¹ Gellius is not concerned with the story's truth or falsity, but declares that the juxtaposed ō-sounds make ora ‘melius suauiusque ad aures’. This reflects the doctrine found in Demetr. eloc. 72–4, the Towneley scholia on X 152, and Eustathius 1702. 19 on X 596 (cf. D. H. comp. 20); one may argue that in Vergil the sequence suggests the richness of Campanian soil. Gellius cites the two Homeric passages, but then attempts to import this feature of the (Demetr. loc. cit.) into Catullan lyric, arguing for ebria acina at 27.4; to Fordyce's refutation I add that the repeated elision of a long vowel in ebrioso acino ebriosioris admirably hits off the slur of a drunkard's speech.

² DS on G. 2.224; Hosius's reference in his Teubner Gellius to Philargyrius is illusory.

³ Cf. the literature quoted by Hosius, vol. i, p. xxxvi.

⁴ See in general Goold, G. P., HSCP 74 (1970), 101–68.Google Scholar

⁵ Goold, pp. 125-6; the grossest example is the alleged rewriting of Georgic 4 when Gallus fell, cf. Anderson, W. B., CQ 27 (1933), 36–45, 73, Goold, p.CrossRefGoogle Scholar 137. So too in Euripides: Zuntz, G., An Inquiry into the Transmission of the Plays of Euripides, pp. 253–4,Google Scholar cited by Goold. The story at Vita Persi 57 that sat. 1. 121 originally read ‘auriculas asiniMida rex habet’ is plausible till one notices that quis non resumes verse 8.

⁶ Pokorny, J., Indogermaniscbes etymo-logiscbes Worterbuch i. 1, cf. Fordyce ad loc. and on verse 684 for such etymological epithets in Vergil. Hemica saxa (7. 684) shows knowledge of a word in Sabine or Marsic, closely related to Oscan; a poet learned enough to know (even if at second hand) the Punic etymology of Kartbago (Aen. 1. 298), who loved Italy as his poetry shows he did, will surely have known at least a word or two of the once great language not yet dead that had made the second of his predecessor Ennius' three hearts (Gell. 17.17.1). Lucilius, writing perhaps no longer a time before Vergil's birth than the Aeneid was written after it, used an Oscan word, correctly declined, in place of a Latin one with identical scansion (1318 M ap. Fest. 384. 30–2 L).Google Scholar

⁷ Nissen, H., Italische Landeskunde II. ii. 755;Google Scholar Abella ‘nostrae hinc sex milibus absita Nolae/altiiugos montes inter iacet’ (Paulin, . carm. 21. 712–13); by contrast ‘campo Nola sedet’ (Sil. 12. 162) at only 40 m (Nissen, p. 756).Google Scholar

⁸ Ausonius, epigr. 79. 5 Peiper; an old Oscan custom (Festus 204. 31–2 L).

⁹ Housman, , Classical Papers, pp. 733 and 1180.Google Scholar

¹⁰ For the date see Fabre, P., Essai surla chronologie de I'œuvre de saint Paulin de Nole, p. 37. The parallel was noticed by Fordyce on Aen. 7. 740.Google Scholar

¹¹ It is not particularly probable that Servius' commentary had appeared by 407 (cf. A. D. E. Cameron, JRS 56 (1966), 30), still less that Paulinus had a copy; but it is more than likely that Servius took over negatum bospitium (cf. negauerant in DS on G. 2. 224) from Donatus, who need not be its coiner. Paulinus speaks of ‘hospitium communis aquae’ because both by law (J. 2.1.1) and by common consent (Cic. off. 1.52) water flowed for all, and was not to be denied save to the outlaw; see in general Schulze, W., Kl. Schr., pp. 189–210.Google Scholar

¹² See Fraenkel, , Horace, p. 275 and n. 1.Google Scholar

¹³ Whether like Lucretius (6. 552, 1072; cf. Italian aequo) Paulinus permitted aqu- to fill a biceps or wrote the tuae aquae that Hartel gives him, with its hideous elision and its inappropriate suggestion of dominium, there are others more qualified to judge.

¹⁴ Nissen, pp. 751, 758.

¹⁵ Cf. Cic. Att. 13. 8 for outsiders' seeking properties in the Nolanum.

¹⁶ Vit. Donat. 45 Hardie.