THE
POEMS AND FRAGMENTS
OF
CATULLUS,

TRANSLATED IN THE METRES OF THE ORIGINAL

BY

ROBINSON ELLIS,

FELLOW OF TRINITY COLLEGE, OXFORD,
PROFESSOR OF LATIN IN UNIVERSITY COLLEGE, LONDON.

LONDON:
JOHN MURRAY, ALBEMARLE STREET.
1871.

LONDON:
BRADBURY, EVANS, AND CO., PRINTERS, WHITEFRIARS.

TO ALFRED TENNYSON.

PREFACE.

The idea of translating Catullus in the original metres adopted by the poet himself was suggested to me many years ago by the admirable, though, in England, insufficiently known, version of Theodor Heyse (Berlin, 1855). My first attempts were modelled upon him, and were so unsuccessful that I dropt the idea for some time altogether. In 1868, the year following the publication of my larger critical edition[A] of Catullus, I again took up the experiment, and translated into English glyconics the first Hymenaeal, Collis o Heliconici. Tennyson's Alcaics and Hendecasyllables had appeared in the interval, and had suggested to me the new principle on which I was to go to work. It was not sufficient to reproduce the ancient metres, unless the ancient quantity was reproduced also. Almost all the modern writers of classical metre had contented themselves with making an accented syllable long, an unaccented short; the most familiar specimens of hexameter, Longfellow's Evangeline and Clough's Bothie of Tober-na-Vuolich and Amours de Voyage were written on this principle, and, as a rule, stopped there. They almost invariably disregarded position, perhaps the most important element of quantity. In the first line of Evangeline—

there are no less than five violations of position, to say nothing of the shortening of a syllable so distinctly long as the i in primeval. Mr. Swinburne, in his Sapphics and Hendecasyllables, while writing on a manifestly artistic conception of those metres, and, in my judgment, proving their possibility for modern purposes by the superior rhythmical effect which a classically trained ear enabled him to make in handling them, neglects position as a rule, though his nice sense of metre leads him at times to observe it, and uniformly rejects any approach to the harsh combinations indulged in by other writers. The nearest approach to quantitative hexameters with which I am acquainted in modern English writers is the Andromeda of Mr. Kingsley, a poem which has produced little effect, but is interesting as a step to what may fairly be called a new development of the metre. For the experiments of the Elizabethan writers, Sir Philip Sidney and others, by that strange perversity which so often dominates literature, were as decidedly unsuccessful from an accentual, as the modern experiments from a quantitative point of view. Sir Philip Sidney has given in his Arcadia specimens of hexameters, elegiacs, sapphics, asclepiads, anacreontics, hendecasyllables. The following elegiacs will serve as a sample.

In these the classical laws of position are most carefully observed; every dactyl ending in a consonant is followed by a word beginning with a vowel or h—afflīctĭŏn holdeth, momēnt ŏf hĭs anguish, caūse ŏf hĭs onely; affliction wasteth, moment of his dolour, cause of his dreary, would have been as impossible to Sir Philip Sidney as moērŏr tĕnebat, momēntă pĕr curae, caūsă vĕl sola in a Latin writer of hexameters. Similarly where the dactyl is incided after the second syllable, the third syllable beginning a new word, the utmost care is taken that that word shall begin not only with a syllable essentially short, but, when the second syllable ends in a consonant, with a vowel: ōf thĭs ĕpistle, but not ōf thĭs dĭsaster, still less ōf thĭs dĭrection. The other element of quantity is less rigidly defined; for (1) syllables strictly long, as I, thy, so, are allowed to be short; (2) syllables made long by the accent falling upon them are in some cases shortened, as rŭīne, pĕrĭshēd, crŭēl; (3) syllables which the absence of the accent only allows to be long in thesi, are, in virtue of the classical laws of position, permitted to rank as long elsewhere—momēnt of his, ōf this epistle. It needs little reflection to see that it is to one or other of these three peculiarities that the failure of the Elizabethan writers of classical metres must be ascribed. Pentameters like

sapphics like

hexameters like

are too alien from ordinary pronunciation to please either an average reader or a classically trained student. The same may be said of the translation into English hexameters of the two first Eclogues of Virgil, appended by William Webbe to his Discourse of English Poetrie (1586, recently reprinted by Mr. Arber). Here is his version of Ecl. I., 1-10.

