conception.

The most general term for the belief in the various orders of gods thus
becomes the belief in invisible, super-material beings, like, and yet
superior to man. It is for this reason that Professor Tylor's definition
of religion as "the belief in spiritual beings--so long as we do not use
the term "spiritual" in its modern sense"--seems to me the moat
satisfactory definition yet offered. It is the one point on which all
religions agree, and for this reason may be regarded as their essential
feature.

This taken for granted, our next point of enquiry is, What was there in
the conditions of primitive life that would give rise to a belief in
this super-material, or in modern language, spiritual existence? Now
there are at least two sets of experiences that seem adequate to the
required explanation. The one is normal, the other abnormal. The first
is connected directly with the universal experience of dreams. The
savage is, as Tylor says, a severely practical person. He believes what
he sees and, one may add, he sees what he believes. Knowing nothing of
the distinction we draw between a fact and an illusion, ignorant of the
functions, or even the existence of a nervous system, the dreams of a
savage are to him as real as his waking experiences. He does not say "I
dreamed I saw So-So," but like the Biblical characters he says, "I saw
So-So in a dream." The two forms of expression carry all the difference
between fact and fancy. One thing is therefore obvious to the savage
mind--something escapes from the body, travels about, and returns. Such
a conviction does not represent the conclusions of a genius speculating
upon the meaning of unexplained facts. It is a conviction steadily built
up by the pressure of unvarying experience, as steadily as is the
conviction that fire burns or that water is wet. The very universality
of the belief is proof that it had some such sub-conscious origin.

A second class of experiences lead to the same conclusion. In temporary
loss of consciousness the savage again sees proof of the existence of a
double. With epilepsy or insanity there is offered decisive proof that
some spirit has taken possession of the individual's body. Even in
civilised countries this belief was widely held hardly more than a
century ago. And both these classes of experience are enforced by the
belief that the shadow of a man, an echo, a reflection seen in water,
etc., are all real things. The proofs that the