represents the results which arise out of rather than are definitely stated in Prof. Zimmer's pages, or that it fails to mark their importance and interest. Should I not have done justice to Prof. Zimmer in these respects it is from lack of skill and not of will. Nor should I fail to note that the value of his investigations depends only slightly upon the correctness of his results. He has cross-examined the documents far more searchingly than any previous scholar; he has been indefatigable in ransacking the records of the sixth-eleventh centuries with a view to providing an historical basis for this or that episode of the romances; he is always ingenious in detail, most ingenious perhaps when he is substantially contradicting himself. It will be understood that merits such as these cannot be adequately exhibited in the few pages at my disposal. Let me then note that much of the evidence is philological; thus, the forms of names in many French Arthurian romances are shown to be Breton and not Welsh, as is also the case, partly, with Geoffrey of Monmouth and William of Malmesbury. Prof Zimmer is a believer in the Northern locale of the original Arthur-tales, and makes ingenious use of the fact that this locale may readily be distinguished in the French romances—the Bretons, whose historical connection with Britain ceased with the seventh century, preserved it better than the Welsh, the centre of whose political history was shifted from Northern to South-Western Britain, and who gradually came to look upon Arthur as a South-Welsh chieftain. Welsh literature, even of the oldest class, is shown to be comparatively modern in its present form; for instance, the tale of Kulhwch and Olwen, and the Triads of the Horses in the Black Book, are shown to allude to post-Conquest personages.
When the reader frees himself from the avalanche of detail under which Prof Zimmer overwhelms him, he is apt, however, to ask himself if the result of the German scholar's labours is quite what the latter thinks it to be.
