This view may throw some light on the fact that the domestic animals which formerly inhabited the several districts in Great Britain, and the half-wild cattle lately kept in several British parks, differed slightly from one another; for these animals were prevented from wandering over the whole country and intercrossing, but would have crossed freely within each district or park[40].
Now, although I allow that Darwin never attributed to this principle of Amixia, or Independent Variability, anything like the degree of importance to which, in the opinion of Delbœuf, Gulick, Giard, and myself, it is entitled, the above passage appears to show that, as soon as the "view" was clearly "suggested" to his mind, he was so far from being unfavourably disposed towards it, that he added a paragraph to the last edition of his Variation for the express purpose of countenancing it. Nevertheless, later on the matter appears to have entirely escaped his memory; for in 1878 he wrote to Semper, that he did "not see at all more clearly than I did before, from the numerous cases which he [Wagner] has brought forward, how and why it is that a long isolated form should almost always become slightly modified[41]." I think this shows entire forgetfulness of the principle in question, because, if the latter is good for explaining the initial divergence of type as between separated stocks of "domesticated animals," much more must it be competent to explain the further divergence of type which is "almost always" observable in the case of "a long isolated form" under nature. The very essence of the principle being that, when divergence of type has once begun, this divergence must ipso facto proceed at an ever-accelerating pace, it is manifestly inconsistent to entertain the principle as explaining the first commencement of divergence, and then to ignore it as explaining the further progress of divergence. Hence, I can only conclude that Darwin had forgotten this principle altogether when he wrote his letter to Semper in 1878—owing, no doubt, as he says in the sentence which immediately follows, to his having "not attended much of late years to such questions."
So much, then, for Darwin's opinions. Next in order of time we must consider Moritz Wagner's essays on what he called the "Law of Migration[42]." The merit of these essays was, first, the firm expression of opinion upon the swamping effects of free intercrossing; and, second, the production of a large body of facts showing the importance of geographical isolation in the prevention of these effects, and in the consequent differentiation of specific types. On the other hand, the defect of these essays was, first, not distinguishing between evolution as monotypic and polytypic; and, second, not perceiving that geographical isolation is only one among a number of other forms of isolation. From these two radical oversights—which, however, were shared by all other writers of the time, with the partial exception of Darwin himself, as previously shown—there arose the following and most lamentable errors.
Over and over again Moritz Wagner insists, as constituting the fundamental doctrine of his attempted reform of Darwinism, that evolution by natural selection is impossible, unless natural selection be assisted by geographical isolation, in order to prevent the swamping effects of intercrossing[43]. Now, if instead of "evolution" he had said "divergence of type," and if instead of "geographical isolation" he had said "prevention of intercrossing," he would have enunciated the general doctrine which it has been the joint endeavour of Mr. Gulick and myself to set forth. But by not perceiving that "evolution" is of two radically different kinds—polytypic and monotypic—he entirely failed to perceive that, while for one of its kinds the prevention of intercrossing is an absolute necessity, for the other of its kinds the permission of intercrossing is a necessity no less absolute. And, again, in missing the fact that geographical isolation is but one of the many ways whereby intercrossing may be prevented, he failed to perceive that, even as regards the case of polytypic evolution, he greatly erred in representing this one form of isolation as being universally a necessary condition to the process. The necessary condition to this process is, indeed, the prevention of intercrossing by some means or another; but his unfortunate insistence on geographical separation as the only possible means to this end—especially when coupled with his no less unfortunate disregard of monotypic evolution—caused him to hinder rather than to advance a generalization which he had only grasped in part. And this generalization is, as now so repeatedly stated, that while the form of isolation which we know as natural selection depends for its action upon the intercrossing of all the individuals which it isolates (i. e. selects), when acting alone it can produce only monotypic evolution; but that when it is supplemented by any of the other numerous forms of isolation, it is furnished with the necessary condition to producing polytypic evolution—and this in as many lines of divergent change as there may be cases of this efficient separation.
Nevertheless, while we must lament these shortcomings on the part of Wagner, we ought to remember that he rendered important services in the way of calling attention to the swamping effects of free intercrossing, and, still more, in that of showing the high importance of geographical isolation as a factor of organic evolution. Therefore, although in an elaborate criticism of his views Weismann was easily able to dispose of his generalizations in the imperfect form that they presented, I do not think it was just in Weismann to remark, "if Wagner had confined himself to the statement that geographical isolation materially assists the process of natural selection, and thus also promotes the origination of new species, he would have met with little or no opposition; but then, of course, in saying this much, he would not have been saying anything new." No doubt, as I have just shown, he ought thus (as well as in other and still more important respects not perceived by Prof. Weismann) to have limited his statement; but, had he done so, it does not follow that he would not have been saying anything new. For, in point of fact, in as far as he said what was true, he did say a great deal that was also new. Thus, most of what he said of the principle of separation (apogamy) was as new as it was true, although, as we have seen, he said it to very little purpose on account of his identifying this principle as a whole with that of but one of its forms. Again, notwithstanding this great error, or oversight, he certainly showed of the particular form in question—viz. geographical isolation—that it was of considerably more importance than had previously been acknowledged. And this was so far a valuable contribution to the general theory of descent.
Prof. Weismann's essay, to which allusion has just been made[44], was, however, in all respects a great advance upon those of Wagner. It was not only more comprehensive in its view of the whole subject of geographical isolation, but likewise much more adequate in its general treatment thereof. Its principal defects, in my judgement, were, first, the inordinately speculative character of some of its parts, and, second, the restriction of its analysis to but one form of isolation—a defect which it shares with the essays of Wagner, and in quite as high a degree. Furthermore, although this essay had the great merit of enunciating the principle of Amixia, it did so in a very inefficient manner. For not only was this principle adduced with exclusive reference to geographical isolation, but even in regard to this one kind of isolation it was presented in a highly inconsistent manner, as I will now endeavour to show.
Weismann was led to perceive the principle in question by the consideration that new specific characters, when they first appear, do not all appear together in the same individuals: they appear one in one individual, another in another, a third in a third, &c.; and it is only in the course of successive generations that they all become blended in the same individuals by free intercrossing. Hence, the eventually emerging constant or specific type is the resultant of all the transitory or varietal types, when these have been fused together by intercrossing. From which Weismann deduces what he considers a general law—namely, that "the constancy of a specific type does not arise suddenly, but gradually; and it is established by the promiscuous crossing of all individuals[45]." From which again it follows, that this constancy must cease so soon as the condition which maintains it ceases—i. e. so soon as free intercrossing is prevented by the geographical isolation of a portion of the species from its parent stock.