The following stanzas are from a Sapphic ode into which Webbe translated, or as we should say, transposed the fourth Eclogue of Spenser's Sheepheardes Calendar.

There are many faults in these verses; over quaintnesses of language, constructions impossible in English, quantities of doubtful correctness, harsh elisions, for Webbe has tried even elisions. Yet, if I may trust my judgment, all of them can still be read with pleasure; the sapphics may almost be called a success. This is even more true of metres, where these faults are less perceptible or more easily avoided, for instance, Asclepiads. Take the verses on solitariness, Arcadia, B. II. fin.

or the hendecasyllables immediately preceding,

It is obvious that a very little more trouble would have converted these into very perfect and very pleasing poems. Had Sir Philip Sidney written every asclepiad on the model of Where man's mind hath a freed consideration, every hendecasyllable like Where sweet graces erect the stately banner, the adjustment of accent and quantity thus attained might, I think, have induced greater poets than he to make the experiment on a larger scale. But neither he nor his contemporaries were permitted to grasp as a principle a regularity which they sometimes secured by chance; nor, so far as I am aware, have the various revivals of ancient metre in this country or Germany in any case consistently carried out the whole theory, without which the reproduction is partial, and cannot look for a more than partial success. Even the four specimens given in the posthumous edition of Clough's poems, two of them elegiac, one alcaic, one in hexameters, though professedly constructed on a quantitative basis, and, in one instance (Trunks the forest yielded, with gums ambrosial oozing, &c.) combining legitimate quantity (in which accent and position are alike observed) with illegitimate (in which position is observed, but accent disregarded) into a not unpleasing rhythm, cannot be considered as more than imperfect realizations of the true positional principle. Tennyson's three specimens are, at least in English, still unique. It is to be hoped that he will not suffer them to remain so. Systems of Glyconics and Asclepiads are, if I mistake not, easily manageable, and are only thought foreign to the genius of our language because they have never been written on strict principles of art by a really great master.

What, then, are the rules on which such rhythms become possible? They are, briefly, these:—(1) accented syllables, as a general rule, are long, though some syllables which count as long need not be accented, as in

blossoms, though only accented on the first syllable, counts for a spondee, the shortness of the second o being partly helped out by the two consonants which follow it; partly by the fact that the syllable is in thesi; (2) the laws of position are to be observed, according to the general rules of classical prosody: (a) dactyls terminating in a consonant like beautiful, bounteous, or ending in a double vowel or a diphthong like all of you, surely may, come to thee, must be followed by a word beginning with a vowel or y or h; dactyls terminating in a vowel or y, like slippery, should be followed, except in rare cases, by words beginning with a consonant; trochees, whether composed of one word or more, should, if ending in a consonant, be followed by a vowel, if ending in the vowel a, by a consonant, thus, planted around not planted beneath, Aurora the sun's not Aurora a sun's (see however, lxiv. 253), but unto a wood, any again, sorry at all, you be amused. (b) Syllables made up of a vowel followed by two or more consonants, each of which is distinctly heard in pronunciation, as long, sins, part, band, waits, souls, ears, must, heart, bright, strength, end, and, rapt, hers, dealt, moment, bosoms, answers, mountains, bearest, tumbling, giving, coming, harbouring, difficult, imminent, stratagems, utterance, happiest, tremblingly, can never rank as short, even if unaccented and followed by a vowel, h or y. Thus, to go back to Longfellow's line,

forĕst, murmurĭng, pines ănd the, are all inadmissible. But where a vowel is followed by two consonants, one of which is unheard or only heard slightly, as in accuse, shall, assemble, dissemble, kindness, compass, affect, appear, annoy, or when the second or third consonant is a liquid, as in betray, beslime, besmear, depress, dethrone, agree, the vowel preceding is so much more short than long as to be regularly admissible as short, rarely admissible as long. On this principle I have allowed disōrdĕrly̆, tēnăntlĕss, heavĕnly̆, to rank as dactyls.

