¹ The passage with which this note is chiefly concerned has been recently discussed by Huxley, G. L. (‘Arsinoe Lokris’, GRBS 21 [1980], 238–44)Google Scholar; though I agree with him in thinking that the relevance of Venus has been underestimated, I take a somewhat different view of its place in Catullus' picture.

² The text is preserved by two papyri, PSI 1092 and P. Oxy. 2258; half-brackets indicate letters supplied from the latter's scholia. In 54 PSI 1092 gives Λοκρικόϲ; for further details see Pfeiffer's apparatus.

³ We may disregard the theory of Ellis, Robinson (Commentary on Catullus ² [Oxford, 1889], 386)Google Scholar, who supposed that the Lock was removed from the Royal Palace at Alexandria. However, though Ellis was not in general a luminous interpreter of Catullus, his discussion of this poem squarely faces difficulties to which others have tended to turn a blind eye and is still worth reading.

⁴ This phrase comes from a slightly corrupt quotation in the scholia to Arat. 146. However, Catullus' version of 8 differs significantly: multis…dearum is more specific than π⋯ϲιν…θεοῖϲ and pollicita est represents an earlier stage in the proceedings than ἔθηκε. I suspect that the commentator quoted from memory and perhaps confused 8 ff. with 33 ff. It sounds as if Berenice made her dedication to Aphrodite-Arsinoe and other named goddesses (perhaps no more than three) and to all the gods collectively.

⁵ See further Fraser, P. M., Ptolemaic Alexandria (Oxford, 1972), i. 239, 729–30, ii. 1023 n. 105Google Scholar.

⁶ Chron. Pasch. 266c (Migne, , PG 92, 652–3Google Scholar): τούτοιϲ τοῖϲ χρόνοιϲ ⋯ν Ἀλεξανδρείᾳ τ⋯ Сεβηριαν⋯ν ⋯κτίϲθη γυμνάϲιον κα⋯ τ⋯ ⋯κεῖϲε ίερ⋯ν μέγα τ⋯ καλούμενον Πάνθεον.

⁷ Astron. 2.24: ‘cuius supra simulacrum proxime Uirginem sunt aliae septem stellae adcaudam Leonis in triangulo collocatae, quas crines Berenices esse Conon Samius mathematicus et Callimachus dicit: cum Ptolemaeus Berenicen Ptolemaei et Arsinoes filiam, sororem suam, duxisset uxorem, et paucis post diebus Asiam oppugnatum profectus esset, uouisse Berenicen, si uictor Ptolemaeus redisset, se crinem detonsuram; quo uoto damnatam crinem in Ueneris Arsinoes Zephyritidos posuisse templo, eumque postero die non comparuisse; quod factum cum rex aegre ferret, Conon mathematicus, ut ante diximus, cupiens inire gratiam regis dixit crinem inter sidera uideri collocatum; et quasdam uacuas a figura septem stellas ostendit, quas esse fingeret crinem’. Nearly all this account might simply derive from Callimachus, but not the king's anger nor (unless Catullus has suppressed the detail) the specific number of stars (which seems to reflect the unrivalled attraction of seven as a symbolic or rhetorical number rather than actual observation; see further Roscher, W. H., Die Sieben- u. Neunzahl im Kultus u. Mythus der Griechen, ASAW 53 [1906], 1)Google Scholar.

⁸ Compare the Homeric description of a newly risen star as λελουμένοϲ Ὠκεανοῖο (Il. 5.6) and the phrase twice used of the Great Bear (Il. 18.489, Od. 5.275) οἴη δ' ἄμμορόϲ ⋯ϲτι λοετρ⋯ν Ὠκεανοῖο.

⁹ Less concerned than Callimachus about the practical details of the process, he created a handling problem for those who are happy to envisage Zephyr as a horse (on which see below n. 13) by suppressing the first half of 55 (where ἤ[λ]αϲε, suggested independently by Skiadas, A. D. (Athena 68 [1965], 44)Google Scholar and Ardizzoni, A. (GIF 27 [1975], 198–9)Google Scholar, is a very attractive supplement).

