passing from the word to the idea, and of being led to believe that,
because there is a word, there is a real thing designated by this
word.

Let us take, for example, the word _electricity_. If we understand by
this term the common law which embraces a certain category of
phenomena, it expresses a clear and useful idea; but as for its
existence, it is not permitted to believe _a priori_ that there is a
distinct agent called electricity which is the efficient cause of the
phenomena. We ought never, says the old rule of philosophy, to admit
entities without an absolute necessity. The march of science has
always consisted in gradually eliminating these provisory conceptions
and in reducing the number of causes. This fact is visible without
going back to the ages of ignorance, when every new phenomenon brought
with it the conception of a special being which caused it and directed
it. In later ages they had _spirits_ in which there was everything:
volatile liquids, gases, and theoretical conceptions, such as
phlogiston. At the end of the last century, and at the beginning of
our own, ideas being more rational, the notion of the "fluid" had been
admitted, a mysterious and still vague enough category (but yet an
already somewhat definite one) in which were ranged the unknown and
ungraspable causes of caloric, luminous, electric, etc., phenomena.
Gradually, the "fluid" has vanished, and we are left (or rather, we
were a short time ago) at the notion of forces--a precise and
mathematically graspable notion, but yet an essentially mysterious
one. We see this conception gradually disappearing to leave finally
only the elementary ideas of matter and motion--ideas, perhaps, which
are not much clearer philosophically than the others, particularly
that of matter taken _per se_, but which, at least, are necessary,
since all the others supposed them.