gives rise to further complication. The majority of the texts are in French, the oldest, paleographically speaking, are in Latin. That the whole superstructure is Celtic no one would affirm, but even the fact that it rests in any large measure upon a Celtic basis is denied by a school at the head of which stands Professor W. Förster of Bonn, and which counts among its ranks, inter alios, the most distinguished German disciple of Professor Bugge, Dr. Wolfgang Golther of Munich. The latter’s article, cited at the head of this report, sums up tersely and ably the views of this school. Moderate opposition thereto is made by the greatest living authority in this field of scholarship, Mons. Gaston Paris, and his group of pupils, chief among them Mons. E. Muret. The French school by no means advocates the Celtic nature of the Arthurian romance in such a thorough-going way as, say, Professor Rhys, perhaps because it has paid less attention to the Irish sources, the examination of which reveals with such startling clearness the fact that well-nigh all the leading scenes and situations of the Arthur myth have their parallels in the Irish saga of the 11th and preceding centuries, and may therefore be considered Pan-Celtic.
The most important work that has recently appeared in this field of research is undoubtedly Mons. Gaston Paris’ survey of the Arthurian metrical romance (Hist. Lit., vol. xxx). Nor is its interest exclusively or even mainly confined to the Celtic student; every “storyologist” should have it at hand. I may be permitted to trench for a moment on Mr. Hartland’s province, and to point out that a theory deriving the current European folk-tales from mediæval romance has far more to say for itself than the Indian importation hypothesis championed by Benfey and Mons. Cosquin and Mr. Clouston. Needless, then, to insist upon the value to the student of a work which affords a bird’s-eye view, admirable in its clearness and precision, over the most important and most intricate section of mediæval romance. One instance must suffice to illustrate how
