This conjecture of Mr Gow's for teque (p. 155 of the April number 〚CR 4 (1890), 155〛) has a special interest for me, as I had pencilled it in my own margin some while ago. I do not even now feel sure that it is right, for although Ovid often appends que in this way to a quoted word, there is no similar instance in Horace; and moreover Meineke's atque may well be the true reading. But my purpose in writing is to present Mr Gow with a parallel passage which seems to tell strongly in his favour: Ovid Trist iv. 2, 51 sq. ‘tempora Phoebea lauro cingentur “io” que ∣ miles, “io” magna uoce “triumphe” canet.’ This has all the air of a copy from the above text, for Ovid is here in a very imitative mood: the pentameter is taken word for word from Tib. 11. 5, 118, and his next distich, ‘ipse sono plausuque simul fremiruque calentes ∣ quadriiugos cernes saepe resistere equos’, comes from Prop. 111. 4, 14, ‘ad uulgi plausus saepe resistere equos’.

Mr Gow's certain emendation cerebrique in Serm. 11. 3, 208, was forestalled by Horkel, a critic who in a short life did more for the text of Horace than any man since Bentley, though the editors, with the significant exception of Meineke, seldom deign even to record his corrections.

Catullus LXIV. 279–87

Before dealing with the last verse I will offer a short defence of the conjecture ‘aperit’ for ‘perit’, v. 282, which has already seen the light in Mr Postgate's edition. The vulgate text is ‘parit’, which a later hand has written in G over the erasure of the original reading; O has the abbreviation which regularly stands for ‘perit’ but may according to the practice of that scribe signify ‘parit’ also. If the tradition is ‘perit’, then ‘aperit’ since an a precedes is an easier change than ‘parit’; while even if ‘parit’ were clearly given by the MSS I should think it hardly suitable: it is surely ‘terra’ or ‘flumen’ that ‘parit flores’, just as ‘campi ferunt’ and ‘ora creat’ in the lines above: the function of ‘aura’ is more properly expressed by ‘aperit’.

The first edition of Baehrens' Catullus, which now that the second has appeared will fetch fancy prices, was in the rigour of the term an epoch-making work. But it exhibited a text of the author much corrupted by unprovoked or unlikely or incredible conjecture; so that the task of revision was delicate, and the choice of a reviser was not easy. It was not easy; but scholars who are acquainted with the history of Catullus' text and with the metres he wrote in, who know how to edit a book and how to collate a manuscript, who are capable of coherent reasoning or at all events of consecutive thought, exist; and to such a scholar the task might have been allotted.

It has been allotted to Mr Schulze, who says, ‘Munus nouae huius libelli editionis post praematuram Aemilii Baehrensii mortem curandae ita suscepi, ut quoad fieri posset quam plurima eorum, quae ille ad Catulli carmina et recensenda et emendanda contulisset, retinerem ac seruarem.’ Out of Baehrens' conjectures Mr Schulze has found it possible to retain six. The first of these is the merely orthographical correction 2 6 lubet for libet or iubet.

In the American Journal of Philology, vol. XIII p. 152 〚this edition p. 192〛, I proposed to write here ἦ πατρόθεν, ὡς ἀκούω, δυσώνυμα λέκτρ' ἐπάσω; and I added i.e. didst thou wed thy father's widow? No. XIX. of Hermathena contains an article by Prof. Tyrrell, full of his usual kindness and generosity, in which the following remarks occur (p. 307) :

I take πατρόθεν with the verb: ἦ πατρόθεν λέκτρ' ἐπάσω means came thy bride unto theefrom thy father's arms? My paraphrase didst thou wed thy father's widow? is in sense the exact equivalent of this; but since its form has misled Prof. Tyrrell, and may therefore mislead others, into thinking that I took πατρόθεν λέκτρα to mean thy father's widow, I desire to make this explanation.