believed to be necessary, with, perhaps, the
exception of the sportive or conjuring class. A spirit helps to
cure and helps to kill. The free spirit of the clairvoyant in
bondage meets other spirits in its wanderings. Anthropologists,
taking it for granted that 'spirits' are a mere 'animistic
hypothesis'--their appearances being counterfeited by imposture--
have paid little attention to the practical magic of savages, as far
as it is not merely sympathetic, and based on the doctrine that
'like cures like'.

Thus Mr. Sproat, in his excellent work, Scenes and Studies of Savage
Life, frankly admits that in Vancouver Island the trickery and
hocus-pocus of Aht sorcery were so repugnant to him that he could
not occupy himself with the topic. Some other travellers have been
more inquisitive; unlettered sojourners among the wilder peoples
have shared their superstitions, and consulted their oracles, while
one or two of the old Jesuit missionaries were close and puzzled
observers of their 'mediumship'.

Thus enough is known to show that savage spiritualism wonderfully
resembles, even in minute details, that of modern mediums and
seances, while both have the most striking parallels in the old
classical thaumaturgy.

This uniformity, to a certain extent, is not surprising, for savage,
classical, and modern spiritualism all repose on the primaeval
animistic hypothesis as their metaphysical foundation. The origin
of this hypothesis--namely, that disembodied intelligences exist and
are active--is explained by anthropologists as the result of early
reasonings on life, death, sleep, dreams, trances, shadows, the
phenomena of epilepsy, and the illusions of starvation. This
scientific theory is, in itself, unimpeachable; normal phenomena,
psychological and physical, might suggest most of the animistic
beliefs. {35}

At the same time 'veridical hallucinations,' if there are any, and
clairvoyance, if there is such a thing, would do much to originate
and confirm the animistic opinions. Meanwhile, the extraordinary
similarity of savage and classical spiritualistic rites, with the
corresponding similarity of alleged modern phenomena, raises
problems which it is more easy to state than to solve. For example,
such occurrences as 'rappings,' as the movement of untouched
objects, as the lights of the seance room, are all easily feigned.
But that ignorant modern knaves should feign precisely the same
raps, lights, and movements as