tions from truth
in pictorial representation already touched on, the amount of essential
agreement is so large and so powerful in its effect that even an
intelligent animal will experience an illusion. Mr. Romanes sends me an
interesting account of a dog, that had never been accustomed to
pictures, having been put into a state of great excitement by the
introduction of a portrait into a room, on a level with his eye. It is
not at all improbable that the lower animals, even when sane, are
frequently the subjects of slight illusion. That animals dream is a fact
which is observed as long ago as the age of Lucretius.

[53] This kind of illusion is probably facilitated by the fact that the
eye is often performing slight movements without any clear consciousness
of them. See what was said about the limits of sensibility, p. 50.

[54] _Mental Physiology_, fourth edit., p. 158.

[55] In persons of very lively imagination the mere representation of an
object or event may suffice to bring about such a semblance of
sensation. Thus, M. Taine (_op. cit._, vol. i. p. 94) vouches for the
assertion that "one of the most exact and lucid of modern novelists,"
when working out in his imagination the poisoning of one of his
fictitious characters, had so vivid a gustatory sensation of arsenic
that he was attacked by a violent fit of indigestion.

[56] Mentioned by Dr. Carpenter (_Mental Physiology_, p. 207), where
other curious examples are to be found.

[57] See _Annales Medico-Psychologiques_, tom. vi. p. 168, etc.; tom.
vii. p. 1. etc.

[58] I have already touched on the resonance of a sense-impression when
the stimulus has ceased to act (see p. 55). The remarks in the text hold
good of all such after-impressions, in so far as they take the form of
fully developed percepts. A good example is the recurrence of the images
of microscopic preparations, to which the anatomist is liable. (See
Lewes, _Problems of Life and Mind_, third series, vol. ii. p. 299.)
Since a complete hallucination is supposed to involve the peripheral
regions of the nerve, the mere fact of shutting the eye would not, it is
clear, serve as a test of the origin of the illusion.

[59] That subjective sensation may become the starting-point in complete
hallucination is shown in a curious instance given by Lazarus, and
quoted by Taine, _op. cit._, vol. i. p. 122, _et seq._ The German
psychologist relates that, on one occasion in Switzerland, after gazing
for some t