² Magnus, H. (P. Ovidi Nasonis Metamorphoseon Libri XV [Berlin, 1914])Google Scholar prints aethere, but honours aequore with a fortasse recte and a cross-reference to 1.22 in his apparatus. While this was at the Press I learnt (what I should have known before) that aequore had been defended by Grilka, C. in Res Publica Litterarum 6 (1983), 143–4Google Scholar; many of my particular points and emphases complement his.

³ The OLD (s.v. seduco 4, ‘To draw apart, divide, split’) lists eight passages, though the earliest of them (Prop. 2.6.41) is conjectural, and diducet seems preferable there, as diductas does here. The other seven include Met. 13.611, Fast. 4.385, and Her. 18.142, of which the last is not certainly Ovidian.

⁴ West, M. L., Textual Criticism and Editorial Technique (Stuttgart, 1973), p. 59.CrossRefGoogle Scholar

⁵ Deferrari, R. J., SisterBarry, M. I. and McGuire, M. R. P., in A Concordance to Ovid (Washington, 1939)Google Scholar, list Met. 2.560, 6.405, and 8.588, Tr. 1.2.21, and Ib. 609, though Anderson prints deduxit in Met. 6.405. (For the variants in the last two passages, I have consulted André's Budé, J.Tristia (Paris, 1968)Google Scholar and Lenz's Paravia, F. W.Ibis (Torino, 1952)Google Scholar, since S. G. Owen's Oxford text (1915) is unhelpful on this point.) There is another probable instance in Am. 1.7.47, where Kenney reads diducere, though the paradosis is unanimous for deducere.

⁶ Cf. Bömer, F., P. Ovidius Naso, Metamorphosen. Kommentar (7 vols., Heidelberg, 1969–1986), ad loc.Google Scholar

⁷ Miller, F. J. (Loeb, 1916)Google Scholar translates: ‘These elements, although far separate in position, nevertheless are all derived each from the other, and each into other falls back again.’ This is not changed in G. P. Goold's revision (1977). Melville, A. D. (Ovid, Metamorphoses [Oxford, 1986])Google Scholar translates: ‘Though spaced apart, all issue from each other / And to each other fall.’

⁸ Examples of qualitative distare are Met. 8.438–9 discite…/ facta minis quantum distent!, Tr. 2.353 crede mihi, distant mores a carmine nostro, and Tr. 5.13.34 meis ut distent tua fata!

⁹ These are listed by Bömer ad loc.: there is another in Nux 75, which, though inauthentic, may still be pertinent to the habits of Ovid's scribes.

¹⁰ Since Anderson reports the reading as ‘sp (ati N²) osoque N’, it may well be that N² was emending suo Marte. However, collation with a lost source cannot be excluded.

¹¹ It may be significant that one of the editors and the anonymous referee split on this point. I wish to thank both of them, and David Kovacs, for their help with this paper, which was not confined to the points on which they are specifically mentioned.