prise, a cycle of stories like the Carlovingian or the Arthurian, with Cambuscan as the central figure. It would have been a worthy enterprise for the last of the Trouvères. The old cycles of romance were hackneyed, worn out; otherwise one might wonder that the great poet of English chivalry never dealt with the legend of Arthur. His predecessors had exhausted every great name known to their histories. But Genghis Khan, whose fame filled Europe in the fourteenth century, was a new and tempting hero; and the far East was an untrodden field for the romancer. Cambuscan might well have been the centre of a new romantic cycle. Hence it strikes one as possible that Chaucer stopped short with the Squire's Tale because he had larger views, and put off completing it as he put off completing the full scheme of the 'Legend' and the 'Canterbury Tales,' because he shrank from long continuance of high-strung labour.

II. HIS LANGUAGE, METRES, AND IMAGERY.

Our philological authorities do not seem to be quite at one about Chaucer's English. Mr Earle (English Philology,' pp. 7597) says that Chaucer and Gower wrote King's English, the language that had grown up at Court about the person of the monarch; and that this was distinguished from all the contemporary dialects by its being formed more under the influence of French. This position is not refuted by counting the number of words derived from the French, as Mr Ellis does for the Prologue to the 'Canterbury Tales,' and finding that the proportion of words so derived is "not quite one word in a line on an average. It is not so much the number of words borrowed that Mr Earle insists upon, as the general strain or rhythm of the language. Not that he means to say that the King's English adopted the French rhythm, but that, growing up as it did among persons familiar with French, it acquired a rhythm of its own, different both from the French rhythm and from the rhythm of the provincial dialects. To understand this, compare the English of Chaucer or Gower with the English of Robert de Brunne, or of Langland's 'Piers Plowman.' Difference in the inflections and in the proportion of French words do not account for the immense indescribable difference in the general movement of the language. This movement, this rhythm, Mr Earle considers the distinctive feature of the King's English.

Whether Dr Morris would accept Mr Earle's position or not, I do not know. Dr Morris lays down that Chaucer wrote in the East Midland dialect, and so far is at variance with Mr Earle's statement, that the language of the Court differed from all other

B

dialects. But perhaps Dr Morris means only that Chaucer uses the inflectional system of the East Midland as distinguished from the Northern and the Southern. These dialects have been defined by Dr Morris with new precision.1 The Midland was the most widely spread, and of its many varieties the East Midland was the most important. It was first cultivated as a literary dialect as early as the beginning of the thirteenth century; and it had then thrown off most of the older inflections, so as to become in respect of inflectional forms and syntactical structure as simple as our own. It was the dialect of Orm and of Robert of Brunne. Wycliffe and Gower added considerably to its importance, and Chaucer's influence raised it to the position of the standard language. In Chaucer's time it was the language of the metropolis, and had probably found its way south of the Thames into Kent and Surrey. Such is Dr Morris's account of the East Midland. So far as appears, he has not been struck with differences of rhythm, and it has not occurred to him whether the language of the Court was different in other respects than inflections from the language of the metropolis. The probabilities seem to be overwhelming in support of Mr Earle's hypothesis, and justify him in saying that the facts admit of no other explanation. The frequenters of the Court, when they were ultimately forced to adopt the language of the people, could not but have found it poorly equipped for the varied needs of polite conversation; and even when they were not under the necessity of importing words from the more cultivated French, must have been compelled to introduce turns of expression to the extent of altering the complexion of the language.

Whatever may be our conclusion as to the sources of Chaucer's language, there can be little doubt that his genius made it the standard language. A poet cannot, of course, invent a language: what he writes must be intelligible to his readers, and his admirers are constrained by the same necessity of being intelligible to their readers. But if a great poet had arisen before Chaucer in the Northern dialect, or if Chaucer himself had written in that dialect, the course of the English language might have been substantially altered. In corroboration of this, Mr Earle remarks that "the Tuscan form of modern Italian was decided by the poetry of Dante, at a time when Florence and Tuscany lay in comparative obscurity, and when more apparent influence was exercised by Venice, or Naples, or Sicily." I suspect, however, that in all such cases the poet must be backed by a cultivated society; and that the only possibility likely to have affected the course of standard English would have been the existence of a high culture

1 Garnett and Guest, however, are still worth reading.

in the Court of Scotland, and the ascendancy of a great poet there before the date of Chaucer.

