^{page 327 note 1} Cf., for example, Ussani's attempt, in his commentary ad loc. ; other vain efforts are listed in Orelli-Hirschfelder's Excursus on the passage. Such endeavours have failed because they are based upon impossible acrobatics in punctuation.

^{page 327 note 2} There is of course no tautology in Odes I. 27. 21 quis Thessalis magus venenis, or in Epod. 5. 61 dira venena.Google Scholar

^{page 327 note 3} He explains : ‘quamvis venena multum (this renders the magnum: cf. Keller-Holder's apparatus) possint, non tamen valent merita in contrarium vertere, ut liberentur poena, qui male mereantur.‘

^{page 327 note 4} On this methodological approach cf. Eranos, 1965, p. 32, and my forthcoming papers Emendation einer horazischen Korruptel (Rhein. Mus. 1967), Lateinische Beiträge (Mnemos. 1968).Google Scholar

^{page 327 note 5} Porphyrio, in fact, saw this meaning in the part of the text that is not corrupt (non valent convertere humanam vicem = non tamen valent merita in contrarium vertere, ut liberentur poena, qui male mereantur) ; he could, of course, do nothing better than interpret the corrupt magnum fas nefasque as an apposition to venena, which indeed is the meaning aimed at by the copyist responsible for the corruption. Of course, apart from its clumsiness, the apposition would be logically impossible : the venena the boy is declaiming against can hardly be termed fas (cf. Hor, , Odes. 2. 13.Google Scholar 8. venena … nefas) and in any case, even assuming that the boy is talking about venena in the archaic sense of ‘juices’ (cf. Lewis-Short, s.v. i) we would then expect a disjunctive (something like magnum fas aut nefas), not the copulative -que. Cf. Mac-leane's rendering of Ritter's interpretation, which significantly reflects these difficulties : poisonous juices, be they good or bad‘ (italics mine : if the juices are poisonous, how can they be good ? and or is not the exact transla tion of the copulative -que).

^{page 327 note 7} ‘Delle tante interpretazioni, con cui il passo è stato tormentato, pare a me la più accettabile quella accolta da Kiessling-Heinze : “gli spedienti magici possono, sí, rovesciare i termini tra it fas e it nefas … ma non possono sovvertire la legge dell’u mana retribuzione”.’ Tescari's note—like the whole of his commentary on Horace—is very good, and usefully supplements Wick-ham's discussion.

^{page 328 note 1} Paraphrasing Manilius 2. 602, we may say that, according to the boy's warning in Horace's passage, Canidia's noxia cannot vanquish the divine poenas.

^{page 328 note 2} Miscent, in other words, was written by Horace in the form and was misread by the copyist as (i.e. = magnum; on the ligature = ma cf. Thompson, , Introd., Facsim. 105, nos. 25-27Google Scholar; Horace's s was mistaken for a g of the ‘S’ type, and his c (of the ‘broken’ type, cf. van Hoesen, Table A, 2178; Table D, seventh c) for an n of the type with a ‘hooked’ left vertical stroke (e.g. Battelli, Lez. Pal., p. 71) and a ‘reduced’ right vertical stroke (confusion between c and n might also be assumed on the basis of the types Hoesen, van, nos. 15–16Google Scholar [facing p. 56], resp. third c and fifth n, if we assume that Horace wrote on papyrus). On t with a short horizontal stroke cf. van Hoesen, Table C, first t.

^{page 328 note 3} Horace likes to use non placed asyndetic ally at the beginning of a sentence : cf., for example, Odes 2. 17, Sat. 1. 9. 56 ff., 2. 3. 321 ff., Ep. 1. 14. 15 ff., I. 17. 34 ff., 1. 20. 5 f., 2. I. 120 ff., 2. 2. 12 ff., 2. 2. 200ff.Google Scholar

^{page 330 note 1} On the construction, cf., for example, Cic. ad Fam. 9. 13. 18, Dolabella merum bellum loquitur.

^{page 330 note 2} In connexion with our discussion of loquor and dico acc., it may be observed here that at Odes 4. 15.Google Scholar I (Phoebus volentem proelia me loqui / victas et urbes increpuit lyra) loqui is a touch of felicitous Selbstironie: the poet professes his earlier intention of ‘talking abundantly’ about epic matter, loqui being jocularly self-derogative (cf., on the other hand, the serious dicam et Alciden puerosque Ledae, Odes I. 12. 25).

^{page 330 note 3} The need for poetic celebration of the Roman victories over the Carthaginians was certainly felt, witness Silius Italicus. Naevius‘ poem was, of course, hardly regarded by Horace as genuine poetry ; Ennius had, in his Annales, celebrated the Punic wars (Hor, . Odes 4. 8. 20), but Augustus wanted epic works in contemporary Latin, written by Horace or his colleagues.Google Scholar

^{page 331 note 1} Hannibal's, speech is an of the type cf. Wickham on 4. 4. 49.Google Scholar

^{page 331 note 2} Silius Italicus, with his characteristic love for antiquarian detail, preserves the name of Hannibal's wife, Imilce, just as he has preserved the name of Atilius Regulus’ wife, cf. Kiessling-Heinze, on Odes 3. 5. 41 : cf. Pun. 3. 97, 106 and 4. 775, 806. Amongst the biographical data on Hannibal known to Horace's public (e.g. Livy 29. 29. t2) there was the warrior's married status (e.g. Livy 24. 41. 7: cf. Pasquali, op. cit., p. 768, with note 2). Both Horace's and Silius’ are based on this information.Google Scholar