calame, op. cit. [n. 2], p. 313, and op. cit. [n. 36], 2.52ff., especially pp. 55–8); and after the gnome there was a second myth of which even less can be read (22–35), but still enough to permit most recent scholars to agree that Alcman told of the destruction of Otos and Ephialtes, punished among other reasons for their attempt to rape Artemis (cf. especially Janni, P.,La cultura di Sparta arcaica. Richerche i[Rome,1965], pp.69–71; Calame, op. cit. [n. 2], pp. 320–1, with further references). So too, in other poems Alcman recounted the combats between the Dioscuri and the Apharetids, rivals for the daughters of Leucippus (8.1–6 = 20, and probably 5.1= 79), and the combats between the Dioscuri and Aphidnos arising from the attempt of the Dioscuri to rescue Helen after she had been abducted by Theseus (21 =210); other Alcmanic myths that may similarly have emphasised the theme of erotic violence are those of Paris (70[b] = 98, 77 = 97) and Circe (80 = 102). To be sure, Alcman seems also to have referred to apparently non-erotic myths on occasion: 5.49 = 83, 7 = 19, 50(b) = 124, 56 = 125, 68 = 95, 74 = 101, 79 = 100, 87(d) = 103; but none of these is certain to have been the central myth of one of his partheneia.Google Scholar

³⁸ On the pedagogic function of Alcman's myths, cf. Calame, op. cit. (n. 36), 1.41 Of.

³⁹ Cf. Calame, op. cit. (n. 2), pp. 431, 442.

⁴⁰ So Calame, p. 442.

⁴¹ So Calame, loc. cit.

⁴² E.g. Barrett, loc. cit. (n. 4); Frankel, op. cit. (n. 7), p. 290 n. 2; Lobel, op. cit. (n. 1), p. 55; Page, loc. cit. (n. 4); Ricciardelli, op. cit. (n. 6), 19; Schwabl, op. cit. (n. 27), 1467; West, op. cit. (n. 6), 4.Google Scholar

⁴³ Harvey, F. D., ‘Oxyrhynchus Papyrus 2390 and Early Spartan History’, JHS 87(1967), 62–73, here 69–70, criticised Lobel's and proposed instead ; parallels in support of Lobel's supplement are provided by Ricciardelli, op. cit. (n. 6), 7 n. 2. In terms of the argument of the present article, either supplement is acceptable; but the parallels offered below may make slightly preferable.CrossRefGoogle Scholar

⁴⁴ He also occasionally uses such evidently synonymous expressions as , cf.Bonitz, H., Index Aristoteticus (Berlin, 1870), ss.vv.Google Scholar

⁴⁵ Cf.Idem.; and W.D., Ross, ed., Aristotle's Metaphysics(Oxford, 1924), 1.130 ad A.3.983b29Google Scholar

⁴⁶ Cf. also : Plut. De Daed. Plat. 4, Schol. Arat. 1 (40.6ff. Martin). : Schol. Pind. N. 4.101b, 107b. Schol. Pind. /. 7.3a. Heraclit. 16.5. vciKr) avoSocic:Philo, De fuga et inventione108 (3.133.11 Cohn-Wendland). physica ratio: SVF2.313.11. physici:Servius ad Aen.1.47. vcioXoyia:Heraclit. 72.1. vcioyvaifioviw.Philo, Desomniis1.164(3.240.2). vcioXoyecu:Diod. Sic. 3.62.3, Philo, Demutatione nominum62 (3.168.9), Schol. Ap. Rhod. 1.1098–1102a. This usage is also extremely frequent among the Church Fathers, e.g. Tertullian, Ad nat. {physiologice2.12.17; argumentations physiologicae2.4.13; physicum theologiae genus2.12.14), Eusebius, Praep. ev. (va.oXoyla3 Praef. 1, 5; 2.1, 3; 6.2, 7; EVCIKT) deoXoyla3.1; vciKWTepa deoXoyla3.6.1; 3.15.1), Augustine, De civ. Dei (interpretationes physicae7.5; interpretationes physiologicae7.5; interpretations naturales7.18; physiologice7.27.2; naturales rationes7.33).

