ctive. _It is a rationalized myth._

The three moments requisite for the constitution of a science are found
here, but in a modified form: reflection replaces observation, the
choice of the hypothesis becomes all-important, and its application to
everything corresponds to scientific proof.

(1) The first moment or preparatory stage, does not belong to our
subject. It requires, however, a word in passing. In all science,
whether well or ill established, firm or weak, we start from facts
derived from observation or experiment. Here, facts are replaced by
general ideas. The terminus of every science is, then, the
starting-point of philosophical speculation:--metaphysics begins where
each separate science ends; and the limits of the latter are theories,
hypotheses. These hypotheses become working material for metaphysics
which, consequently, is an hypothesis built on hypotheses, a conjecture
grafted on conjecture, a work of imagination superimposed on works of
imagination. Its principal source, then, is imagination, to which
reflection applies itself.

Metaphysicians, indeed, hold that the object of their researches, far
from being symbolic and abstract, as in science, or fictitious and
imaginary, as in art, is the very essence of things,--absolute reality.
Unfortunately, they have never proven that it suffices to seek in order
to find, and to wish in order to get.

(2) The second stage is critical. It is concerned with finding the
principle that rules and explains everything. In the invention of his
theory the metaphysician gives his measure, and permits us to value his
imaginative power. But the hypothesis, which in science is always
provisional and revocable, is here the supreme reality, the fixed
position, the _inconcussum quid_.

The choice of the principle depends on several causes: The chief of
these is the creator's individuality. Every metaphysician has a point of
view, a personal way of contemplating and interpreting the totality of
things, a belief that tends to recruit adherents.

Secondary causes are: the influence of earlier systems, the sum of
acquired knowledge, the social _milieu_, the variable predominance of
religions, sciences, morality, esthetic culture.

Without troubling ourselves with classifications, otherwise very
numerous, into which we may group systems (idealism, materialism,
monism, etc.) we shall, for our purpose, divide metaphysicians into the
imaginative and rational, according as th