Page:Folk-lore - A Quarterly Review. Volume 3, 1892.djvu/401

This page needs to be proofread.

Celtic Myth and Saga. 393

in the i6th century Welsh tale of Taliesin, and has been boldly claimed as a late loan from the East. Its presence in an Irish tale, which can hardly be younger than the seventh century, and which is probably centuries older, shows how baseless this claim is.

The number of Irish texts which have been translated is considerable, but they are either scattered through the pages of specialist periodicals, or, as a rule, unaccom- panied by such critical comment as enables the layman to judge of their place in Celtic literature. It is the merit of M. D'Arbois de Jubainville's Epopee celtique to make the general reader free of a domain hitherto reserved for the specialist, or only thrown open to the public at large without the necessary sign-posts. A selection is given of the more important texts, the MS. sources are enumerated, the authenticity and development of the versions are discussed, the scientific value of the ancient Irish sagas is set forth briefly but clearly. I think too well of this book not to hope that a second edition may soon be called for, and in anticipation thereof would point out what I cannot but consider serious defects, defects that may, however, be easily remedied. A large amount of space is wasted upon French versions of Mac- pherson's Ossian. Whatever Macpherson's merits as a writer may be, he throws absolutely no light upon the origin and early form of the Gaelic heroic epos. Again, the volume, due to the collaboration of M. D'Arbois and his pupils, has not been brought up to date. Thus the trans- lation and comment upon the Voyage of Mael Duin, due to M. Ferd. Lot, date from before Prof. Zimmer's masterly account of the imniran literature (cf. ante, i, p^. 237). M. Lot cannot be blamed for not having anticipated Prof. Zim- mer's results, but the public is justly entitled to complain that he has not revised his study in accordance with the latest and best information.

I pass from the disagreeable task of fault-finding to the consideration of some views advanced by the editor which,