⁴⁷ Cf. also (pvciKwc:Heraclit. 8.5, 15.2, 43.7, 46.1, 66.10; Philo, Legum alleg.2.5 (1.91.11). vctKu>Tepoc:Heraclit. 25.1, 56.1; Philo, Legum alleg.1.37 (1.70.11). physice:Servius ad Aen.10.5. VCIKOI: Heraclit. 22.2. VCIKOI deiopia:Heraclit. 25.12, 36.1. Aoyoc UCIKOITQTOC: Philo, Quod deus sit immutabilis 11(2.73.19), De plantatione120 (2.157.8). physica ratio: Servius ad Aen.1.52, 78, 142. frvaoXoyew: Georg. Syncel., Eclog. Chron.30.9–10 Mosshammer (cf. Alex. Polyhist. FGrHist273 F 79). <^uctoAoyia: Philo, De somniis1.120 (3.230.24).

⁴⁸ Apparently the same work is referred to at 5KF2.212.38f. under the title .

⁴⁹ Cf. Buffiere, F., Les Mythes d'Homere et la pensee grecque(Paris, 1956), pp.155–86.Google Scholar

⁵⁰ Compend. Theol. 17 (27.2–17 Lang). Cf. Buffiere, op. cit. (n. 49), pp. 176–7.

⁵¹ Cf. Schol. ad II. 18.434a Erbse, Eustath. 1152.9f. ad loc.

⁵² Fr. 210 M.-W.

⁵³ N. 3.35, 4.62. Cf. Kaiser, J.,Peleus und Thetis. Eine sagengeschichtliche Untersuchung.(Munich,1912), pp.44–63.Google Scholar

⁵⁴ Soph. Fr. 150, 618 R a d t; Eur. Andr. 1253, 1A 705, 1040, Fr. 1093 Nauck².

⁵⁵ Hdt. 7.191; Xen. Cyn. 1.8; Apollodor. Bibl. 3.13.4f.; Ovid, Met. 11.237ff.; Val. Flacc. 1.130; Paus. 5.18.5; Quint. Sm. 3.618–24, 4.131 ff.; Tzetzes Schol. ad Lycophr. 175, 178, Chiliad. 4.519; Eustath. 1685.64 ad Od. 11.285.Google Scholar

⁵⁶ Cf.Brommer, F.,Vasenlisten zur griechischen Heldensage³(Marburg, 1973), pp.321–9Google Scholar, Denkmalerlisten zur griechischen Heldensage(Marburg, 1976), 3.365–7; Paribeni, E. inEnciclopedia delFarte antica classica e orientale(Rome, 1958–73)Google Scholar, s.v. Teti; Schneider, K.,Thetis im Verwandlungskampf mil Peleus in der griechischen Vasenmalerei(Breslau,1941).Google Scholar

⁵⁷ Rep. 2.38 lbff. (especially 381d5: Proteus and Thetis).

⁵⁸ In Remp. 1.109.20 (Proteus and Thetis), 112.14–113.19 (Proteus) Kroll.

⁵⁹ Schol. a d N. 4.101b.

⁶⁰ That this whole myth in Pindar's Fourth Nemean was most probably the subject of an extended physical allegorical interpretation is suggested by the passage in the subsequent lemma (Schol. a d N. 4.107b) in which another ‘more physical’ explanation is cited, this time so as to be refuted.

⁶¹ Cf. Schol. ad Od. 4.384, Eustath. 1503.7ff. adloc, Sextus Emp. Adv. Math. 9.5; and Buffiere, op. cit. (n. 49), pp. 179–86.Google Scholar

⁶² E.g. Burkert, op. cit. (n. 6), 828; already Lobel, op. cit. (n. 1), p. 55 n. 1; Penwill, op. cit. (n. 5), 13; West, op. cit. (n. 6), 4.Google Scholar

⁶³ Cf. 14(a) = 4; 8.7–11 = 2 1; 27 = 84; 28 = 85; probably 3.1 = 26.1–12; near the beginning, 30 = 86; probably Apollo, 51 = 1. So too, Pindar seems to have begun his partheneia with an invocation to the Muses (probably Fr. 94b Sn.-M.) or Apollo (probably Fr. 94c Sn.-M.).

⁶⁴ For parallels in Pindar, cf. the openings of P. 4, N. 3. 9; and, near the beginnings, O. 10.3f., /. 8.6f. That Alcman referred explicitly in this part of the poem to the chorus as being Dymainian is rendered probable by the commentator's discussion of the chorus immediately after his gloss on the poem's opening: for how else will he have known who the constituents of the chorus were unless Alcman had announced the fact, here as he does in other poems (10[b].89 = 82.8–9, 45.4 = 61.4), and why else should the commentator make mention of the fact at this point in his commentary?.

⁶⁵ Frankel, op. cit. (n. 7), p. 291 n. 4; Penwill, op. cit. (n. 5), 16.

⁶⁶ N. 4.62–5.

