Page images

ing with it-võlvěrě | Pārcās ;' (2) caesura between the short syllables of the dactyl-saeviquě | dòlòrēs. All other closing rhythms are with him exceptional, sometimes in imitation of Greek rhythm (e.g. hýměneaos, cýpărissīs, Lāðdămiā), sometimes for special effect (e.g. 'quādrůpědāntūm’ Aen. xi. 614, “pudeāt sõlă neve' G. i. 80, 'procūmbịt hủmi bos;' or spondaic endings, as "ābscondāntūr' G. i. 226, 'purpureo nārcīssõ'Ecl. v. 38).

3. A purely dactylic line, common enough in Greek (e. g. Iliad i. 13, 25, 31, 32, 34, 54, and so on in like proportion), is comparatively rare in Virgil's epic poetry; such a line as 'Quadrupedante putrem sonitu quatit ungula campum' (Aen. viii. 596, cp. G. iii. 201) being a conscious imitation of the sound of galloping, and its jerky movement being foreign to the stately march of epic rhythm. The greater length of many Greek words, with a greater abundance of long compounds, prevented the Homeric hexameter from moving too rapidly or jerkily—e. g. Iliad i. 87 Eůxóuevos Aaναοίσι θεοπροπίας αναφαίνεις : but in Virgil the rapid movement is almost always checked and the rhythm, as it were, collected and steadied by a spondaic fourth foot-e. g. Aen. i. 45 ‘Turbine corripuit scopuloque infixit acuto.' The spondaic fourth foot is indeed specially characteristic of Virgil's epic rhythm, as the dactylic fourth foot ending a word (“Bucolic caesura') is of his Eclogues in imitation of Theocritus : and wherever, for the sake of variety, the fourth foot is dactylic, one or more of the preceding feet is spondaic to restore the balance. The student can verify this for himself on any page of the Aeneid.

It thus seems that Virgil, in adapting the Homeric hexameter to the Latin language, realised that its dactylic rhythm must be modified by a large admixture of 'spondei stabiles, as Horace calls them (A. P. 256). A considerable majority of his verses have at least three spondees (including the last foot); and the proportion of fifteen such lines in Aen. i. 1-20 to nine in Iliad i. 1–20 may be taken as a rough measure of the extent to which he carried out this modification of Homeric rhythm. A spondee in the first foot, contained in a single word and followed by a pause in sense, is almost the only circumstance under which he seems to shrink from spondaic rhythm in the first four feet : and the somewhat slow and ponderous movement thus given to the verse at starting is reserved, as a rule, for the special expression of solemnity or emotion (see note to Ecl. v. 21).

4. The hexameters of Ennius are a first experiment to reproduce, in a rough unpolished material, the rhythm of Homer. The con

ditions under which the metre could be adapted to Roman usage
had yet to be discovered : caesura, cadence, proportion-all the
niceties of rhythm which combine to form the charm of Virgil's
verse-were to him unknown. The rude and tentative imitation
of a great model by a vigorous and powerful hand struck out
indeed here and there a line which Virgil did not disdain to
borrow (e. g. 'Tuque pater Tiberine tuo cum flumine sancto '), or a
of grave solemnity, as the lament for Romulus

'o Romule, Romule die,
Qualem te patriae custodem di genuerunt!
O pater o genitor o sanguen dis oriundum,

Tu produxisti nos intra luminis oras' (Enn. Ann. 115-118): but it also produced much that was harsh and abhorrent to the culture of after years (see Hor. A. P. 258 sqq.), and much that could scarcely be distinguished from prose. But in settling the quantity of Latin words and moulding them into forms suitable for hexameter verse he paved the way for others, and was deservedly reverenced as the pioneer

‘qui primus amoeno Detulit ex Helicone perenni fronde coronam' (Lucr. i. 117). 5. Lucretius marks à great advance upon Ennius, though in some respects his rhythm is (perhaps intentionally) more archaic than that of his contemporary Catullus ; whose hexameters, however, with their monotonous cadence (“prognatae vertice pinus, Neptuni nasse per undas, Argivae robora pubis ' lxiv. 1, 2, 4), are far less effective. The following points of contrast between the Lucretian and Virgilian hexameter are noted by Munro : (1) the first two feet separated from the rest—Religionibus atque minis, Ergo vivida vis,' etc.; such rhythms being rare in Virgil (^ Armentarius Afer' G. iii. 344, 'Sed tu desine velle' G. iv. 448); (2) in the last two feet, such endings as 'principiorum,' 'materiai,' 'quandoquidem exstat;' (3) elision after the fourth foot—Perdelirum esse videtur, nisi concilio ante coacto;' (4) fourth foot wholly contained in a word, and ending with it—'quae terras frugiferentes,' tibi suaves daedala tellus. (not "terras quae ... suaves tibi '), etc.; (5) copious use of alliteration and assonance, occasionally adopted by Virgil under Lucretian influence.

