Here then the idea is, as more fully expressed by Mr. Wallace in the context, that all the innumerable, frequently considerable, and generally elaborate "peculiarities of form, structure, colour, and ornament," which Darwin attributed to sexual selection, are really due to "the laws of growth." Diverse, definite, and constant though these specific peculiarities be, they are all but the accidental or adventitious accompaniments of "vigour," or "vital power," due to natural selection. Now, without waiting to dispute this view, which has already been dealt with in the chapter on Sexual Selection in Part I, it necessarily follows that "a large proportional number of specific characters," which, while presenting "no imaginable use," are very much less remarkable, less considerable, less elaborate, &c., must likewise be due to this "correlation with vital power." But if the principle of correlation is to be extended in this vague and general manner, it appears to me that the difference between Mr. Wallace and myself, with respect to the principle of utility, is abolished. For of course no one will dispute that the prime condition to the occurrence of "specific characters," whether useful or useless, is the existence of some form which has been denominated a "species" to present them; and this is merely another way of saying that such characters cannot arise except in correlation with a general fitness due to natural selection. Or, to put the case in Mr. Wallace's own words--"This development [of useless specific characters] will necessarily proceed by the agency of natural selection [as a necessary condition] _and the general laws which determine the production of colour and of ornamental appendages_." The case, therefore, is just the same as if one were to say, for example, that all the ailments of animals and plants proceed from correlation with life (as a necessary condition), "and the general laws which determine the production" of ill-health, or of specific disease. In short, the word "correlation" is here used in a totally different sense from that in which it is used by Darwin, and in which it is elsewhere used by Wallace for the purpose of sustaining his doctrine of specific characters as necessarily useful. To say that a useless character A is correlated with a useful one B, is a very different thing from saying that A is "correlated with vital power," or with the general conditions to the existence of the species to which it belongs. So far as the present discussion is concerned, no exception need be taken to the latter statement. For it simply surrenders the doctrine against which I am contending. IV. _Isolation._