⁶⁷ A useful summary of the various traditions is found in Ninck, M.,Die Bedeutung des Wassers im Kult und Leben der Allen. Eine symbolgeschichtliche Untersuchung = Philologus Supplementband 14.2(Leipzig, 1921), pp. 138–80, especially pp. 161ff.Google Scholar

⁶⁸ This psychological meaning for seems otherwise not to be found before Aeschylus: Ag. 1216, Choe. 289. So too, Alcman's usage of the word has no parallels before Aeschylus: cf. n. 71 below.

⁶⁹ One more guess: perhaps the commentator's (III. 1516) refers to her regaining her familiar shape and identity after her metamorphoses.

⁷⁰ E.g. Frankel, op. cit. (n. 7), pp. 1845; Penwill, op. cit. (n. 5), 1724; Ricciardelli, op. cit. (n. 6), 1821; Vernant, op. cit. (n. 6), 4456; Voelke, op. cit. (n. 9), pp. 1821; West, op. cit. (n. 5), 155f.

⁷¹ Outside of Alcman, this meaning for seems not to be found before Aeschylus: PV 59, 111,477. But it is supported by such archaic words as ; and in any event it is guaranteed for Alcman by the Louvre Partheneion.

⁷² For ‘goal’, cf.//. 13.20, Pind. P. 2.49; for ‘termination’, Pind. Fr. 165 Sn.M. Other archaic usages of the word-‘fixed line of separation’ (Hes. Fr. 273.2 M.-W.), ‘sure sign or token’ (//. 1.526, Pind, N. 11.44) - have a no less obvious pertinence to Thetis‘ ceasing to transform herself and readopting her true identity.

⁷³ Heraclit. Quaest. Horn. 43.3, 7: n o t e in t h e Alcman papyrus (III.234) and in Heraclitus.

⁷⁴ For example, it is odd that the words return as part of the next lemma together with day and moon (26). Scholars are practically unanimous - rightly so, in my opinion - in deleting the second occurrence of the phrase as an interpolation: for the commentator's explanation at its first occurrence must mean that the darkness was described before the appearance of sun and moon: so most recently Calame, op. cit. (n. 2), p p. 44950. The only exceptions were early: Barrett, loc. cit. (n. 4); Page, op. cit. (n. 4), 20; Treu, op. cit. (n. 5), 86. The decisive arguments were formulated by West, op. cit. (n. 5), 156.

⁷⁵ Il. 10.252; Od. 12.312; Pind. P. 4.256.

⁷⁶ But I find no parallel for such a usage of .

⁷⁷ As far as I know, no ancient source reports the length of the wrestling-match between Peleus and Thetis; hence there are no close parallels in support of these hypotheses, but neither is there any evidence against them. Various, more general parallels for such triads are provided e.g. by Miiller, R., Die Zahl 3 in Sage, Dichtung und Kunst(Teschen, 1903)Google Scholar; Usener, H., ‘Dreiheit’,Rh Mus. 58(1903),1–47Google Scholar, 161208, 32162; cf. also Stith, Thompson,Motif-Index of Folk Literature. A Classification of Narrative Elements in Folktales, Myths, Fables, Mediaeval Romances, Exempla, Fabliaux, Jest-Books and Local Legends(Bloomington and London, 1966), H.1463, D.758.1, and especially Z.71.1, and cf. E.162.1, H.1472, and T.165.Google Scholar

⁷⁸ Some scholars have argued that this fragment too is cosmogonic and may belong in the same context as the poem discussed here: H. Lloyd-Jones apud West, op. cit. (n. 5), 156; Penwill, op. cit. (n. 5), 14; Treu, op. cit. (n. 5), 85. But, as Ricciardelli points out (op. cit. [n. 6], 26), the tone here is jocular: this fragment belongs much more in the context of Alcman's various ‘eating’ poems, e.g. 17 = 9, 19 = 11, 94 = 132, 95(b) = 92, 96 = 130.

⁷⁹ No ancient source corroborates such a suggestion; but Alcman was famous for his idiosyncratic versions of myths, cf. e.g. Page, op. cit. (n. 37), pp. 31ff., 44.

⁸⁰ As R. Nicolai (Rome) has suggested to me, these three allies of Peleus might be a divine version of the three mortal allies Menelaus uses to subdue Proteus (Od. 4.408–9, 433–4). In other regards too, Alcman's poem may well have been modelled fairly closely upon its Homeric analogue, cf. the use of a\Od. 4.373 and 466, and of at 4.417 to refer to the variety of forms Proteus will take on.

⁸¹ /. 8.44: .

