think alike, but men at different periods indulged different fancies about the same matter. And the Irish writers are disconcerting. When we have satisfactorily proved to our own mind that such and such was the theory of the Irish race on a certain subject at a certain time, some chance phrase or obscure passage springs up before us and belies all we have done. There is hardly any theory that cannot by some occasional phrase be overset; not even the theory I am about to propound to you to-night. That is the disadvantage of a copious literary output. I may as well say at once that in regard to the question that we have to discuss to-night, namely, whether the pagan Celt believed in a Hades in the sense in which M. D'Arbois de Jubainville and Professor Rhys invariably use the word, as a place of departed spirits, a land of shades and of death, a dark land ruled by what Rhys constantly calls the "dark divinities," the gods of death and of night, as opposed to the gods of light and knowledge and life—while I utterly disagree with their main theory, and hold that the Irish Gael, at all events, and probably his Welsh and Gallic cousins, were not at all possessed by such an idea, did not, in fact, so far as I can see, in general believe in a world of departed spirits at all, much less believe in it as a place of gloom and darkness, there are one or two passages which seem to contradict this theory and make distinctly for the belief held by Rhys and de Jubainville. But these passages are so rare and so surprising that to build a theory upon them seems to me to get matters out of proportion altogether. Most of them are obscure, and may almost as easily be interpreted in another way; indeed, many points relied upon seem to me to bear a quite different meaning, as I hope to show by one or two examples. It seems to me as dangerous to build up a theory from a single passage (and there is only one explicit passage brought forward, which I shall now at
