Often the word "mongrel" is limited in its use to signify an undesired or undesirable result of the cross-breeding of individuals of established races. But this is not quite fair to mongrels in general, since, as we have seen, the name really refers only to the fact they are crosses between two breeds. When they happen to suit some artificial and arbitrary requirement they are favoured, and made the starting-point of a new breed, and kept pure in their own line; but when they do not fit some capricious demand of the breeder they are sneered at and condemned, although they may be fine and capable animals. No doubt some mongrels between races differing greatly from one another, or having some peculiar mixture of incompatible qualities the exact nature of which we have not ascertained, are wanting in vigour, and cannot be readily established as a new breed. In nature the success of the mongrel depends on whether or not its mixture of qualities makes it fitter than others to the actual conditions of its life, and able to survive in the competition for food and place. In man's breeding operations with varieties of domesticated animals and "cultivated" plants, the survival of the mongrel depends upon its fitting some arbitrary standard applied by man, who destroys those which do not suit his fancy, and selects for survival and continued breeding those which do. What is called "miscegenation," or the inter-breeding of human races, must be looked at from both these points of view. We require to know how far, if at all, the mixed or mongrel offspring of a human race A with a human race B is really inferior to either of the original stocks A and B, judged by general capacity and life-preserving qualities in the varied conditions of the great area of the habitable globe. And how far an arbitrary or fanciful standard is set up by human races, similar to that set up by the "fancier" or cultivator of breeds of domestic animals. The matter is complicated by the fact that what we loosely speak of as "races" of man are of very various degrees of consanguinity or nearness to one another in blood, that is, in stock or in ultimate ancestry. It is also complicated by the fact that we cannot place any reliance upon the antipathies or preferences shown by the general sentiment of a race in this (or other matters) as necessarily indicating what is beneficial for humanity in general or for the immediate future of any section of it. Nor have we any assurance that what is called "sexual selection"–the preference or taste in the matter of choosing a mate–is among human beings necessarily anything of greater importance–so far as the prosperity of a race or of humanity in general is concerned–than a mere caprice or a meaningless persistence of the human mind in favouring a choice which is habitual and traditional. I have referred to this point again in the last paragraph of this chapter.