Although these and similar facts indicate strongly that isolation favours the evolution of new species, it would be erroneous to imagine that every isolation of a species-colony conditions its transmutation to a new species, or, as has been maintained, first by Moritz Wagner, and later by Gulick and by Dixon, that isolation is a necessary preliminary to the variation of species--that not selection but isolation alone renders the transmutation of a species possible, and thus admits of its segregation into several different groups of forms. Romanes went so far as to regard the natural selection of Darwin and Wallace as a sub-species of isolation, and isolation in its diverse forms he regarded as the sole factor in the formation of species. He assumed that it was only by the segregation of individuals which did not vary that the constant reversion to the ancestral species could be prevented, and he regarded the process of selection as essentially resulting in the 'isolation' of the fittest through the elimination of the less-fit. The idea is correct in so far, that selection undoubtedly aids the favourable variation to conquest over the old forms, precisely because the latter, being less favourably placed in the struggle for existence, are gradually more completely overcome and weeded out, so that a constant mingling of the new forms with the old is prevented, just as it is by isolation of locality. Obviously the new and fitter forms could not become dominant, could not even become permanent, if they were always being mingled again with the old. But whether it serves any useful purpose to bring this under the category of 'isolation,' and to say that mingling with the ancestral form during transmutation is prevented by natural selection, in that favourably varying individuals _are isolated_ by their superiority from the inferior ones, that is, the non-varying individuals which are doomed to elimination, is somewhat doubtful. For my part, I should prefer to retain the original meaning of the word, and to call 'isolation' the separation of a species-colony by spatial barriers. Whether this factor by itself prevents the mingling with the ancestral form as effectually as selection does, and whether isolation alone and by itself can lead to the evolution of new forms, or perhaps must lead to them, must now be investigated.