artistic suggestion
depends on this. A clever draughtsman can indicate a face by a few rough
touches, and this is due to the fact that the spectator's mind is so
familiarized, through recurring experience and special interest, with
the object, that it is ready to construct the requisite mental image at
the slightest external suggestion. And hence the risk of hasty and
illusory interpretation.

These observations naturally conduct us to the consideration of the
second great group of sense-illusions, which I have marked off as active
illusions, where the action of a pre-existing intellectual disposition
becomes much more clearly marked, and assumes the form of a free
imaginative transformation of reality.

CHAPTER VI.

ILLUSIONS OF PERCEPTION--_continued_.

B. _Active Illusions._

When giving an account of the mechanism of perception, I spoke of an
independent action of the imagination which tends to anticipate the
process of suggestion from without. Thus, when expecting a particular
friend, I recognize his form much more readily than when my mind has not
been preoccupied with his image.

A little consideration will show that this process must be highly
favourable to illusion. To begin with, even if the preperception be
correct, that is to say, if it answer to the perception, the mere fact
of vivid expectation will affect the exact moment of the completed act
of perception. And recent experiment shows that in certain cases such a
previous activity of expectant attention may even lead to the illusory
belief that the perception takes place before it actually does.[47]

A more palpable source of error resides in the risk of the formation of
an inappropriate preperception. If a wrong mental image happens to have
been formed and vividly entertained, and if the actual impression fits
in to a certain extent with this independently formed preperception, we
may have a fusion of the two which exactly simulates the form of a
complete percept. Thus, for example, in the case just supposed, if
another person, bearing some resemblance to our expected friend, chances
to come into view, we may probably stumble into the error of taking one
person for another.

On the physical side, we may, agreeably to the hypothesis mentioned
above, express this result by saying that, owing to a partial identity
in the nervous processes involved in the anticipatory image and the
impression, the two tend to run one into the other, consti