⁸² Barrett, loc. cit. (n. 4).

⁸³ Cf. above n. 74.

⁸⁴ Jul. Gal. 357a; Dam. Princ. 213.

⁸⁵ Od. 8.265; Horn. H. Apollo 203.

⁸⁶ Cf. Schol. ad Theocr. 2.10b = Pind. Fr. 104 Sn.-M.

⁸⁷ E.g. 1.41, 6263 = 3.41, 6263; 3.3 = 26.66; cf. also perhaps 1.60 = 3.60.

⁸⁸ It may be of interest in this connection that Proclus reports that sun, moon and stars played an important role in the ceremony accompanied by another kind of partheneion, the daphnephoricon {apudPhotius, Bibl. 321b821).

⁸⁹ On this problem, cf. Calame, op. cit. (n. 36), 2.46ff., 86ff., 137ff.

⁹⁰ In the surviving part of the commentary on this poem, the commentator says nothing about its historical background; contrast the commentary, from the same papyrus, to 5.2 1.122 = 80. This may be due to our papyrus’ using different sources for the commentaries to the different poems; or, more likely, we may suppose that this poem, in the part whose commentary survives, gave no occasion for discussing historical matters. In the latter case our commentator displays a familiarity with a considerable variety of interpretative techniques which he adapts to the particular exigencies of the passages he is commenting on.

⁹¹ Cf. Calame. op. cit. (n. 2), pp. 400, 449.

⁹² E.g. 1.1321 =3.1321; 64= 105; 102= 108; 146= 106. Cf. in general A. Piatkowski, ‘Personificari si abstractii la Alcmana’, An. Univ. Bucurepi, Ser. ^tiinf. soc. filologie 9.18 (1960), 31924; and, on the Louvre Partheneion, Calame, op. cit. (n. 36), 2.59f.; Garzya, op. cit. (n. 6), pp. 214; Page, op. cit. (n. 37), pp. 337; C. O. Pavese, ‘Alcmane, il Partenio del Louvre’, St/CC 4 (1967), 11333, here 11620.Google Scholar

⁹³ We have no way of knowing for certain what kind of interpretation earlier scholars had given Alcman's myth of Thetis (pace Penwill, op. cit. [n. 5], 15); but the guess may be hazarded that, had their approach been allegorical, our commentator might not have simply dismissed their views.

⁹⁴ Schol. ad Pind. N. 3.60, 4.101b; Tzetzes ad Lycophr. 175; Paradox. Vat. 33

⁹⁵ Cf. Buffiere, op. cit. (n. 49), pp. 15565.Google Scholar

⁹⁶ As L. Koenen suggests, the cosmogonic implications of the prophecy that Thetis’ son would be stronger than his father (e.g. Pind. /. 8.26aff.) may have also helped lead the commentator in this direction.

⁹⁷ It is tempting to suggest that he may have been using uptodate scholarship: this papyrus can be dated with certainty to the second century AD. (Lobel, op. cit. [n. 1], p. 49), and Heraclitus’ Homeric Questions probably date from the reign of Augustus or Nero (F. Buffiere, ed., Heraclile. Allegories d'Homere [Paris, 1962], p. x): if the former is directly dependent upon the latter, it must have been composed not too long afterwards. But both texts may be drawing on a common source or sources, and the Alcman commentary may be earlier than Heraclitus. Yet it remains striking that the closest source for both the cosmogonic and the nocturnal parts of the Alcman commentary is Heraclitus.

⁹⁸ E.g. Phys. 2.3.194b236, 7.3.245b8f., Part. An. 1.1.640b25f., Gen. An. 1.18.724a236, Metaph. A.3.984a225, Pol. 1.8.1256a610.

⁹⁹ Cf. Voelke, op. cit. (n. 9), pp. 1415.

¹⁰⁰ So e.g. Harvey, op. cit. (n. 43), 62; Penwill, op. cit. (n. 5), 13, 26.

¹⁰¹ Cf. above n. 5

¹⁰² Cornutus, Compend. Theol. P, Schol. Horn. //. 1.399^106.

¹⁰³ The commentator gives three versions of the sequence of Alcman's alleged cosmogony: (I) III.78: (1) organised matter, (2) (II) III.814: (1) matter, (2) an organiser, (3) (4) (III) III.1517: (1) Thetis, (2). At III. 1720 he also identifies (1) matter with TOI -rravra, (2) the craftsman with Thetis, (3) the . Correlating these various versions leads to the result that is disorganised matter and Thetis is the organiser (despite the masculine gender at III. 11).