ous work of the mind in the course of biological evolution;
they too sum up and give concrete form and expression to a system
of enlightening theories. But that is not all. The most elementary
psychology shows us the amount of thought, in the correct sense of the
term, recollection, or inference, which enters into what we should be
tempted to call pure perception.

Establishment of fact is not the simple reception of the faithful
imprint of that fact; it is invariably interpreted, systematised, and
placed in pre-existing forms which constitute veritable theoretical
frames. That is why the child has to learn to perceive. There is an
education of the senses which he acquires by long training. One day,
which aid of habit, he will almost cease to see things: a few lines, a
few glimpses, a few simple signs noted in a brief passing glance, will
enable him to recognise them; and he will hardly retain any more of
reality than its schemes and symbols.

"Perception," says Mr Bergson on this subject, "becomes in the end only
an opportunity of recollection." ("Matter and Memory", page 59.)

All concrete perception, it is true, is directed less upon the
present than the past. The part of pure perception in it is small, and
immediately covered and almost buried by the contribution of memory.

This infinitesimal part acts as a bait. It is a summons to recollection,
challenging us to extract from our previous experience, and construct
with our acquired wealth a system of images which permits us to read the
experience of the moment.

With our scheme of interpretation thus constituted we encounter the few
fugitive traits which we have actually perceived. If the theory we have
elaborated adapts itself, and succeeds in accounting for, connecting,
and making sense of these traits, we shall finally have a perception
properly so called.

Perception then, in the usual sense of the word, is the resolution of a
problem, the verification of a theory.

Thus are explained "errors of the senses," which are in reality errors
of interpretation. Thus too, and in the same manner, we have the
explanation of dreams.

Let us take a simple example. When you read a book, do you spell each
syllable, one by one, to group the syllables afterwards into words, and
the words into phrases, thus travelling from print to meaning? Not at
all: you grasp a few letters accurately, a few downstrokes in their
graphical outline; then you guess the remainder, travel