der Erkenntniss die in
uns _a priori_ angetroffen werden."--_Kritik der reinen Vernunft.
Elementarlehre_, p. 135.

Without a glossary explanatory of Kant's terminology, this passage would
be hardly intelligible in a translation; but it may be paraphrased thus:
All knowledge is founded upon experiences of sensation, but it is not
all derived from those experiences; inasmuch as the impressions of
relation ("reine Anschauungen"; "reine Verstandesbegriffe") have a
potential or _a priori_ existence in us, and by their addition to
sense-experiences, constitute knowledge.

CHAPTER IV.

THE CLASSIFICATION AND THE NOMENCLATURE OF MENTAL OPERATIONS.

If, as has been set forth in the preceding chapter, all mental states
are effects of physical causes, it follows that what are called mental
faculties and operations are, properly speaking, cerebral functions,
allotted to definite, though not yet precisely assignable, parts of the
brain.

These functions appear to be reducible to three groups, namely:
Sensation, Correlation, and Ideation.

The organs of the functions of sensation and correlation are those
portions of the cerebral substance, the molecular changes of which give
rise to impressions of sensation and impressions of relation.

The changes in the nervous matter which bring about the effects which we
call its functions, follow upon some kind of stimulus, and rapidly
reaching their maximum, as rapidly die away. The effect of the
irritation of a nerve-fibre on the cerebral substance with which it is
connected may be compared to the pulling of a long bell-wire. The
impulse takes a little time to reach the bell; the bell rings and then
becomes quiescent, until another pull is given. So, in the brain, every
sensation is the ring of a cerebral particle, the effect of a momentary
impulse sent along a nerve-fibre.

If there were a complete likeness between the two terms of this very
rough and ready comparison, it is obvious that there could be no such
thing as memory. A bell records no audible sign of having been rung five
minutes ago, and the activity of a sensigenous cerebral particle might
similarly leave no trace. Under these circumstances, again, it would
seem that the only impressions of relation which could arise would be
those of co-existence and of similarity. For succession implies memory
of an antecedent state.[24]

But the special peculiarity of the cerebral apparatus is, that any given
function which