It is the opinion of many naturalists who are well entitled to have an opinion upon the subject, that, in the words of Mr. Dixon, "Isolation can preserve a non-beneficial as effectually as natural selection can preserve a beneficial variation[118]." The ground on which this doctrine rests is thus clearly set forth by Mr. Gulick:--"The fundamental cause of this seems to lie in the fact that no two portions of a species possess exactly the same average characters; and, therefore, that the initial differences are for ever reacting on the environment and on each other in such a way as to ensure increasing divergence in each generation, as long as the individuals of the two groups are kept from intergenerating[119]." In other words, as soon as a portion of a species is separated from the rest of that species, so that breeding between the two portions is no longer possible, the general average of characters in the separated portion not being in all respects precisely the same as it is in the other portion, the result of in-breeding among all individuals of the separated portion will eventually be different from that which obtains in the other portion; so that, after a number of generations, the separated portion may become a distinct species from the effect of isolation alone. Even without the aid of isolation, any original difference of average characters may become, as it were, magnified in successive generations, provided that the divergence is not harmful to the individuals presenting it, and that it occurs in a sufficient proportional number of individuals not to be immediately swamped by intercrossing. For, as Mr. Murphy has pointed out, in accordance with Delbœuf's law, "if, in any species, a number of individuals, bearing a ratio not infinitely small to the entire number of births, are in every generation born with a particular variation which is neither beneficial nor injurious, and if it be not counteracted by reversion, then the proportion of the new variety to the original form will increase till it approaches indefinitely near to equality[120]." Now even Mr. Wallace himself allows that this must be the case; and thinks that in these considerations we may find an explanation of the existence of certain definite varieties, such as the melanic form of the jaguar, the brindled or ring-eyed guillemot, &c. But, on the other hand, he thinks that such varieties must always be unstable, and continually produced in varying proportions from the parent forms. We need not, however, wait to dispute this arbitrary assumption, because we can see that it fails, even as an assumption, in all cases where the superadded influence of isolation is concerned. Here there is nothing to intercept the original tendency to divergent evolution, which arises directly out of the initially different average of qualities presented by the isolated section of the species, as compared with the rest of that species[121]. [118] _Nature_, vol. xxxiii. p. 100.