grave man of science, go on telling distorted legends out of my own
head, while the facts are plain in print before me,' Herr Parish
may reason, 'how much more are the popular tales about coincidental
hallucinations likely to be distorted?' It is really a very strong
argument, but not exactly the argument which Herr Parish conceives
himself to be presenting.[15]

This unlucky inexactitude is chronic, as we have shown, in Herr Parish's
work, and is probably to be explained by inattention to facts, by
'expectation' of suitable facts, and by 'anxiety' to prove a theory. He
explains the similar or identical reports of witnesses to a collective
hallucination by 'the case with which such appearances adapt themselves
in recollection' (p. 313), especially, of course, after lapse of time. And
then he unconsciously illustrates his case by the case with which
printed facts under his very eyes adapt themselves, quite erroneously, to
his own memory and personal bias as he copies them on to his paper.

Finally he argues that even if collective hallucinations are also 'with
comparative frequency' coincidental, that is to be explained thus:
'The rarity and the degree of interest compelled by it' (by such an
hallucination) 'will naturally tend to connect itself with some other
prominent event; and, conversely, the occurrence of such an event as the
death or mortal danger of a friend is most calculated to produce memory
illusions of this kind.'

In the second case, the excitement caused by the death of a friend is
likely, it seems, to make two or more sane people say, and _believe,_
that they saw him somewhere else, when he was really dying. The only
evidence for this fact is that such illusions occasionally occur, _not_
collectively, in some lunatic asylums. 'It is not, however, a form of
mnemonic error often observed among the insane.' 'Kraepelin gives two
cases.' 'The process occurs sporadically in certain sane people, under
certain exciting conditions.' No examples are given! What is rare as an
_individual_ folly among lunatics, is supposed by Herr Parish to explain
the theoretically 'false memory' whereby sane people persuade themselves
that they had an hallucination, and persuade others that they were told
of it, when no such thing occurred.

To return to our old example. Jones tells me that he has just seen his
aunt, whom he knows to be in Timbuctoo. News comes that the lady died when
Jones beheld her in his smoking-room. 'Oh,