ime on a chain of snow-peaks, he saw an apparition of an
absent friend, looking like a corpse. He goes on to explain that this
phantom was the product of an image of recollection which somehow
managed to combine itself with the (positive) after-image left by the
impression of the snow-surface.

[60] For an account of Mr. Galton's researches, see _Mind_, No. xix.
Compare, however, Professor Bain's judicious observations on these
results in the next number of _Mind_. The liability of children to take
images for percepts, is illustrated by the experiences related in a
curious little work, _Visions_, by E.H. Clarke, M.D. (Boston, U.S.,
1878), pp. 17, 46, and 212.

[61] A common way of describing the relation of the hallucinatory to
real objects, is to say that the former appear partly to cover and hide
the latter.

[62] Griesinger remarks that the forms of the hallucinations of the
insane rarely depend on sense-disturbances alone. Though these are often
the starting-point, it is the whole mental complexion of the time which
gives the direction to the imagination. The common experience of seeing
rats and mice running about during a fit of _delirium tremens_ very well
illustrates the co-operation of peripheral impressions not usually
attended to, and possibly magnified by the morbid state of sensibility
of the time (in this case flying spots, _muscae volitantes_), with
emotional conditions. (See Griesinger, _loc. cit._, p. 96.)

[63] Wundt (_Physiologische Psychologie_, p. 652) tells us of an insane
woodman who saw logs of wood on all hands in front of the real objects.

[64] It is stated by Baillarger (Memoires de l'Academie Royale de
Medicine, tom. xii. p. 273, etc.) that while visual hallucinations are
more frequent than auditory in healthy life, the reverse relation holds
in disease. At the same time, Griesinger remarks (_loc. cit._, p. 98)
that visual hallucinations are rather more common than auditory in
disease also. This is what we should expect from the number of
subjective sensations connected with the peripheral organ of vision. The
greater relative frequency of auditory hallucinations in disease, if
made out, would seem to depend on the close connection between
articulate sounds and the higher centres of intelligence, which centres
are naturally the first to be thrown out of working order. It is
possible, moreover, that auditory hallucinations are quite as common as
visual in states of comparative health, tho