Page:Folk-lore - A Quarterly Review. Volume 28, 1917.djvu/55

 Presidential Address. 23

reference to the literary records of the past traces the development of a custom from stage to stage. It might seem hardly necessary to formulate so obvious a principle of research were it not that the kind of material interesting to the folklorist, consisting in the sayings and doings of those whom Hume describes as "the lowest vulgar," is precisely such as official historians will be likely to slur over or_.misrepresent; so that a positive rule is required to remind us that the accidents of history are the opportunities of folklore. The historical method is Tylor's name for this straightforward way of hunting up the pedigree of a survival ; and, by way of illustration, he applies it very prettily to the explanation of the led horse at the soldier's funeral.^^ Historical research, then, in this plain sense of the term, has always been a main concern of this Society. We have enjoyed many demonstrations of the value of this method not only for constructive, but likewise for critical, purposes ; as, for instance, when Miss Burne, in a striking Presidential Address, showed us how, by the aid of recorded history, it was possible " to distinguish between one survival and another, between survivals from mediaeval days and survivals from totemic days, between local variations and radical differences.""' As for Gomme, his examination of the archives of British custom was so systematic and fruitful as to entitle him to rank high among the historians of this country. But it is not this aspect of his work that I propose to consider to-night. He was likewise a follower of the historical method in the sense in which it is contrasted with the evolutionary; and, since the relative value of these methods for the science of culture is even to-day by no means clear, it may be useful to enquire how the argument from historical connexion took shape under the hand of a great pioneer.

^=In "The Study of Custom,^' Macmillan s Magazine, xlvi. (1S82), 79. '^^ I-'olk-Loie, x.xi. (1910), 32.