In order at once to restate the theory of physiological selection, and to do so in a form which cannot be suspected of being in any way influenced by Mr. Wallace's more recent criticisms, I will begin by reproducing the main features of the theory in the words which were employed for this purpose more than three years ago, when I supplied an article to the _Nineteenth Century_ in answer to one by him in the _Fortnightly Review_. Moreover, for the most part this restatement of the theory is quoted _verbatim_ from my original paper—the differences being due only to the conditions imposed by limits of an article. The following, then, is quoted from the _Nineteenth Century_ for January, 1887: "According to the Darwinian theory [which, as elsewhere fully explained, the present theory is in no way capable of supplanting, but only of supplementing, and this among other ways, by explaining why it is that some degree of mutual infertility is so general a phenomenon as between allied species—a phenomenon which Darwin expressly regarded as not explicable by the theory of natural selection], it is for the most part only those variations which happen to have been useful that have been preserved: yet, even as thus limited, the principle of variability is held able to furnish sufficient material out of which to construct the whole adaptive morphology of nature. How immense, therefore, must be the number of unuseful variations. Yet these are all, for the most part, still-born, or allowed to die out immediately by intercrossing. Should such intercrossing be prevented, however, there is no reason why unuseful variations should not be perpetuated by heredity quite as well as useful ones when under the nursing influence of natural selection—as, indeed, we see to be the case in our domesticated productions. Consequently, if from any reason a section of a species is prevented from intercrossing with the rest of its species, new varieties of a trivial or unuseful kind might be expected to arise within that section. And this is just what we find. Oceanic islands, for example, are well known to be extraordinarily rich in peculiar species; and this can best be explained by considering that a complete separation of the fauna and flora on such an island permits them to develop varietal histories of their own, without interference by intercrossing with their originally parent forms. We see the same principle exemplified by the influence of geographical barriers of any kind, and also by the consequences of migration. Therefore, given an absence of overwhelming intercrossing, and the principle of what I term independent variability may be trusted to evoke new species, without the aid of natural selection. [Homogamy.]