hdraws, death follows, and the
soul, no longer warmly housed in the body, ceases to exist. For the
'film' (ghost) is not the soul, and the soul is not the film,
whereas savage philosophy identifies the soul with the ghost. Even
Lucretius retains the savage conception of the soul as a thing of
rarer matter, a thing partly separable from the body, but that thing
is resolved for ever into its elements on the death of the body.
His imaginary 'film,' on the other hand, may apparently endure for
ages.

The Lucretian theory had, for Lucretius, the advantages of being
physical, and of dealing a blow at the hated doctrine of a future
life. For the public it had the disadvantages of being incapable of
proof, of not explaining the facts, as conceived to exist, and of
being highly ridiculous, as Plutarch observed. Much later
philosophers explained all apparitions as impressions of sense,
recorded on the brain, and so actively revived that they seemed to
have an objective existence. One or two stock cases (Nicolai's, and
Mrs. A.'s), in which people _in a morbid condition_, saw
hallucinations which they knew to be hallucinations, did, and do, a
great deal of duty. Mr. Sully has them, as Hibbert and Brewster
have them, engaged as protagonists. Collective hallucinations, and
the hallucinations of the sane which coincide with the death, or
other crisis in the experience of the person who seemed to be seen,
were set down to imagination, 'expectant attention,' imposture,
mistaken identity, and so forth.

Without dwelling on the causes, physical or psychological, which
have been said by Frazer of Tiree (1707), Ferrier, Hibbert, Scott,
and others, to account for the hallucinations of the sane, for
'ghosts,' Mr. Tylor has ably erected his theory of animism, or the
belief in spirits. Thinking savages, he says, 'were deeply
impressed by two groups of biological phenomena,' by the facts of
living, dying, sleep, trance, waking and disease. They asked:
'What is the difference between a living body and a dead one?' They
wanted to know the causes of sleep, trance and death. They were
also concerned to explain the appearances of dead or absent human
beings in dreams and waking visions. Now it was plain that 'life'
could go away, as it does in death, or seems to do in dreamless
sleep. Again, a phantasm of a living man can go away and appear to
waking or sleeping people at a distance. The conclusion was reached
by savages that the ph