6. The most common licences or metrical irregularities in Virgil are

(i.) Lengthening of short final syllables. This occurs only in arsi (i. e. in the emphatic syllable of a foot, upon which the metrical ictus falls), and seldom where there is not a pause


[ocr errors]

or slight break in the sentence : and it is used by him as a purely antiquarian ornament. With Ennius, on the contrary, whom Virgil seems to follow in this licence, the apparent violation of quantity as fixed in Augustan prosody is no 'licence,' because the syllables in question were originally long, and were subsequently shortened by a familiar tendency of the Latin language, due mainly to the fact that final syllables were never accented.

(a) Nouns, etc. in '-or;' 'Amor et' Ecl. x. 69, 'labör: aeque' G. iii. 118, melior insignis G. iv. 92. The corresponding Greek -wp, and the prosody of oblique cases (“amoris,' etc.), point to the original length of this syllable, which is always so in Ennius, even in thesi, i.e. the unemphatic syllable of a foot, e. g.

• Clamor ad caelum volvendus per aethera vagit' (Ann. 520). The same applies to 'patēr' (natúp) Aen. v. 521 : but 'puēr' (Ecl. ix. 66), 'supēr' (Aen. vi. 254), 'ebūr' (Aen. xii. 68), show that Virgil uses the licence as mere matter of form, with no thought of etymology.

(6) Verb terminations in 'r:''-or' of first pres. pass. is naturally long, and is so used by Ennius and Plautus. Virgil does not follow them in this; but has in 3 sing. “ingreditūr' (G. iii. 76), 'datūr' (Aen. v. 284), and in I plur. 'obruimūr' (Aen. ii. 211), neither of which has any precedent in Ennius.

(c) Noun terminations in '-s;' sanguis always in Lucretius, once only in Virgil (Aen. X. 487): '-ūs' from '.-' stems (G. iii. 189, Aen. v. 337, etc.) is found also in Ennius, perhaps from imitation of Homeric use in e. g. Iliad i. 244 χωόμενος, ότ' άριστον 'Αχαιών vůdèvětigas : 6-būs' dat. plur. (Aen. iv. 64) has no example in Ennius and few in Plautus.

(d) Verb terminations in '-s :' only 'fatigamus' (Aen. ix. 610), which finds no analogy in Ennius or in the corresponding Greek μες, μεν.

(e) Verb-endings in '-t' (3 sing.); "āt' of pres. ind. ist conj. generally in Ennius and often in Plautus, never in Virgil; '-āt' of imperf. in Plautus and Ennius even in thesi (“Noenum rumores ponebāt ante salutem' 314); in Virgil only in arsi (Ecl. i. 39, Aen. v. 853, xii. 722, etc.): '-ēt’ pres. indic. Aen. i. 308 ; imp. subj. ib. 651 : 6-īt' pres. indic. 3rd conj. Ecl. vii. 23, Aen. X. 433 ; 'erīt' (fut.) Ecl. iii. 97, Aen. xii. 883: '-īt' perf. indic. (as originally) G. ii. 211, Aen. viii. 363.

(f) Miscellaneous : 'procūl'Aen. viii. 98, 'capūt'x. 394. (8) Vowel-endings: only 'graviā'Aen. iii. 464, ‘Getā' (nom.



sing.) ib. 702, "animā' (nom. sing.) xii. 648; and thirteen instances of 'quē' (see on Aen. iii. 464), in imitation of Homeric usage making te long before double consonants, liquids, and sibilants.

[Full lists of examples, from Wagner, 'Quaest. Virg. xii, are given in Professor Nettleship’s ‘Excursus’to Aen. xii. (ed. Conington), and Dr. Kennedy's Appendix, C, II. (pp. 622-4, 2nd ed.)]

7. (ii.) Hiatus, i. e. non-elision of a vowel or diphthong before another vowel or aspirate. This licence appears from a statement of Cicero (Orator 45. 152 1) to have been common with the older Latin poets, and occurs frequently in the dialogue of Plautus and Terence ? : but it is used sparingly by later poets. The most common conditions for its admission are (1) after long monosyllables, retaining their quantity in arsi, or shortened in thesi, (e. g. the interjections 'heu !' 'o!' spē inimica' Aen. iv. 235, 'te Corydon ŏ Alexi' Ecl. ii. 85, 'ān qui àmant’ viii. 108, “tě amice' Aen. vi. 507): (2) at the regular caesuras ; (3) wherever there is a distinct pause in the sense (e. g. in dialogue, at the change of speaker).

