the higher anthropoids. Doubtless it is
hard to get anything like scientific evidence out of people so
uncultivated, whose language and modes of conception are so alien to our
own. Individual travellers, moreover, have been the victims of their own
credulity, stupidity, self-conceit, and prejudice. "But the best
testimony of the truth of the reports as to the actual belief in the
facts, is the undesigned coincidence of the evidence from all quarters.
When the stories brought by travellers, ancient and modern, learned and
unlearned, pious or sceptical, agree in the main, we have all the
certainty that anthropology can offer."

From this ever-growing mass of evidence, it would appear that the
universal belief among savages in a spirit-world is mainly strengthened
and sustained, not by the phenomena of dreaming but by what Mr. Spencer
would call "alleged" supernormal manifestations, such as those of
clairvoyance, crystal-gazing, apparitions, miracles, prophecies,
possession, and the like. For belief in such marvels exists beyond
doubt, and furnishes a very obvious and logical basis for the further
belief in the invisible causes of these visible effects; nor should we
have recourse to an hypothetical and more indirect explanation of belief
in a spirit-world when an actual and direct explanation is at hand. If
we see the branch growing out of the tree, we need not inquire what
trunk it sprang from, unless we have strong evidence that it is only a
graft. All investigation tends to show that savages believe in spirits
and in the spirit-world because they witness, or firmly believe they
witness, supernormal phenomena.

Besides this, it must be allowed that together with the _normal_
phenomena of dreaming, there are abnormal dreams which even to
cultivated minds seem at times as supernormal as second-sight or
prophecy. But it is not on supernormal, but on normal dreams that
animists base their explanation. We need not deny that dreams and
delirium may have given palpable shape to the conception of a ghost, and
may also have helped forward the notion of a spirit by furnishing
something intermediary between the grossness of our waking
sense-experiences, and the altogether elusive and difficult thought of
unembodied will and intelligence independent of space and time.

In the main then it seems more plausible to maintain that the idea of
unembodied or disembodied spirits was shaped by that instinctive law of
our mind which makes