26 Journal of Philology. publish some of these ; at present I will content myself with noting that at i. 141 he too suggests sufferre, but he may have seen the edition of Faber published in 1662 ; at v. 747 he sug- gests didit for the redit of MSS. ; at iu. 1016 jactu' reorum; at vr. 1195 he anticipates Lachmann in proposing tumebat. The extent and accuracy of Lachmann's knowledge is so great, that it may well be said of him, as of Bentley, that you often learn more from him when wrong, than from others when right ; but certainly he often lays down arbitrary rules to which he forces Lucretius to conform by a Procrustean process. He observes that it is illiberal to expect too great uniformity in spelling ; it appears to me no less illiberal to refuse to Lucretius such slight idiomatic deviations from strict grammar, as all Greek and Roman writers allowed themselves. Lachmann asserts that it violates the "antiqui sermonis castitas" to use et for etiam, and then remorselessly alters some ten passages ; not one of which has the least appearance of being corrupt. Why Ber- nays in some of these instances retains et, in others rejects it, I cannot tell. Not only did Lucretius use et for etiam in these passages, but I believe that in two others it is to be restored. No confusion in our MSS. is more common than that between ut and et ; therefore in iv. 638 I would read : Est itaque et serpens, for the ut of MSS., " thus there is a serpent also f a and e and 03 too are continually interchanged ; in vi. 604, therefore, I would read subdit et hunc stimulum, for Lachmann's read- ing does not seem to me satisfactory. Nor can I accept uncon- ditionally his dicta about the omission of the substantive verb ; certainly I should not read est for e, n. 194, nor for in, n. 431 ; iv. 271 and 278, he corrupts the sense by changing vere, which is certainly the true reading, into sunt bene ; here Bernays does not follow him ; fl. 137, I believe qum porro is right, not pro- porro; v. 720, I should not read sit for *t forte; nor am I con- vinced of the necessity of inserting est in I. Ill and vi. 746. Again, I am not sure that Lucretius, because he uses fulgZre, fulgit, 4rc, would never have used fulget, fulgent, refulget, fulgere. Virgil has fulgtre, fervSre, yet he does not fivoid fulgent, fervet, &c. I am equally indisposed to believe that Lucretius denied himself the liberty of lengthening a short syllable by caesura, as in n. 27, fidget or fulgit auroque, and in m. 21, where I retain semper in- nubilus with the MSS. Again in v. 1049 I cannot believe that
