Page:Folk-lore - A Quarterly Review. Volume 28, 1917.djvu/58

 26 Presidential Address.

and Descent."'" Gomme's special merit, however, consists in havinj^ formulated the principle of method that institu- tions need, first and foremost, to be studied in their local context. Intensive ethnographical research is the necessary prius of comparative ethnology. Before we proceed to trace Iiistorical connexions between different areas of culture on the strength of the geographical distribution of customs, we must have worked out the topographical distribution''of customs within the several areas concerned, so as to make sure that in each case the things to be com- pared are themselves envisaged in the light of their authentic development. Such a method, then, as applied to a region with a recorded past such as this countr)-, will be historical in two senses at once ; because it is the only way of proving the historical transmission of customs, and at the same time because it involves the testing of each custom by its historical pedigree. It is likewise essentially sociological, since it insists that social organization rather than belief or stor}^ brings us directly into touch with that continuous life of the people of which the various customs are but the expression.

Further, such a method is no less characteristically ethnological. Even if we concentrate on a single area, we can hardly fail to discover, in its institutional history, the effects of culture-contact. We are proud to remember that under Gomme's Presidency this Society was to the fore in promoting an ethnological survey of Britain.-^ Gomme's own work, too, had led him straight to the explanation of the British village-community in terms of culture-contact. Into the particular merits of this explanation we cannot go now ; but it will serve as an excellent example of an ethnological hypothesis as employed by the historical method of folklore. Having tried to eliminate the effects

'"\x\ Journ. Aiith. lust, xviii. (lS88), 245 f. ; see Gomme in Folk-Lorc, ii. .(1S91), 4^S7.

"■'Compare Folk-Lorc, v. (1S94), 50.