of change from which it is a divergence, and is
therefore excluded where there is no habitual course of change. In the
absence of that cyclical series of metamorphoses which even the simplest
living thing now shows us, as a result of its inherited constitution,
there could be no _point d'appui_ for natural selection. How, then, did
organic evolution begin?

If a primitive mass of organic matter was like a mass of inorganic
matter in respect of its passivity, and differed only in respect of its
greater changeableness; then we must infer that its first changes
conformed to the same general law as do the changes of an inorganic
mass. The instability of the homogeneous is a universal principle. In
all cases the homogeneous tends to pass into the heterogeneous, and the
less heterogeneous into the more heterogeneous. In the primordial units
of protoplasm, then, the step with which evolution commenced must have
been the passage from a state of complete likeness throughout the mass
to a state in which there existed some unlikeness. Further, the cause of
this step in one of these portions of organic matter, as in any portion
of inorganic matter, must have been the different exposure of its parts
to incident forces. What incident forces? Those of its medium or
environment. Which were the parts thus differently exposed? Necessarily
the outside and the inside. Inevitably, then, alike in the organic
aggregate and the inorganic aggregate (supposing it to have coherence
enough to maintain constant relative positions among its parts), the
first fall from homogeneity to heterogeneity must always have been the
differentiation of the external surface from the internal contents. No
matter whether the modification was physical or chemical, one of
composition or of decomposition, it comes within the same
generalization. The direct action of the medium was the primordial
factor of organic evolution.

* * * * *

And now, finally, let us look at the factors in their _ensemble_, and
consider the respective parts they play: observing, especially, the ways
in which, at successive stages, they severally give place one to another
in degree of importance.

Acting alone, the primordial factor must have initiated the primary
differentiation in all units of protoplasm alike. I say alike, but I
must forthwith qualify the word. For since surrounding influences,
physical and chemical, could not be absolutely the same in