These rules are after all only an outline, and perhaps can never be made more. It will be observed that they are more negative than positive. The reason of this is not far to seek. The main difference between my verses and those of other contemporary writers—the one point on which I claim for myself the merit of novelty—is the strict observance throughout of the rules of position. But the strict observance of position is in effect the strict avoidance of unclassical collocations of syllables: it is almost wholly negative. To illustrate my meaning I will instance the poems written in pure iambics, the Phaselus ille and Quis hoc potest uidere. Heyse translates the first line of the former of these poems by

and this would be a fair representation of a pure iambic line, according to the views of most German and most English writers. Yet not only is Die no short syllable, but ihr, itself long, is made more hopelessly long by preceding three consonants in schauet, just as the last syllable of schauet, although in itself short, loses its right to stand for a true short in being followed by the first consonant of liebe. My own translation,

whatever its defects, is at least a pretty exact representation of a pure iambic line. xxix. 6-8, are thus translated by Heyse:—

by me thus,

The difference is purely negative; I have bound myself to avoid certain positions forbidden by the laws of ancient prosody. To some I may seem to have lost in vigour by the process; yet I believe the sense of triumph over the difficulties of our language, the satisfaction of approaching in a novel and perceptibly felt manner one of those excellences which, as much as anything, contributes to the permanent charm of Catullus, his dainty versification, will more than compensate for any shortcomings which the difficulty of the task has made inevitable. The same may be said of the elaborately artificial poem to Camerius (c. lv.), and the almost unapproachable Attis (c. lxiii.). Here, at least half the interest lies in the varied turns of the metre; if these can be represented with anything like faithfulness, the gain in exactness of prosody is enough, in my judgment, to counterbalance the possible loss of freedom in expression.

There is another circumstance which tends to make modern rules of prosody necessarily negative. Quantity, in English revivals of ancient metre, depends not only on position, but on accent. But accent varies greatly in different words; heavy level ever cometh any, have the same accent as empty evil either boometh penny; but the first syllable in the former set of words is lighter than in the latter. Hence, though accented, they may, on occasion, be considered and used as short; as, on the same principle, dolorous stratagem echoeth family, usually dactyls, may, on occasion, become tribrachs. But how lay down any positive rule in matter necessarily so fluctuating? We cannot. All we can do is to refuse admission as short syllables to any heavier accented syllable. Here, then, much must be left to individual discretion. My translation of the Attis will best show my own feeling in the matter. But I am fully aware that in this respect I have fallen far short of consistency. I have made any sometimes short, more often long; to, usually short, is lengthened in lxi. 26, lxvii. 19, lxviii. 143; with is similarly long, though not followed by a consonant, in lxi. 36; given is long in xxviii. 7, short in xi. 17, lxiv. 213; are is short in lxvii. 14; and more generally many syllables allowed to pass for short in the Attis are elsewhere long. Nor have I scrupled to forsake the ancient quantity in proper names; following Heyse, I have made the first syllable of Verona short in xxxv. 3, lxvii. 34, although it retains its proper quantity in lxviii. 27. Again, Pheneos is a dactyl in lxviii. 111, while Satrachus is an anapaest in xcv. 5. In many of these instances I have acted consciously; if the writers of Greece and Rome allowed many syllables to be doubtful, and almost as a principle avoid perfect uniformity in the quantity of proper names, a greater freedom may not unfairly be claimed by their modern imitators. If Catullus could write Pharsăliam coeunt, Pharsălia regna frequentant, similar license may surely be extended to me. I believe, indeed, that nothing in my translation is as violent as the double quantity just mentioned in Catullus; but if there is, I would remind my readers of Goethe's answer to the boy who told him he had been guilty of a hexameter with seven feet, and applying the remark to any seeming irregularities in my own translation would say, Lass die Bestie stehen.

It would not be difficult to swell this Preface by enlarging on the novelty of the attempt, and indirectly panegyrising my own undertaking. I doubt whether any real advantage would thus be gained. If I have merely produced an elaborate failure, however much I might expatiate on the principles which guided me, my work would be an elaborate failure still. I shall therefore say no more, and shall be contented if I please the, even in this classically trained country, too limited number of readers who can really hear with their ears—if, to use the borrowed language of a great poet, I succeed in making myself vocal to the intelligent alone.

CATULLUS.

I.

II.

III.

IV.

1.

2.

V.

VI.

VII.

VIII.

IX.

X.

1.

2.

3.

XI.

XII.

XIII.

XIV.