have been due to evolution only, or, as the single other alternative, to an immense series of special creations, made not only when life began, but also during its entire continuance, in exact imitation of a process of evolution—a preposterous theory, for which there is no more warrant than for the hypothesis that our various breeds of dogs have not descended from a common ancestry, but were separately created.
The third great mass of evidence is that furnished by experiment. It is therefore the most conclusive of all, since any theory which can be proved experimentally may be said to be proved decisively. For instance, if we have a theory that the boiling-point of water varies with the atmospheric pressure, and find on experiment that it does so vary, we may conclude that we have proved our theory beyond cavil. As a disputed point is involved, which it will be advantageous to discuss thus early, we will consider this part of the evidence more at length than we did the rest. The theory of evolution may be put to the test of experiment. We may take any species of plant and animal, and by copying the process of natural selection, by destroying those individuals of the species which we regard as unfit,—i.e. those which vary unfavourably as regards any peculiarity we wish to develop,—and by allowing only those individuals who vary favourably as regards that peculiarity to continue the race, we may develop the chosen peculiarity in the chosen species to almost any extent; the one condition being, that our operations shall extend over a sufficient number of generations. The efforts of man to improve his cultivated plants and domesticated animals may be described as a lengthened series of such experiments, conducted on a gigantic scale, with the result that he has been able to vary plant and animal structures almost at will, and so greatly that, in the absence of scientific and historic testimony, no one