Of the fifty-three examples cited by Wagner (Q. V. xi.) from Virgil, forty show the unelided syllable in arsi, and therefore metrically emphatic. Of thirteen examples in thesi, eleven are cases of a long vowel shortened (Ecl. ii. 65, iii. 79, vi. 44, viii. 108 ; G. i. 281, 332, 437, iv. 461 ; Aen. iii. 211, v. 261, vi. 507), which thus seems the necessary condition for hiatus in thesi ; the remaining two (Ecl. iii. 53, Aen. i. 405) being justified by a distinct pause in the sense. Of the forty examples in arsi, twentyseven occur at one or other of the two important caesurae, the penthemimeral and hephthemimeral (see above, footnote to $ 2): ten of the remainder being at the ennehemimeral caesura (e. g. ‘Amphion Dircaeus in Actaeo | Aracyntho’Ecl. ii. 24), in obvious imitation of the frequent Homeric cadence IInanıádew ’Axiños (II.

* Cicero is speaking of the tendency of Roman speech to run together vowels, contrary to Greek practice, which allows hiatus: "Sed Graeci yiderint; nobis ne si cupiamus quidem distrahere voces conceditur. Indicant, ... omnes poëtae praeter eos qui ut versum facerent saepe hiabant, ut Naevius “Vos qui accolitis Histrum flumen, atque algidam ...”' citing also from Ennius and his own poems.

2 Ritschl and others, who, by alteration of text, restoration of obsolete final consonants, etc. try to minimise hiatus in the comic writers, allow it in about one out of twenty-two lines in Plautus and one out of sixty-six in Terence. Wagner (Q. V. xi.) citęs fifty-three examples from Virgil.




i. 1). Five of these latter, unlike the bulk of Homeric examples, are spondaic endings (Ecl. viii. 53 'castaneae hirsutae,' Aen. iii. 74, vii. 631, ix. 647, xi. 31), due especially to the less dactylic character of the Latin language (see above, § 3); similar endings in Homer being generally quadrasyllabic words ('Ατρείδαο, Πηλειώνα, etc.) and never with trisyllable words embracing hiatus.

8. (iii.) Hypermetric lines—i. e. with an additional syllable after the final trochee or spondee; this syllable being always one that may be elided, while the next line must begin with a vowel or aspirate. This licence (resting apparently on a false assumption that the scansion of hexameter verses is continuous, as in Greek anapaests, and that the sixth foot is complete, i. e. a real spondee) was unknown to Homer; oủk oid at the close of a line of Callimachus being the only known instance in Greek hexameters. Lucretius employs it once (v. 849), Catullus rarely, in lyric metre -e. g. lxi. 147 (Glyconic), and perhaps xi. 19 (Sapphic)--but there omnium'may be dissyllable by synizesis, as 'precantia' Aen. vii. 237 (cp. 'omnia' vi. 33). It is found in an iambic line of Pacuvius preserved by Cicero, Tusc. iii. 12. 26; and occasionally in Terence --e. g. Phorm. ii. 1. 63 ; Ad. ii. 2. 9, iii. 3. 21 (iambic); And. iv. I. 9 (cretic); Eun. iv. I. II (trochaic). Wagner on G. ii. 69 hardly proves its use by Ennius : nor do Greek dramatists use it, as he says, 'infinitis locis. By whomever introduced into Latin hexameters it is a purely artificial licence, and as such is used by Virgil. In eighteen out of twenty-two instances in his poems, the hypermetric syllable is 'que;' in G. i. 295 he has decoquit umorjem,' and in Aen. vii. 160, 'tecta Latino]rum.' In all these twenty examples the preceding syllable is long, making the last foot a spondee: but in G. ii. 69, iii. 449, we have, if MSS. are to be trusted, hypermetric syllables preceded by trochees (see note to G. ii. 69).

9. Like all great masters of poetic rhythm, Virgil shows his power in the accommodation of sound to sense. Familiar examples of single lines are Aen. V. 481 (the sudden collapse of a stricken ox), viii. 452 (the steady swing of the Cyclops' hammers), viii. 596 (the sound of galloping horses): but the poet's art is also shown in passages of varied length, from the two lines expressive of the ‘moping owl's' complaint (Aen. iv. 462, 463), to the fine description of a storm in G. i. 316-334 (see especially 328-334). Among innumerable examples the following are noticeable : of single lines, G. ii. 441 (gusts of wind assaulting a tree), iii. 201 (the swift rush of the wind as of a horse let loose); of longer passages, G. i. 108-110

« PreviousContinue »