a thing it has never seen, we
say, just because the innate knowledge is in this case of a definite
object, that it belongs to _instinct_ and not to _intelligence_.
Intelligence does not then imply the innate knowledge of any object. And
yet, if intelligence knows nothing by nature, it has nothing innate.
What, then, if it be ignorant of all things, can it know? Besides
_things_, there are _relations_. The new-born child, so far as
intelligent, knows neither definite objects nor a definite property of
any object; but when, a little later on, he will hear an epithet being
applied to a substantive, he will immediately understand what it means.
The relation of attribute to subject is therefore seized by him
naturally, and the same might be said of the general relation expressed
by the verb, a relation so immediately conceived by the mind that
language can leave it to be understood, as is instanced in rudimentary
languages which have no verb. Intelligence, therefore, naturally makes
use of relations of like with like, of content to container, of cause to
effect, etc., which are implied in every phrase in which there is a
subject, an attribute and a verb, expressed or understood. May one say
that it has _innate_ knowledge of each of these relations in particular?
It is for logicians to discover whether they are so many irreducible
relations, or whether they can be resolved into relations still more
general. But, in whatever way we make the analysis of thought, we always
end with one or several general categories, of which the mind possesses
innate knowledge since it makes a natural use of them. Let us say,
therefore, that _whatever, in instinct and intelligence, is innate
knowledge, bears in the first case on_ things _and in the second on_
relations.

Philosophers distinguish between the matter of our knowledge and its
form. The matter is what is given by the perceptive faculties taken in
the elementary state. The form is the totality of the relations set up
between these materials in order to constitute a systematic knowledge.
Can the form, without matter, be an object of knowledge? Yes, without
doubt, provided that this knowledge is not like a thing we possess so
much as like a habit we have contracted,--a direction rather than a
state: it is, if we will, a certain natural bent of attention. The
schoolboy, who knows that the master is going to dictate a fraction to
him, draws a line before he knows what numerator and what d