§ 100. One more preliminary is needful before proceeding. We have still to study the conditions under which alone, Evolution can take place.

The process to be interpreted is, as already said, a certain change in the arrangement of parts. That increase of heterogeneity commonly displayed throughout Evolution, is not an increase in the number of kinds of ultimate or undecomposable units which an aggregate contains; but it is a change in the distribution of such units. If it be assumed that what we call chemical elements, are absolutely simple (which is, however, an hypothesis having no better warrant than the opposite one); then it must be admitted that in respect to the number of kinds of matter contained in it, the Earth is not more heterogeneous at present than it was at first—that in this respect, it would be as heterogeneous were all its undecomposable parts uniformly mixed, as it is now, when they are arranged and combined in countless different ways. But the increase of heterogeneity with which we have to deal, and of which alone our senses can take cognizance, is that produced by the passage from unity of distribution to variety of distribution.