endings, were both introduced into it from France. The foreign versification was naturalized, after many preludes, by Chaucer. Almost at once it then became dominant; soon it drove the native English type underground, only to reappear as a substantive thing in occasional reversions, but continuing to exercise a great influence over the rhythm and handling of the new poetry; so that English verse is a different thing from French, as Latin verse (though to a rather less extent) is a different thing from Greek.

To adapt Latin to the new rhythm and the quantitative treatment was a very difficult task. Not only did it mean forcing the native rhythms into a new mould; it meant a very serious loss of poetical vocabulary. A Latin poet writing in hexameter was absolutely debarred from the use of many words which were the very staple of his language. Words like civitas, vigilia, providentia, veritas, contumelia, multitudo, testimonium cannot be got into hexameter verse at all. With countless words of common and almost necessary use, he was restricted to one or two of their many inflections. He could, to take a few actual instances from Virgil himself, use imago, but

not imagines, exercitus as a nominative, but not as a genitive; he could say destinat, imprecor, adfert, invideo, but he could not say destinavit imprecatur, adferunt, invidere. In the first chapter of Livy's History, which gives a rapid sketch of the Aeneas-legend, and in which the general vocabulary, and many of the actual words and phrases, are almost the same as those of the contemporary Aeneid, there are, in some fifty lines, twenty-five words which Virgil could not use, besides half a dozen more which he could use only by very awkward elisions. The tendency therefore was for Latin hexameter verse to become cramped, restricted, and monotonously artificial in its diction.

Lucretius had already done miracles with this recalcitrant material. He had by sheer force of genius hammered Latin out into the new mould; he had made the hexameter a vehicle which could and did express close argument, splendid eloquence, thrilling emotion. It remained for Virgil to take the one further step and to transmute this bronze into gold.

He did so, and in a sense, if we look to what happened to Latin poetry after him, almost too successfully. He exhausted the possibilities of his medium. The artifice, though not

the art, of the Virgilian hexameter became the common property of all his successors. Ovid, only a few years later, handles it with amazing adroitness and what seems like effortless ease. But for Virgil himself it remained to the last a matter of perpetual labour and endless experiment. By elaboration of periodic structure, by constant variation of stress and pause, by avoidance of tripping runs and heavy masses, by innumerable verbal or syntactical devices, he made a stubborn material flexible and supple: he gave the language a new music. In this he is like Milton in English poetry. But their methods were different. Milton, debarred by blindness from the help of pen and paper, had to compose Paradise Lost in his head and dictated a passage only when he had got it into satisfactory form. Wordsworth, it is interesting to note, did the same thing without the same reason; he did not draft on paper, and when by mental labour that amounted to agony he had got a poem into shape, would make his sister or his wife write it down from his dictation. Virgil, we are told, wrote a first draft and then worked on it until perhaps there was not a word of the first draft left. The Aeneid at his death was still full of gaps and stop-gaps,

of alternative lines, of passages tentatively cancelled or provisionally inserted, of erasures and interlineations. The editors made, as it seems, no detailed record of what they actually did, whether in choosing between alternative drafts, or in piecing together fragments, or in omitting altogether passages which were unplaced or too obviously incomplete. There are either fifty-five or fifty-six unfinished lines in the Aeneid as they published it.43 But in many other places the substance is incomplete though there is no gap in the metre; there are several paragraphs which seem disconnected, and a few passages which are barely if at all grammatical; there are lines, as well as mere beginnings of lines, which are only rough sketches; there are some metrical inconsistencies; there are tags, sometimes put in to fill a gap, sometimes imperfect provisional solutions of metrical difficulties. It is only by appreciation of this and such appreciation comes only by long familiarity and minute study that the genius of Virgil as a master of language can be fully grasped. But it repays the labour. It is a fruitful lesson in the art of poetry.

Almost from the first, Virgil's ear was fault

less, his melodiousness exquisite. The progress to be traced in his art is the development of richer harmonies through a larger handling and an ampler movement. The magical single lines of his early work —

Molli paulatim flavescet campus arista

Audiit Eurotas iussitque ediscere lauros

Hic gelidi fontes, hic mollia prata, Lycori

melt more and more into periods; the music gains more and more of smooth, sweet and rushing freedom. As a masterpiece of full periodic movement, the invocation at the opening of the Georgics has never been surpassed. But in his later work at its finest, the art is further sublimated. It even tends to discard the rich ornamentation which is a steady note of Virgil's middle period. In his earlier poetry its use wavers, and his touch on it is uncertain: phrases of almost bald simplicity are mixed with others in which one may feel the decoration to be over-loaded. In the Georgics he has fully mastered this among the other problems of technique; and we may say unhesitatingly that both in style and diction he has reached perfection. It would be idle to cite instances