what I would venture to call an anthropological borrowing theory. He has suggested the application of the latest theories of comparative philology to explain the phenomena of folk-lore. Comparative philology has before now had "new" theories, which were readily accepted in years gone by, and now they are rejected, not without some show of petulant scorn, by those who have learnt the new ways. At least it seems to me to be premature to accept the latest " ew" theories of comparative philology as a guide to folk-lorists. Why should folk-lore be perpetually asked to lean upon philology? I altogether reject an alliance upon such a basis. I believe that folk-lore has methods of its own quite as exact as those of philology, and that the true course to pursue is to keep to those methods. They are to be determined by folk-lore data, and not by the data of other sciences, however closely allied; they depend upon the inter-relationship of the various component parts of folk-lore, and must be ascertained and set forth in scientific order and precision—an order and precision attainable, as I believe, to a much greater degree of perfection than most of us have any idea of. It may well be that by its own methods folk-lore will be in a position to teach something to philology and the other allied sciences.
Thus, then, it seems that our work in the future must lie more and more in the direction of analysis and classification. To do this properly we want first of all absolutely exhaustive collection. Collection is twofold: (1) among the people for those items which are even yet unrecorded, as, for instance, such an item as Professor Haddon a week or two ago told me he had noted in Ireland—the custom of loosening the nails of a coffin just before consignment to the grave, so that the spirit may have less trouble in getting to the spirit world; (2) among the scattered literature not specially devoted to folk-lore. This last need has been noted in the Handbook, but I will recall to the members the admirable paper which Miss Burne sent up to the