One of the many directions of thorough antiquarian study in this century has been towards the remains of old English; and one of the most valuable triumphs of this patient scholarship has been to restore the versification of Chaucer. Chaucer's metre had been a vexed question ever since he had been reduced to print. The change of pronunciation had seriously affected the number of syllables in his verses; and his editors, Thynne and Speght, while they valiantly abused detractors, could not, or at least did not, show how to supply the missing syllables. Ascham, Sidney, Spenser, and others of that age, recorded their admiration. of the old poet; but they omitted to say on what principles they scanned his lines. A century later, Dryden, with his vigorous habit of saying what he could not as well as what he could admit, expressed the greatest veneration for Chaucer, but considered that the rough diamond needed smoothing for modern senses. He laid down in the most positive manner that Chaucer's lines often have a syllable too many, and often a syllable too few. His peremptory opinion received a considerable shake from the publication of Tyrwhitt's edition in 1778, with its "Essay on the Language and Versification of Chaucer;" but even then a certain amount of scepticism might have been pardonable. One might have been forgiven for entertaining some doubts about Tyrwhitt's liberal restoration of final es, to make up for the deficiency of feet when the text is pronounced with the modern allowance of syllables. Tyrwhitt did not apply sufficiently convincing scholarship to show that these es were "survivals of French and old English (or Anglo-Saxon) terminations, and, still more largely, of old English inflections. This was first thoroughly done by Guest, who pointed out that the dropping of the final e is the exception, and expressed a hope that we should one day have a list of all the words in which Chaucer has taken that liberty. I am not aware that the more recent labours of Mr Ellis, Mr Skeat, and Mr Furnivall, have realised this hope, but they have gone a considerable way towards its fulfilment. The general reader should see the sections added to Tyrwhitt's Essay by Mr Skeat, in the Aldine edition of Chaucer, vol. i.

But not only have the missing syllables been recovered by modern scholarship: an attempt has been made by Mr Ellis, in his elaborate and ingenious work on Early English Pronunciation, to recover the sounds of the vowels, so that Chaucer may be declaimed as he was by his contemporaries. Mr Ellis proceeds on the supposition that no two words were set down as rhymes unless their sounds agreed perfectly; and he starts from known

vowel-sounds of Latin, French, and English of the sixteenth century. The investigation is interesting, and the results are as trustworthy as the undertaking is courageous; but nothing can restore for us the old music of Chaucer's verse. It is musical to us-exquisitely so-but the music is not the music that delighted the Court of Richard II. We may learn to repeat the articulations of his contemporaries, but we cannot hear with their ears.

Chaucer has three principal metres: four-accent couplets (the metre of Milton's "Comus"), in the "Romance of the Rose," "Chaucer's Dream," the "Book of the Duchess," and the "House of Fame"; rhyme-royal 1-the metre of Shakespeare's "Lucrece -(a stanza of seven decasyllables, with rhymes of 1, 3+2, 4, 5+ 6, 7), in the "Court of Love," the "Parliament of Birds," the "Flower and the Leaf," four of the Canterbury Tales,' and "Troilus and Cresside"; and heroic, or five-accent couplets, in the "Legend of Good Women," and most of the 'Canterbury Tales.'