¹⁰ The Aristarchean distinction between αἰθήρ and ⋯ήρ as upper and lower air does not hold for Callimachus: cf. h. 4.176. But aether and aetherius usually refer to the upper regions of space (or to the air of earth as contrasted with Hades, which is irrelevant here).

¹¹ πάντα τ⋯ν ⋯ν γραμμαῖϲιν ἰδὼν ὅρον ᾗ τε ɸέρονται ~ omnia qui magni dispexit lumina mundi.

¹² Cf. Verg. A. 6.849–50.

¹³ ‘Minime gratus redit, quem in perpetuum e textu expulsum esse sperabamus, equus mirus’ comments Pfeiffer. After the objections which he and Housman had assembled against a horse as Aphrodite's emissary (Philol. 87 [1932], 197 ffGoogle Scholar. = Kallimachos ed. Skiadas, A. D. [Darmstadt, 1975], 116 ff.Google Scholar), CR 43 (1929), 168Google Scholar (= Classical Papers of A. E. Housman [Cambridge, 1972], iii. 1157Google Scholar), it is unnerving to find in P. Oxy. 2258 apparent confirmation for equos of Catullus' MSS. However, anyone who would still prefer a horseman could restore the Homeric ἱππό[τ'] instead of ἵππο[ϲ]; the scribe does not normally write elided vowels, and this elision does not appear to violate Callimachean metrical practice: for elision of final a with nouns, adjectives or verbs after the first trochee cf. h. 2.82; 4.202; 5.55; fr. 66.7. Of course, the cogency of this supplement depends on one's view of the corresponding line in Catullus.

¹⁴ It is difficult altogether to dismiss the suspicion that the Lock's disappearance was engineered by the man who knew where to look for it; but he would have needed to supply his accomplices with a cover-story.

¹⁵ Musae. 249–50.

¹⁶ E.g. Antip. Thess. 93 (597 ff.) Gow–Page (= A.P. 9.143), Gaet. 1 Page (= A.P. 5.17), Phld. 15 (3246 ff.) Gow–Page (= A.P. 5.306); for further material see Nisbet and Hubbard on Hor. O. 1.5.16.

¹⁷ On Aphrodite Euploia see Roscher, i. 401–2, Nilsson, , GGR i. 521Google Scholar.

¹⁸ Call. Epigr. 5 (1109 ff. Gow–Page), Posidipp. Epigr. 12, 13(3110 ff.) Gow–Page; see further Fraser, op. cit. i. 568–9, 667–8, ii. 388–9 nn. 389–90; 811–2 nn. 139–40; 935–5 nn. 400–4.

¹⁹ Ardizzoni's restoration [θ⋯κεν ἄɸαρ καθαρού] (loc. cit. n. 9) deserves serious consideration.

²⁰ First attested in the fourth century: see LSJ s.v. Ἀɸροδίτη III, RE xx 2112 ff. s.v. Planeten, viii A 887 ff. s.v. Venus (2). On the wider Middle Eastern background to these divine names see Eilers, W., Sinn u. Herkunft der Planetennamen, SBAW (1975), 5Google Scholar.

²¹ Met. 1073b30–2: εἶναι δ⋯ τ⋯ϲ τρίτηϲ ϲɸαίραϲ τοὺϲ πόλουϲ τ⋯ν μ⋯ν ἄλλων ἰδίουϲ, τοὺϲ δ⋯ τ⋯ϲ Ἀɸροδίτηϲ κα⋯ το⋯ Ἑρμο⋯ τοὺϲ αὐτούϲ.

²² Cf. Call. fr. 291.

²³ οἱ δ⋯ ποιητα⋯ Ὠκεαν⋯ν αὐτ⋯ν (τ⋯ν ⋯ρίζοντα) καλο⋯ϲιν (Isag. ad Arat. Maass, p. 95, 10). Failure to recognise this convention at Il. 23.227, where Homer describes the sun rising out of the sea, has led to the inference that the composer of the Iliad must have lived on an eastward-facing coast.

