terrentia saepe figuris Cum subito emersere furenti corpora ponto: for thus Haupt punctuates; and the position of Cum is a good parallel to 22 9. freti for feri is an easy correction, as r, t, tr, rt, as we have again and again had reason to shew, are among the letters most frequently confounded in our Mss. 16 Illac (quaque alia?): this I think is a more elegant correction and gives a better rhythm than Schwabe's, or older corrections, tho' Ellis takes no notice of it: t and c are often interchanged in our Mss. and all Mss. alike are apt to omit one syll. of a word like quaque: 36 14 we find Colisque for Colis quaeque: 'on that day-and on what other in all time?-did mortal men cast eyes on the naked nymphs, as they rose breast-high out of the hoar deep'. I must say both Mueller's and Baehrens' violent corrections to my taste greatly spoil the picture. 23: The Virgilian scholia of the Verona palimpsest give us in a correct form the end of this line and half of the next, which has disappeared entirely in our Mss. Ellis alone among recent editors has rejected this gift with contumely: 'The weight of the Veronese Scholia' he says, 'imperfect and full of lacunae as they are, is not to be set against our Mss.; it is difficult to imagine any mode of filling up the lacuna which would not either be weak or load the sentence unnecessarily'. It is thus he can find in his heart to speak of what was once one of the most glorious codices that have come down from ancient times, written in the full blaze of the old classical world. Not to be set against our Mss. ! bad transcripts all of an archetype written when the gloom of mediaeval barbarism was at its deepest: and where too it preserves a line which they have lost, tho' Ellis does not hesitate to impeach these very Mss. of scandalous absurdity, in the way of omission, when he เ เ is dealing with our 54th poem. It is true these scholia are now in a very tattered state; but both Mai and after him Keil print: Catullus, Saluete deum gens o bona matrum Progenies saluete iter: without a hint that there is any doubt about any one of the magnificent letters of the original. Of the genuineness of this half verse I have no more doubt than of that of any other verse whatever in Catullus. Nay more, I do not see why all editors reject its 'deum gens' for the genus' of V; as I feel pretty sure that Virgil had Catullus in mind, when he wrote 'deum gens, Aenea'. matrum' too must be the poet's: nay the double reading 'mater' and 'matre' in G indicates that the final letters were obscured in V or in V's predecessor. Nor do I think it 'difficult' to fill up the verse as the poet may have written it; tho' none of the editions satisfies my mind: for the 'bona matrum' has no point unless the next line contained an epithet of matrum, which was as emphatic as bona, or more so. My reading then surely gives us what we want: 'right worthy progeny of right worthy mothers'. The joining of the mothers with the fathers is not without a purpose; for Catullus may well have thought as Euripides did in his Meleager: Stob. OB 12 ἡγησάμην οὖν εἰ παραζεύξειέ τις Χρηστῷ ovv ει πονηρὸν λέκτρον, οὐκ ἂν εὐτεκνεῖν, Εσθλοῖν δ ̓ ἀπ ̓ ἀμφοῖν ἐσθλὸν ἂν φύναι γόνον. I have never comprehended Ellis' defence of mater. 28 Nereine: this is nearer the Mss. and in other respects far preferable to the very suspicious Neptunine. All the patronymics quoted by Ellis are from Greek words: Neptunus is a pure Latin word. 기 αν 31 optatae finito. optato finito G, optato finite 0: another of the many many proofs of o and e being almost indistinguishable in our Mss.: this fact makes. Guarinus' correction in 309 roseae niueo' for the 'roseo niuee' of V highly probable. 48 Indo quod dente politum: 'which formed of the Indian tusk and finely wrought'. Comp. Virgil's ' pictas abiete puppes'. 82 quam talia Cretam Funera Cecropiae nec funera portarentur: comp. Ov. met. VIII 231 At pater infelix nec iam pater 'Icare' dixit: the nec seems really the same as non, of which I have spoken at 30 4: it may therefore perhaps be compared with the 'per non medium', the 'a non sensu' and the like which I have illustrated in my note on Lucr. 1 1075. ib. 105-11 Nam uelut in summo quatientem bracchia Tauro prona cadit lateque comeis obit obuia frangens), 109 comeis obit obuia scripsi. cum eius obuia V. I confess to setting some store on my emendation of 109, on which so many conjectures have been made. comeis might pass at once into cum eius, especially when the latter was written compendiously, as it is in O at all events; and obit might readily be absorbed in obuia: nay it may represent the double reading' omnia' in G: comp. my emendation obit for omnia (ouit for oža) in Lucil. XXVII 35 M. Whoever has seen a tree fall to the ground with its leaves on, must have marked the sweep and crash made by them as they first come into contact with the ground and spread themselves out. With 105 bracchia, and 109 comeis, comp. Aen. XII 209 posuitque comas et bracchia ferro; Georg. II 368 tum stringe comas, tum bracchia tonde. ib. 272, 273 Quae tarde primum clementi flamine pulsae leuiterq; sonant O, leviter sonant G. leni resonant uulgo. That O here too is right against G and other Mss. we have a strong confirmation in Sen. Agam. 680 licet Alcyones Ceyca suum Fluctu leuiter plangente sonent: see my note on 23 10 for this and other apparent reminiscences of Catullus' language in Seneca. ib. 276 Sic tum uestibuli linquentes regia tecta 'Vestibuli' cannot be right: the 'regia tecta the inner apartments of the palace, in which the 'puluinar' and the famous 'uestis' or tapestry, described at such a length, were exhibited; 'sedibus in mediis' (v. 48). The 'uestibulum' was the open court before the 'ianua', formed by the façade of the building, and the walls or rooms run out beyond this façade at each end of it. Ellis says: 'the shelter of the royal porch', inversion of the adjective. This to me explains nothing. They would pour out through the 'ianua' into the open 'uestibulum'; Vitr. VI 7 5 πpółupa Graece dicuntur quae sunt ante ianuas uestibula. Schrader's 'uestibulo' or 'uestibulis' may be right. But, seeing the persistency with which this wondrous'uestis' is dwelt on: 50 Haec uestis priscis hominum uariata figuris cet.; 265 Talibus amplifice uestis uariata figuris cet.: and remembering Virgil's 'pinguis ubi et placabilis ara Dianae', and 'pinguis ubi et placabilis ara Palici', I think Catullus may have written 'Sic tum, uestis ubi, linquentes' cet. The 'uestibulum' of 293, which by the way with its context really refutes 'uestibuli' of 276, may have caused the corruption. The front approach to the older part of Hampton Court gives a lively image of an ancient 'ues tibulum '. Catullus must have taken great pains to improve the rhythm and prosody of his two hexameter poems, as we may see if we compare him with any of his predecessors, such as Ennius or Cicero. In respect of elisions he is much less harsh than he is either in his hendecasyllables or in his elegiacs; and comes much nearer in these two poems to the rules which prevailed after his time. This is very remarkable and contrary to the usage of subsequent masters, Virgil for instance, if he be compared with Ovid and Martial. It is another proof too, in addition to those which I have given in my Lucretius, that 64 is one of his latest poems. his elegiacs, even in the last half of the pentameter, he has the very harshest rhythms and elisions, such as 'perdito amore fore'. In his hendecasyllabic poems, even in the sweetest of them, his elisions are quite as harsh, judged by the standard of Martial and Statius: even in his 45th he does not balk at such rhythms as 'Ni te perdite ámo átque amare porro', where a long vowel is elided before the accentuated short syllable of In |