accepted, it would be the next step to ask whether hallucinations
representing the dead show any signs of being caused by some action on the
side of the departed. That is a topic on which the little that we have to
say must be said later.

In the meantime the reader who has persevered so far is apt to go no
further. The prejudice against 'wraiths' and 'ghosts' is very strong; but,
then, our innocent phantasms are neither (as we understand their nature)
ghosts nor wraiths. Kant broke the edges of his metaphysical tools
against, not these phantasms, but the logically inconceivable entities
which were at once material and non-material, at once 'spiritual' and
'space-filling.' There is no such difficulty about hallucinations, which,
whatever else may be said about them, are familiar facts of experience.
The only real objections are the statements that hallucinations are
always _morbid_ (which is no longer the universal belief of physiologists
and psychologists), and that the alleged coincidences of a phantasm of a
person with the unknown death of that person at a distance are 'pure
flukes.' That is the question to which we recur later.

In the meantime, the defenders of the theory, that there is some not
understood connection of cause and effect between the death or other
crisis at one end and the perception representing the person affected by
the crisis at the other end, point out that such hallucinations, or other
effects on the percipient, exist in a regular rising scale of potency and
perceptibility. Suppose that 'A's' death in Yorkshire is to affect the
consciousness of 'B' in Surrey before he knows anything about the fact
(suppose it for the sake of argument), then the effect may take place
(1) on 'B's' emotions, producing a vague _malaise_ and gloom; (2) on his
motor nerves, urging him to some act; (3) or may translate itself into his
senses, as a touch felt, a voice heard, a figure seen; or (4) may render