href="@public@vhost@g@gutenberg@html@files@46234@[email protected]#Footnote_3_3" class="fnanchor pginternal" tag="{http://www.w3.org/1999/xhtml}a">[3] and in this character the parallels carry us far back to the earliest stages of Celtic tradition as preserved in ancient Irish romance.

A special feature of these Breton lais, to be noted in this connection, is that they often combine two features which are more generally found apart, and which, as represented by their most famous mediæval forms, are wont to be considered by us as belonging to two different families of tradition, i.e., the Tannhäuser legend (the carrying off of a knight by the queen of the other world), and the Lohengrin legend (the rupture of a union between a mortal and an immortal, and the penalties incurred by the former by the transgression of a prohibition imposed by the latter). Two of the stories given in this volume, Guingamor and Lanval, in common with others which will be found noted in Dr. Schofield's studies, combine both motifs.

Now that such tales as these, in themselves independent popular folk-tales, sometimes became incorporated with, at other times by the loan of incident and feature strongly influenced, the Arthurian story, cannot I think be denied. Fairies such as the mistresses of Guingamor and Lanval were, as I have said above, residents or visitors at Arthur's court. Arthur himself is, like those knights, carried to Avalon; even as Guingamor in the extremity of mortal weakness. That like Guingamor he was thought of as recovering, and reigning with undiminished vigour over his fairy kingdom, is clear from numerous references in mediæval romance. The authors of La Bataille de Loquifer and Ogier le Danois knew him as King of Avalon; in Huon de Bordeaux he has been promised the reversion of Oberon's kingdom; in Lohengrin he reigns with Parzival, in a mysterious other-world realm; he is as completely lord of Fairyland as any knight beloved of fairy queen. The boyhood of Tyolet is the boyhood of Perceval; the mysterious stag guarded by lions wanders in and out of the mazes of Arthurian romance.

Some might, of course, suggest that these stories are really fragmentary borrowings from the Arthurian legend; but such a view is scarcely compatible with the fact that in their earlier forms they are entirely unconnected with that story. Thus we see that the lai of Guingamor in the solitary version we possess knows nothing of Arthur; neither the king or the queen, the fairy or her kingdom is named; Chrétien de Troyes knew the lady as Morgain, and her land as Avalon, and brings Guingamor to Arthur's court. The same remark applies to Graalent, while Lanval is in an Arthurian setting. If the stories had originally formed part of the cycle it is difficult to see why they should have been separated from it; while we can well understand that already existing folk-tales would be swept into the vortex of an increasingly popular tradition.

The story of Tyolet as preserved in the lai is certainly not in its earliest form; it is in some points incomprehensible, and as I have suggested in the Notes, the real meaning of the tale has been already forgotten. But Tyolet is never elsewhere mentioned as one of Arthur's knights, and the adventure achieved by him when transferred to Lancelot loses even the measure of coherence and plausibility it had preserved. Thus Lancelot, though knowing what is to be the guerdon of the successful knight, and voluntarily undertaking the adventure, when achieved, leaves the lady under the pretext of summoning his kinsmen and never returns; on no account would he be faithless to Guinevere.

In the Were-Wolf, again, the characters are anonymous; but Malory's reference leaves no room for doubt that the hero later on figured as one of Arthur's knights.

It is, I think, impossible to avoid the conclusion that the Arthurian legend, in the process of evolution, borrowed with both hands from already existing stores of popular folk-lore and tradition; and an examination of the parallels with this folk-lore element makes it equally clear that it was largely of Celtic origin.

But in what form was this popular tradition when the literary masterpieces of the Arthurian cycle, the poems of Chrétien de Troyes and his German rivals, were composed? We know that many of these tales were told as Breton lais, and in this original form they have practically disappeared. Those we possess are French translations, and of these the best and largest collection we owe to the skill and industry of Marie de France, an Anglo-Norman poetess who lived in the reign of Henry II. and was therefore a contemporary of Chrétien de Troyes. Of the four lais here given, two, Lanval and Were-Wolf (Bisclaveret), are undoubtedly by her, and Guingamor is very generally considered to be also her work. The metre in which she wrote was the eight-syllable verse, in rhymed pairs, adopted also by Chrétien in common with most of the poets of his time. As we see, Marie, like Chrétien, connected some of these lais with Arthur. They are Breton lais; Arthur is a Breton king; his legend certainly came to the Northern French poets partly, if not entirely, from Breton sources; the probability, therefore, is that the connection took place, in the first instance, on Breton rather than on French ground—i.e., it is due neither to Marie nor to Chrétien, but to the sources they used.

Setting hypothesis aside, however, this may be stated as an absolute matter of fact: at the time that the longer Arthurian romances took shape there were also current a number of short poems, both in Breton and in French, the latter in the precise metre adopted for the longer poems, connecting the Arthurian story with a great mass of floating popular folk-tale, which short poems were known to the writers of the longer and more elaborate romances. Are we seriously called upon to believe that they made absolutely no use of them? That they left all this wealth of material rigidly on one side, and combined for themselves out of their inventive faculties and classical knowledge the romances that won such deserved repute? Such a solution of the Arthurian problem I can scarcely think likely in the long run to be accepted by serious students; certainly not by those whom the study of comparative religion and folk-lore has taught how widely diffused in extension, and how persistent in character, are the tales which belong to the childhood of the race. That a large and important body of genuine existing tradition should be, not merely superseded, but practically beaten out of the field and destroyed by the power of mere literary invention, would be a curious phenomenon at any date; in the twelfth century it is absolutely inconceivable. The Arthurian legend has its roots in folk-tradition, and the abiding charm of its literary presentment is in reality due to the persistent vitality and pervasive quality of that folk-lore element. Children of a land of eternal youth, Arthur and his knights are ever young; it is true that some of the romances tell us that in the last great war with Lancelot Arthur was over ninety years old and Gawain above seventy, but one feels that even for the writer such figures had no significance; their words and actions are the words and actions of youth—we have here no Charlemagne and his veterans à la barbe fleurie.

But this is an element which in our rightful appreciation of the literary masterpieces of the cycle we are apt to ignore, nor is it other than scantily represented in English literature; it has therefore been thought well, in such a series as this to include a volume which shall direct attention less to the completed Arthurian epic than to the materials from which that epic was formed, since if we mistake not, it is to the nature of that material even more than to the skill of its