stance as well as of direction is
clearly obtainable by means of this organ. There seems to me no reason
why an animal endowed with fine olfactory sensibility, and capable of an
analytic separation of sense elements, should not gain a rough
perception of an external order much more complete than our auditory
perception, which is necessarily so fragmentary. This supposition
appears, indeed, to be the necessary complement to the idea first
broached, so far as I am aware, by Professor Croom Robertson, that to
such animals, visual perception consists in a reference to a system of
muscular feelings defined and bounded by strong olfactory sensations,
rather than by tactual sensations as in our case.

[15] It may be said, perhaps, that the exceptional direction of
attention, by giving an unusual intensity to the impression, causes us
to exaggerate it just as in the case of a novel sensation. An effort of
attention directed to any of our vague bodily sensations easily leads us
to magnify its cause. A similar confusion may arise even in direct
vision, when the objects are looked at in a dim light, through a want of
proper accommodation. (See Sir D. Brewster, _op. cit._, letter i)

[16] They might also be distinguished as objective and subjective
illusions, or as illusions _a posteriori_ and illusions _a priori_.

[17] _Die Schein-Bewegungen_, von Professor Dr. J.I. Hoppe (1879); _cf._
an ingenious article on "Optical Illusions of Motion," by Professor
Silvanus P. Thompson, in _Brain_, October, 1880. These illusions
frequently involve the co-operation of some preconception or
expectation. For example, the apparent movement of a train when we are
watching it and expecting it to move, involves both an element of
sense-impression and of imagination. It is possible that the illusion of
table-turning rests on the same basis, the table-turner being unaware of
the fact of exerting a certain amount of muscular force, and vividly
expecting a movement of the object.

[18] _Physiologische Optik_, p. 316.

[19] It is plain that this supposed error could only be brought under
our definition of illusion by extending the latter, so as to include
sense-perceptions which are contradicted by reason employing idealized
elements of sense-impression, which, as Lewes has shown (_Problems of
Life and Mind_, i. p. 260), make up the "extra-sensible world" of
science.

[20] An ingenious writer, M. Binet, has tried to prove that the fusion
of homoge