In many cases a Linnaean species has been split up till it became, first, a genus, then a family, and, in some cases, an order. What one naturalist considers a species is considered by another a genus containing several species, and, vice versa, the species of one authority is described as merely a variety by another. The older naturalists might have said with truth: we do not know what the species are, but we are quite certain that whatever they are they have never undergone any change in their distinguishing characters. At the same time we know that whether we call related forms varieties or species or genera in different cases, we find, whatever organisms we study, whether plants or animals, definite types distinguished by special characters of form, colour, and structure, and that individuals of one species or type never give rise by generation to individuals of any other known species or type. We do not find wolves producing foxes, or bulldogs giving birth to greyhounds. As a general rule the distinguishing characters are inherited, and it is by no means easy even in domesticated animals and plants to obtain an exact and complete record of the descent of a new variety from the original form. Among species in a state of nature it is the exception to find two recognised species which can be crossed or hybridised. In the case of the horse and the ass, although mules are the hybrid offspring of the two, the mules themselves are sterile, and there are many similar cases, so that some naturalists have maintained that mutual infertility should be recognised as the test of separation in species. Darwin founded his theory on the assumption that differences of species were differences of adaptation. His theory of natural selection is a theory of the origin of adaptations, and only a theory of the origin of species on the assumption that their distinguishing characters are adaptations to different modes and conditions of life, to different requirements. He pointed out that there is always a considerable range of variation in the specific characters, that, as a rule, no two individuals are exactly alike, even when produced by the same two parents. The central principle of his theory was the survival of individuals possessing those variations which were most useful in the competition of species with species and of individual with individual. He thus explained adaptation to new conditions and divergence of several species from a common ancestor. Characters which were not obviously adaptive were explained either by correlation or by the supposition that they had a utility of which we were ignorant. Darwin also admitted the direct action of conditions as a subordinate factor.