This theory is most cogently presented by Mr. Tylor, and is confirmed by examples chosen from his wide range of reading.  But, among these normal and natural facts, as of sleep, dream, breath, life, dying, Mr. Tylor includes (not as facts, but as examples of applied animistic theory) cases of ‘clairvoyance,’ apparitions of the dying seen by the living at a distance, second sight, ghostly disturbances of knocking and rapping, movements of objects, and so forth.  It is not a question for Mr. Tylor whether clairvoyance ever occurs:  whether ‘death-bed wraiths’ have been seen to an extent not explicable by the laws of chance, whether disturbances and movements of objects not to be accounted for by human agency are matters of universal and often well-attested report.  Into the question of fact, Mr. Tylor explicitly declines to enter; these things only concern him because they have been commonly explained by the ‘animistic hypothesis,’ that is, by the fancied action of spirits.  The animistic hypothesis, again, is the result, naturally fallacious, of savage man’s reasonings on life, death, sleep, dreams, trance, breath, shadow and the other kindred biological phenomena.  Thus clairvoyance (on the animistic hypothesis) is the flight of the conscious ‘spirit’ of a living man across space or time; the ‘deathbed wraith’ is the visible apparition of the newly-emancipated ‘spirit,’ and ‘spirits’ cause the unexplained disturbances and movements of objects.  In fact it is certain that the animistic hypothesis (though a mere fallacy) does colligate a great number of facts very neatly, and has persisted from times of low savagery to the present age of reason.  So here is a case of the savage origin and persistent ‘survival’ of a hypothesis,—­the most potent hypothesis in the history of humanity.

From Mr. Tylor’s point of view, his concern with the subject ceases here, it is not his business to ascertain whether the abnormal facts are facts or fancies.  Yet, to other students, this question is very important.  First, if clairvoyance, wraiths, and the other alleged phenomena, really do occur, or have occurred, then savage man had much better grounds for the animistic hypothesis than if no such phenomena ever existed.  For instance, if a medicine-man not only went into trances, but brought back from these expeditions knowledge otherwise inaccessible, then there were better grounds for believing in a consciousness exerted apart from the body than if there were no evidence but that of non-veridical dreams.  If merely the dream-coincidences which the laws of chance permit were observed, the belief in the soul’s dream-flight would win less favourable and general acceptance than it would if clairvoyance, ’the sleep of the shadow,’ were a real if rare experience.  The very name given by the Eskimos to the hypnotic state, ‘the sleep of the shadow,’ proves that savages do make distinctions between normal and abnormal conditions of slumber.