26 NATURE AND LIFE. 
mere experimentalism is that it may sentence men to the 
fixed and stubborn contemplation of this scratch. What 
folly! All the history of the development of the sciences 
proves that important discoveries all proceed from a differ- 
ent feeling, which is that of continuation of forces beyond 
the limits of observation, and of a harmony in relations, 
overruling the singularities and deformities of detached 
experiences. To hedge one’s self within what can be com- 
puted, weighed, and demonstrated, to trust such evidence 
only, and bar one’s self inside the prison of the senses, to 
hush or scorn the suggestions of the spirit, our only true 
light, because it is the spark of the flame that vivifies all— 
this is, deny it or not, the condition and the subject state 
of materialism. Only reason can conceive the fixity, the 
generality, and the universality of relations, and all sa- 
vants admit that the destiny of science is to establish laws 
possessing these three characteristics; but to admit that is 
to confess by implication that partial, incoherent, imper- 
fect, relative details must undergo a refining, a thorough 
conversion in the alembic of the mind, whence they issue, 
with so new an aspect and meaning, that what before 
seemed most important becomes as mere an accessory as it 
is possible to be, and that which looked most ephemeral 
takes its place among eternal things. 
The conception of atoms dates from the highest antiq- 
uity. Leucippus and Democritus, the masters of Epicurus, 
several centuries before the Christian era, taught that mat- 
ter is composed of invisible but indestructible corpuscles, 
the number of which is as boundless as the vastness of the 
space in which they are diffused. These corpuscles are 
solid, endowed with shape and motion. The difference of 
their forms regulates the difference of their movements, 
and consequently of their characteristics. The conception 
of a principle guiding these diversities, that is, of an intel- 
ligence as the supreme cause of differentiation, is not less 