²⁴ ‘It is now evident’ wrote Lobel in his introduction to the relevant fragments of P. Oxy. 2258 ‘that it is impossible to depend on the Latin, which too often, as at 11. 45, 67 seq., 72, 77, 80-fin., recedes far from the Greek’; Catullus' interpreters seem strangely reluctant to admit this. But in addition to Lobel's catalogue of divergences (which relates only to the new evidence afforded by the Oxyrhynchus fragment) we must note Catullus' less technical version of v. 1, a small but significant discrepancy over the Chalybes' contribution to iron-working (48–9), and the lack of sufficient detail to make it clear that Aphrodite's emissary is a wind (53). The proportion of text over which we can make a fair comparison is small, and whatever may be the correct explanation for the considerable discrepancies between the Oxyrhynchus fragment and Catullus at 80 ff., the differences in the earlier part of the poem can hardly be explained by the hypothesis that Catullus used a text varying significantly from what our papyri offer. (It is commonly held that Callimachus added the nuptial rite of 79–88 when he incorporated this poem in the Aetia I cannot see that this explanation does much credit to his judgement: ‘A poem of which the αἴτιον is the forming of the constellation Βερενίκηϲ πλόκαμοϲ is not improved by the superposition of an αἴτιον concerning a marriage custom’ (Lobel). As regards the end of the poem, either Catullus' MSS. are corrupt, or his model was, or both.)

²⁵ The heliacal rising of a star or constellation is the date on which it rises immediately before sunrise; some imprecision is inevitable when we are dealing with constellations, because they cover a wider area than individual stars. Dates for the rising and setting of Coma in the mid-third century at the latitude of Alexandria were supplied to Lobel for the ed. pr. of P. Oxy. 2258 by H.M. Nautical Almanac Office: see Pfeiffer, , Callimachus, i. 120Google Scholar. The movements of Venus in 246 b.c. (–245 for astronomical purposes) are calculated from the tables given by Tuckerman, B., Planetary, Lunar and Solar Positions, 601 b.c. to a.d. 1 (Philadelphia, 1962)Google Scholar.

²⁶ Three are of fourth magnitude, the rest of fifth or sixth. Stars of sixth magnitude are the faintest visible to the naked eye.

²⁷ His interests being purely astrological, he was content to discover what time the star appeared (Ev. Matt. 2.7).

²⁸ I wonder whether the presence of both the sun and Venus in Virgo at what I believe to be the relevant date is to be connected with the emphasis on Berenice's lost virginity (11 ff., 77 ff., cf. 26), natural enough in a hymeneal context but (as Ellis remarks) a little strange several months after her wedding. The apparent paradox of Aphrodite's star in Virgo might well inspire reflection on the antithesis of maid and wife.

²⁹ Ep. 27; cf. Arat. 145–6.

³⁰ On the rival candidature of the Lesbian Maidens see Wilamowitz, , Hellenistische Dichtung (Berlin, 1924), i. 217 n. 1Google Scholar. Imitators have not been so successful: who now knows where to look for Lyra Georgii and Scutum Sobieski?

³¹ Though Conon is often (rather misleadingly) described by modern writers as ‘the court astronomer’ or ‘the astronomer royal’, his mathematical interests are better attested than his astronomical work; it is hard to feel confident that Seneca had evidence independent of Callimachus for his statement (nat. quaest. 7.3.3) that Conon compiled a list of solar eclipses from Egyptian records. See further Fraser, op. cit. i. 400–1, ii. 580–1 nn. 188–96.

* I could not have written this article without astronomical assistance from my husband, Martin West, who not only did for me what I envisage Conon doing for Callimachus, but also calculated the movements of Venus during 246 and 245 and saved me from various errors and obscurities. I am also indebted to Adrian Hollis for his comments.