essay, to decide whether the parallel supernormal phenomena believed
in by savages, and said to recur in civilisation, are facts of actual
experience, or not.

Now, this question is not otiose. Mr. Tylor, like other anthropologists,
Mr. Huxley, Mr. Herbert Spencer, and their followers and popularisers,
constructs on anthropological grounds, a theory of the Origin of Religion.

That origin anthropology explains as the result of early and fallacious
reasonings on a number of biological and psychological phenomena, both
normal and (as is alleged by savages) supernormal. These reasonings led to
the belief in souls and spirits. Now, first, anthropology has taken for
granted that the Supreme Deities of savages are envisaged by them as
'spirits.' This, paradoxical as the statement may appear, is just what
does not seem to be proved, as we shall show. Next, if the supernormal
phenomena (clairvoyance, thought-transference, phantasms of the dead,
phantasms of the dying, and others) be real matters of experience, the
inferences drawn from them by early savage philosophy may be, in some
degree, erroneous. But the inferences drawn by materialists who reject the
supernormal phenomena will also, perhaps, be, let us say, incomplete.
Religion will have been, in part, developed out of facts, perhaps
inconsistent with materialism in its present dogmatic form. To put it less
trenchantly, and perhaps more accurately, the alleged facts 'are not
merely dramatically strange, they are not merely extraordinary and
striking, but they are "odd" in the sense that they will not easily fit in
with the views which physicists and men of science generally give us of
the universe in which we live' (Mr. A.J. Balfour, President's Address,
'Proceedings,' S.P.R. vol. x. p. 8, 1894).

As this is the case, it might seem to be the business of Anthropology, the