The four-accent couplet is the original metre of the Roman de la Rose and nearly all the French fabliaux, and was the most common metre in English poetry during the thirteenth and the fourteenth centuries. It is the prevailing form in Dr Morris's 'Specimens of Early English,' being used for all kinds of narrative matter, fables, romances, tales, 'Sunday Sermons,' and political songs. Its facility is rather a temptation than a check to a loquacious narrator; and Chaucer dances along in his jingling fetters with the greatest vivacity, as if his fertile invention wanted some severer restraint. He dashes off every now and then into digressive reflections, and recalls himself with an effort, as if he could hardly refrain from throwing in a few more of the facile couplets before he had done.

"Troilus verse was the favourite form of the fifteenth century, and was frequently used in the sixteenth. It was, indeed, the great stanza of English poetry, till Spenser superseded it with the fuller music of his more elaborate structure, and is a simpler

1 "The epithet royal seems to be derived from the chant-royal of the French, a short poem in ballet-stave, written in honour of God and the Virgin Mary; and by which, according to French critics, the abilities of the king were tested in the poetical contests at Rouen."-Guest's English Rhythms, ii. 359. According to Warton, the title "Balad-Royal" was first used in English by Caxton, being applied to Herbert de Burgh's stanza in his Translation of Cato's Morals. Gascoigne calls the same stave "rythm-royal." It is sometimes said that the epithet "royal" was derived from the circumstance that the stave was used by James I. of Scotland in the King's Quhair'; but this is probably a later supposition based only upon the coincidence. If it had been the tradition, James VI. of Scotland would hardly have applied the term " ballet-royal" to the stave of eight, designating the seven-line stave-after its first memorable appearance English-Troilus verse.

and homelier dress corresponding with the youth of the English muse. It may be disputed in which of the two stanzas the greatest quantity of English poetry has been written; for while Troilus verse was used by many poets of the fourteenth century whose names even are seldom repeated now, the Spenserian was the favourite stanza of the great revival at the close of the eighteenth century, numbering among its patrons Shenstone, Thomson, and Beattie, Campbell, Scott, Byron, Shelley, and Keats. Chaucer used the seven-line stanza to embody graceful allegories, and tales of sweet pathos. His choice of this medium for the touching stories of Constance and Griselda, and the woful legends of Cecilia and little Nicholas, is approved by the later employment of it in Spenser's "Ruins of Time," Daniel's "Complaint of Rosamond," and Shakespeare's "Lucrece" and "Lover's Complaint." It is worth noting that though Lydgate adopted the same stanza for his tragical "Falls of Princes," Chaucer employs a heavier eight-line stanza for his Monk's "Tragedies," or brief commiserations of potentates that had fallen out of great prosperity. The seven-line stave with its single repeat and appended couplet, a form that lends itself naturally to the expression of graceful after-thought or irrepressible sob, is too quick a measure for the embodiment of statelier feeling. This was felt by Michael Drayton, when, after composing his poem on the wars of the Barons of Richard II. in Troilus verse, he recast the whole into heavier ottava rima. But ottava rima (the stanza of 'Don Juan'), though strong enough for the vigorous march of Drayton's narrative, would not have been sufficiently inwoven for the grave reflective sentiments of the Monk's Laments, which are written in eight banded lines (1, 3+2, 4, 5, 7+6, 8)—the Spenserian stanza without the concluding Alexandrine.

Five-accent couplets are more suited for comedy and the comic epic, than for tragedy and the grand epic. This can be called "heroic verse" only when heroism is taken to imply a minimum of dignified feeling. There is, doubtless, a certain strenuousness in its movement when the matter is heavy; it may be used to convey the impression of bold splendid energy: but dignity and stateliness are out of the question. If the ear attends to the rhyme at all, the expectation of it must be more or less of a distraction from the feeling of massive grandeur. With the lighter material of comedy, the regular beat of the rhyme is cheerful and animating, if the couplet is occasionally divided and other means are used to prevent the regularity from becoming tedious. If one were disposed to venture on fine distinctions, one might say that the ottava rima of Anster Fair' and 'Don Juan' is the peculiarly appropriate metre of the light epic, while the couplet is the predestined vehicle of dramatic comedy. In