THE THE QUASI-CAESURA IN VERGIL. HE importance to the rhythm of the hexameter of a caesura in the third foot is generally recognized Some confine the definition of the penthemimeral caesura to the 'strong' species, where the break is after the first long syllable of the third foot. In this paper I include also the weak' or 'trochaic,' the roun Kатà трíтоν троxatov, where the break is after the first short syllable of a dactyl in the third foot. When we consider the prevalence1 of this trochaic caesura in Greek hexameters, and reflect how sensitive the Roman ear was to its rhythmic effect-so sensitive that Cicero wrote mihi crede, lest crede mihi should give a suggestion of verse-and when we feel where our metrical sense is first awakened in such a line as the following: Apparet domus intus, et atria longa patescunt Aen. ii. 483; and when we find that in Verg. Aen. ii., out of 804 lines, 673 have the strong, 85 the weak caesura, and 46 neither, and in Aen. xii., of 950 lines, 792 have the strong, 129 the weak, and 29 neither (i.e. that in these books there is a 1 In Hom. I. xxii., in the first 100 lines, there are 54 with trochaic, or weak caesura, 35 with strong caesura, 10 with both and one with neither. In the second 100 lines there are 50 with trochaic, 42 with strong, 7 with both, and one (in which a proper name fills the third foot) with neither. 2 Cf. Aen. iii. 644, 707; iv. 316, 486; v. 781, 785; vii. 711; x. 24, 95, where the trochaic caesura of the third is supported only by the trihemimeral caesura, or the quasi-caesura. In iv. 486, Spargens humida mella soporiferumque papaver, it has no support except another trochaic caesura that is latent in the compound word. weak, and not a strong, caesura in about one-eighth of the total number of lines), I think we must admit the metrical power of this trochaic caesura, and allow it to be regarded as a penthemimeral caesura, though the term may be not strictly applicable. We are truly told that the penthemimeral caesura is of itself sufficient to make a verse harmonious, e.g. Pars stupet innuptae donum exitiale Minervae. Aen. ii. 51. Res Agamemnonias victriciaque arma secutus. Aen. iii. 54. But when it is added that, when the penthemimeral caesura is absent there must be found both the trihemimeral and the hepthemimeral-and by penthemimeral caesura is here meant only the strong form-we must ask leave to dissent. Praecipitat, suadentque cadentia sidera somnos. Aen. ii. 9, is surely a musical line; and if we propose to ourselves the question where we begin to feel that we are reading verse, we must, I think, answer that it is when we perceive that the ictus falls on the unaccented final syllable of praecipitat, and that we become sure of the rhythm when we come to the trochaic caesura of the third foot. Is there then any law which in Vergil governs the omission of the penthemimeral caesura in this wide sense? On examining the lines which have no caesura in the third foot, we shall find that there is an elision before the third foot in about 70 per cent. of them, and that the rest admit of explanation on special grounds. The analogy of this phenomenon to the quasi-caesura' in Greek iambics will strike everyone. Before developing the analogy I must seek to establish the law. The following are all the lines in Aen. v. which have no break in the third foot. Those marked with an asterisk are also without elision before the third foot, and are seeming exceptions to the proposed rule: 1. Interea medium Aeneas iam classe tenebat. 18. Spondeat, hoc sperem Italiam contingere caelo. 133. Ductores longe effulgent ostroque decori. 170. Radit iter laevum interior, subitoque priorem. 250. Victori chlamydem auratam, quam plurima circum. *260. Loricam, quam Demoleo detraxerat ipse. 295. Euryalus forma insignis viridique iuventa. 316. Corripiunt spatia auxilio, limenque relinquunt. 327. Iamque fere spatio extremo fessique sub ipsam. *343. Tutatur favor Euryalum, lacrimaeque decorac. 399. Haud equidem pretio inductus pulchroque iuvenco. 405. Terga boum plumbo insuto ferroque rigebant. *407. Magnanimusque Anchisiades et pondus et ipsa. *418. Idque pio sedet Aeneae, probat auctor Acestes. 434. Multa cavo lateri ingeminant et pectore vastos. 439. Ille, velut celsam oppugnat qui molibus urbem. 443. Ostendit dextram insurgens Entellus, et alte. 468. Ast illum fidi aequales genua aegra trahentem. 472. Accipiunt; palmam Entello taurumque relinquunt. 514. Tela tenens, fratrem Eurytion in vota vocavit. 522. Hic oculis subitum obicitur, magnoque futurum. 625. Infelix! cui te exitio Fortuna reservat. 626. Septima post Troiae excidium iam vertitur aestas. 641. Haec memorans, prima infensum vi corripit ignem. 654. At matres primo ancipites, oculisque malignis. *662. Coniciunt. Furit immissis Vulcanus habenis. *675. Accelerat simul Aeneas, simul agmina Teucrum. *678. Saxa petunt; piget incepti lucisque, suosque. 714. Pertaesum magni incepti rerumque tuarum est. 716. Et quidquid tecum invalidum metuensque pericli est. 725. Care magis, nate, Iliacis exercite fatis. 760. Fundatur Veneri Idaliae, tumuloque sacerdos. In these 36 instances of the omitted caesura in the third foot, there are only eight (marked by an asterisk) which have not an elision before it. Of these eight, five (260, 343, 407, 418, 675) admit of the all-covering plea of a proper name, which gains indulgence for even an anapaest in the fourth foot of a senarius. The other three I would defend on the principle of the detachable character of a preposition, or similar word, in composition. This latter, so common in Greek, is, in Latin, familiar to us from, Hostile aratrum exercitus insolens HORACE, C. I. xvi. 21; Dum flagrantia de torquet ad oscula ID. ib. II. xii. 25; and, as we can see in line 127 above, applies to the negative prefix in- (cp. Hor., A. P. 263). It will be observed that the elision generally occurs before a similarly compounded word (e.g. 133, 170, 295, &c.), and it may be asked why we should not adopt this generally as a sufficient compensation for the lack of the caesura. I have no more doubt that the occurrence of such a compound after the elision helps the metre than I have that the trihemimeral and hepthemimeral caesuras (and, of course, the cadence in the fifth and sixth), which also nearly always are found along with the quasi-caesura, help the metre. But the principle of separable compound' will not account for such cases as 250, 316, 468; and the number of instances where elision before the third foot is found to compensate for absence of a caesura in that foot forces us to believe that Vergil felt the metrical effect of that elision. If this were not so, what was to prevent his writing in 790, Quam molem excierit subito; maria omnia caelo ? I proceed now to examine Aen. vii., which was perhaps the last book Vergil composed. In it we find a far smaller number of lines without a caesura in the third foot, but the proportion of these remedied by quasi-caesura about the same: 177. Quin etiam veterum effigies ex ordine avorum. 748. Armati terram exercent, semperque recentes. 775. Sedibus, et nymphae Egeriae nemorique relegat. Of the four seeming exceptions to the law in these seventeen instances of omitted caesura, 257 and 333 are explained by the 'detachable prefix,' 294 by the principle of the proper name, and 704 perhaps by an extension of the latter principle; aeriam being a word of Greek origin might be felt to have some claim to indulgence, though its use by Lucretius suggests a different explanation. It is worth remarking, that the proportion of lines with only the weak caesura in the third foot has not sensibly diminished in this book, there being 85 to 715 with the strong. There are 49 cases where elision occurs before a strong caesura in the third foot, as 16. Vincla recusantum et sera sub nocte rudentum, which tends to show that the Ancients felt its metrical |