3C Journal of Philology. Lucretius describes the innubilus aether, the fire-laden empyrean, far withdrawn from winds and rains; yet, like other poets, he again and again speaks of aetheriae nubes, and of the rains of father aether : yet Lachmann, in. 405, will not let him or others speak of aetheriae aurae, and II. 1115 he changes aetheraque [aether] into aeraque aer, though in the rhythm of this and the preceding verse Lucretius seems to be imitating the well- known verses of EmpedocleS, Tala pev yap yalav onairapcv v8an 8' v8a>p Aldfpi 8* aWepa 8lov arap nvpl iri>p atdrfkov ; V. 514 Lucretius says, aeterni sidera mundi, though it is one of the cardinal points of his philosophy, that our mundus, and every other mundus, are perishable, vi. 15 and 16, querellis is found at the end of both these verses ; in one of the two the true word has been displaced by a very common error of copyists. Lachmann, in the second, reads periclis; but saevire querellis is surely better than saevire periclis ; but he also supplants the genuine-looking Pausa atque by passimque ; read sine ulla for the first querellis, and all will be clear : animi is governed of ingratis, as in Plaut. Cas. n. 5. 7 : amborum ingratiis. There can be no question that Lachmann is right in saying that the archetype of our MSS. was written in thin capitals, like the Medicean Virgil ; a glance at the various readings will prove this ; letters of similar sound are also often interchanged. But there is likewise a perpetual confusion of letters which resemble each other when written in a small rather than a capital character. My experience in such matters is slight, but it has always ap- peared to me that none of our MSS. comes immediately from the archetype : that at least one written in the " littera minuscula'* has intervened. Lachmann has frequently done good service by observing the law of change ; but seeing that e and i are perhaps more frequently confused than any other letters, I don't know why he believes that he better represents the MS. readings transere, sentere, unguente, se, tale, by writing transeire, &c, than by keeping the common forms ; nor why tripodi should be read in the first book, but tripode in the fourth ; nor why Lachmann and Bernays, vi. 66, retain rationi, though the poet everywhere else uses ratione. Indeed, I cannot understand on what principle Bernays rejects or retains these unusual forms. Why should he refuse unguentei, set, &c. but keep talei ? Why reject ni for ne, but retain nive? Many corrupt passages may still, I think, be
