ears no relation to exactness of reasoning and
firmness of method.

Neglecting species and varieties, we may reduce the fundamental
characters of the scientific imagination to the following:

For its material, it has concepts, the degree of abstraction of which
varies with the nature of the science.

It employs only those associational forms that have an objective basis,
although its mission is to form new combinations, "the discoveries
consisting of the relation of ideas, capable of being united, which
hitherto have been isolated."[117] (Laplace.) All association with an
affective basis is strictly excluded.

It aims toward objectivity: in its conjectural construction it attempts
to reproduce the order and connection of things. Whence its natural
affinity for realistic art, which is midway between fiction and reality.

It is unifying, and so just the opposite of the esthetic imagination,
which is rather developmental. It puts forward the master idea (Claude
Bernard's _idee directrice_), a center of attraction and impulse that
enlivens the entire work. The principle of unity, without which no
creation succeeds, is nowhere more visible than in the scientific
imagination. Even when illusory, it is useful. Pasteur, scrupulous
scientist that he was, did not hesitate to say: "The experimenter's
illusions are a part of his power: they are the preconceived ideas
serving as guides for him."

V

It does not seem to me wrong to regard the imagination of the
metaphysician as a variety of the scientific imagination. Both arise
from one and the same requirement. Several times before this we have
emphasized this point--that the various forms of imagination are not the
work of an alleged "creative instinct," but that each particular one has
arisen from a special need. The scientific imagination has for its prime
motive the need of _partial_ knowledge or explanation; the metaphysical
imagination has for its prime motive the need of a _total_ or complete
explanation. The latter is no longer an endeavor on a restricted group
of phenomena, but a conjecture as to the totality of things, as
aspiration toward completely unified knowledge, a need of final
explanation that, for certain minds, is just as imperious as any other
need.

This necessity is expressed by the creation of a cosmic or human
hypothesis constructed after the type and methods of scientific
hypotheses, but radically subjective in its origin--only apparently
obje