century, despite (or perhaps because of) the enthusiasm which one or two devotees have shown for their literary qualities, it does not seem to me that fair justice, or anything like it, has yet been generally done. German critics care little for literary merit, and are perhaps not often trained to appreciate it; in England the chansons have been strangely little read. But the most singular thing is the cold reception, slightly if at all thawed recently, which they have met in France itself. It may give serious pause to the very high estimate generally entertained of French criticism by foreigners to consider this coldness, which once reached something like positive hostility in M. Ferdinand Brunetière, the chief French literary critic of our generation. I regret to see that M. Lanson, the latest historian of French literature, has not dared to separate himself from the academic grex. "On ne saurait nier,” he says, "que quelques uns aient eu du talent;" but he evidently feels that this generous concession is in need of guards and caveats. There is no "beauté formelle" in them, he says-no formal beauty in those magnificently sweeping laisses, of which the ear that has once learnt their music can no more tire thereafter than of the sound of the sea itself. The style (and if it be objected that his previous words have been directly addressed to the later chansons and chanson writers, here he expressly says that this style "est le même style que dans le Roland," though "moins sobre, moins plein, moins sur ") has "no beauty by itself," and finally he thinks that the best thing to do is "to let nine-tenths of the chansons follow nine

tenths of our tragedies." I have read many chansons and many tragedies; but I have never read a chanson that has not more poetry in it than ninety-nine French tragedies out of a hundred.

The fact is that it is precisely the beauté formelle, assisted as it is by the peculiar spirit of which so much has been said already, which constitutes the beauty of these poems: and that these characteristics are present, not of course in uniform measure, but certainly in the great majority of the chansons from Roland to the Bastard. Of course if a man sits down with a preconceived idea of an epic poem, it is more likely than not that his preconceived idea will be of something very different from a chanson de geste. And if, refusing to depart from his preconceived idea, and making that idea up of certain things taken from the Iliad, certain from the Eneid, certain from the Divina Commedia, certain from Paradise Lost,-if he runs over the list and says to the chanson, "Are you like Homer in this point? Can you match me Virgil in that?" the result will be that the chanson will fail to pass its examination.

But if, with some knowledge of literature in the wide sense, and some love for it, he sits down to take the chansons as they are, and judge them on their merits and by the law of their own poetical state, then I think he will come to a very different conclusion. He will say that their kind is a real kind, a thing by itself, something of which if it were not, nothing else in literature could precisely supply the want. And he will decide further that while the best

of them are remarkably good of their kind, few of them can be called positively bad in it. And yet again, if he has been fortunately gifted by nature with that appreciation of form which saves the critic from mere prejudice and crotchet, from mere partiality, he will, I believe, go further still, and say that while owing something to spirit, they owe most to form itself, to the form of the single-assonanced or monorhymed tirade, assisted as it is by the singular beauty of Old French in sound, and more particularly by the sonorous recurring phrases of the chanson dialect. No doubt much instruction and some amusement can be got out of these poems as to matters of fact: no doubt some passages in Roland, in Aliscans, in the Couronnement Loys, have a stern beauty of thought and sentiment which deserves every recognition. But these things are not all-pervading, and they can be found elsewhere: the clash and clang of the tirade are everywhere here, and can be found nowhere else.

86

CHAPTER III.

THE MATTER OF BRITAIN.

ATTRACTIONS OF THE ARTHURIAN LEGEND DISCUSSIONS ON THEIR SOURCES THE PERSONALITY OF ARTHUR-THE FOUR WITNESSESTHEIR TESTIMONY -THE VERSION OF GEOFFREY ITS LACUNÆ→→→ HOW THE LEGEND GREW-WACE-LAYAMON-THE ROMANCES PROPER -WALTER MAP-ROBERT DE BORRON-CHRESTIEN DE TROYESPROSE OR VERSE FIRST?-A LATIN GRAAL-BOOK-THE MABINOGION THE LEGEND ITSELF THE STORY OF JOSEPH OF ARIMATHEA MERLIN-LANCELOT THE LEGEND BECOMES DRAMATIC-STORIES OF GAWAIN AND OTHER KNIGHTS-SIR TRISTRAM-HIS STORY ALMOST CERTAINLY CELTIC-SIR LANCELOT-THE MINOR KNIGHTS-ARTHUR -GUINEVERE-THE GRAAL-HOW IT PERFECTS THE STORY-NATURE OF THIS PERFECTION-NO SEQUEL POSSIBLE-LATER EPISODES THE LEGEND AS A WHOLE-THE THEORIES OF ITS ORIGIN-CELTIC— FRENCH-ENGLISH-LITERARY-THE CELTIC THEORY-THE FRENCH CLAIMS-THE THEORY OF GENERAL LITERARY GROWTH-THE ENGLISH OR ANGLO-NORMAN PRETENSIONS-ATTEMPTED HYPOTHESIS.

To English readers, and perhaps not to English readers only, the middle division of the three great romancesubjects ought to be of far higher interest than the

1 See the quotation from Jean Bodel, p. 26, note. The literature of the Arthurian question is very large; and besides the drawbacks referred to in the text, much of it is scattered in periodicals. The most useful recent things in English are Mr Nutt's Studies on the

Attractions of

Legend.

others; and that not merely, even in the English case, for reasons of local patriotism. The medithe Arthurian æval versions of classical story, though attractive to the highest degree as evidence of the extraordinary plastic power of the period, which could transform all art to its own image and guise, and though not destitute of individual charm here and there, must always be mainly curiosities. The cycle of Charlemagne, a genuine growth and not merely an incrustation or transformation, illustrated, moreover, by particular examples of the highest merit, is exposed on the one hand to the charge of a certain monotony, and on the other to the objection that, beautiful as it is, it is dead. For centuries, except in a few deliberate literary exercises, the king à la barbe florie has inspired no modern singer-his geste is extinct. But the Legend of Arthur, the latest to take definite form of the three, has shown

Legend of the Holy Grail (London, 1888); Professor Rhys's Arthurian Legend (Oxford, 1891); and the extensive introduction to Dr Sommer's Malory (London, 1890). In French the elaborate papers on different parts which M. Gaston Paris brings out at intervals in Romania cannot be neglected; and M. Loth's surveys of the subject there and in the Revue Celtique (October 1892) are valuable. Naturally, there has been a great deal in German, the best being, perhaps, Dr Kölbing's long introduction to his reprint of Arthour and Merlin (Leipzig, 1890), Other books will be mentioned in subsequent notes; but a complete and impartial history of the whole subject, giving the contents, with strictly literary criticism only, of all the texts, and merely summarising theories as to origin, &c., is still wanting, and sorely wanted. Probably there is still no better, as there is certainly no more delightful, book on the matter than M. Paulin Paris's Romans de la Table Ronde (5 vols., Paris, 1868-77). The monograph by M. Clédat on the subject in M. Petit de Julleville's new History (v. supra, p. 23, note) is unfortunately not by any means one of the best of these studies.