belief; it
developed highly specialised forms, took its part in the formation of
a great social force, a great fighting and conquering force, a great
migratory force. In accomplishing this task it grew into a solid
system, each part in touch with all other parts, each part an
essential factor in the ever-active forces which it helped to fashion
and control.

It is in this wise that we must study its survivals wherever they are
to be found, and the study must be concentrated within certain
definite ethnographic areas. If I were to pursue the subject and
choose for my study the folklore of Britain, I should have to object
to the treatment accorded to British custom, rite, and belief by even
so great an authority as Mr. Frazer, because they are used not as
parts of a tribal system but as mere detritus of a primitive system of
science, or philosophy. According to my views they had long since
become separated from any such system and it is placing them in a
wrong perspective, giving them a false value, associating them with
elements to which they have no affinity to divorce them from their
tribal connection. The custom, rite, and belief which were tribal,
when they were brought to their present ethnographic area, cannot be
considered in the varied forms of their survival except by restoration
to the tribal organisation from which they were torn when they began
their life as survivals.

What I have endeavoured to explain in this way are the principles
which should govern folklore research in relation to ethnological
conditions. The differing races which made up the peoples of Europe
before the era of political history must have left their distinctive
remains in folklore, if folklore is rightly considered as the
traditional survivals of the prehistory period. To get at and classify
these remains we must be clear as to the problems which surround
inquiry into them. The solution of these problems will place us in
possession of a mass of survivals in folklore which are naturally
associated with each other, and which stand apart from other survivals
also naturally associated with each other. In these two masses we may
detect the main influences of the great tribal races and the
non-tribal races. We cannot, I think, get much beyond this. We may,
perhaps, here and there, detect smaller race divisions--Celtic,
Teutonic, Scandinavian or other distinctions, according to the area of
investigation--but these will be less apparent, le