nonsense,' Herr Parish would
argue, 'you, Jones, saw nothing of the kind, nor did you tell Mr. Lang,
who, I am sorry to find, agrees with you. What happened was _this_: When
the awful news came to-day of your aunt's death, you were naturally,
and even creditably, excited, especially as the poor lady was killed by
being pegged down on an ant-heap. This excitement, rather praiseworthy
than otherwise, made you _believe_ you had seen your aunt, and _believe_
you had told Mr. Lang. He also is a most excitable person, though I
admit he never saw your dear aunt in his life. He, therefore (by virtue of
his excitement), now _believes_ you told him about seeing your unhappy
kinswoman. This kind of false memory is very common. Two cases are
recorded by Kraepelin, among the insane. Surely you quite understand my
reasoning?'

I quite understand it, but I don't see how it comes to seem good logic to
Herr Parish.

The other theory is funnier still. Jones never had an hallucination
before. 'The rarity and the degree of interest compelled by it' made Jones
'connect it with some other prominent event,' say, the death of his aunt,
which, really, occurred, say, nine months afterwards. But this is a mere
case of _evidence_, which it is the affair of the S.P.R. to criticise.

Herr Parish is in the happy position called in American speculative
circles 'a straddle.' If a man has an hallucination when alone, he was in
circumstances conducive to the sleeping state. So the hallucination is
probably a dream. But, if the seer was in company, who all had the same
hallucination, then they all had the same _points de repere_, and the same
adaptive memories. So Herr Parish kills with both barrels.

If anything extraneous could encourage a belief in coincidental and
veridical hallucinations, it would be these 'Oppositions of Science.' If a
learned and fair opponent can find no better proofs than logic and
(unconscious) perversions of facts like the logic and the statements of
Herr Parish, the case for telepathic hallucinations may seem strong
indeed. But we must grant him the existence of the adaptive and mythopoeic
powers of memory, which he asserts, and also illustrates. I grant, too,
that a census of 17,000 inquiries may only have 'skimmed the cream off'
(p. 87). Another dip of the net, bringing up 17,000 fresh answers, might
alter the whole aspect of the case, one way or the other. Moreover, we
cannot get scientific evidence in this way of inquir