em of totemism. Totemism, at one stage or another
of its development, is, in fact, one of the universal elements of
man's life, and all consideration of its traces in civilised countries
must begin with some conception of its origin. Its origin must refer
back to conditions of human life which are also universal. Special
circumstances, special peoples, special areas could not have produced
totemism unless we proceed to the somewhat violent conclusion that
beginning in one area it has spread therefrom to all areas. I know of
no authority who advocates such a theory and no evidence in its
favour. We are left therefore with the proposition that the origin of
totemism must be sought for in some universal condition of human life
at one of its very early stages, which would have produced a state of
things from which would inevitably arise the beliefs, customs, and
social organisations which are included under the term totemism.

There is therefore ample ground for a consideration of anthropological
conditions as part of the necessary equipment of the study of folklore
as an historical science. Unfortunately, authorities are now greatly
divided on several important questions in anthropology, and it is not
possible to speak with even a reasonable degree of certainty on many
things. This compels further research than the mere statement of the
present position, and I find myself obliged even for my present
limited purpose to suggest many new points beyond the stage reached
by present research. There is one advantage in this. It allows of a
hypothesis by which to present the subject to the student, and a
working hypothesis is always a great advantage where research is not
founded entirely on actual observation by trained experts in the
field. Where, therefore, I depart from the guidance of conclusions
already arrived at by scholars in this department of research, it will
be in order to substitute an opinion of my own which I think it is
necessary to consider, and the whole study of the anthropological
problems in their relation to folklore will assume the shape of a
restatement of the entire case.

I am aware that a subject of this magnitude is too weighty and
far-reaching to be properly considered in a chapter of a book not
devoted to the single purpose, but it is necessary to attempt a rough
statement of the evidence, though it will take us somewhat beyond the
ordinary domain of folklore; but, while dealing with the
anthropolo