Now, to begin with, this statement of the principle in question is not a good statement of it. There was no need while stating the doctrine that separation induces differentiation, to found the doctrine on any such highly speculative basis. In point of fact, there is no real evidence that specific types do attain their constancy in the way supposed; nor, for the purposes of the doctrine in question, is it necessary that there should be. For this doctrine does not need to show how the constancy has been attained; it only has to show that the constancy is maintained by free intercrossing, with the result that when free intercrossing is by any means prevented, divergence of character ensues. In short, the correct way of stating the principle is that which has been adopted by Delbœuf and Gulick—namely, the average characters of a separated portion of a species are not likely to be the same as those of the whole species; with the result that divergence of type will be set up in the separated portion by intercrossing within that portion. Or the principle may be presented as I presented it under the designation of "Independent Variability"—namely, "a specific type may be regarded as the average mean of all individual variations, any considerable departure from this average mean being, however, checked by intercrossing," with the result that when intercrossing is prevented between a portion of a species and the rest of the species, "this population is permitted to develop an independent history of its own, shielded from intercrossing with its parent form[46]."
Not only, however, is Weismann's principle of "Amixia" thus very differently stated from that of my "Independent Variability" (apogamy), or Gulick's "Independent Generation"; but, apparently owing to this difference of statement, the principle itself is not the same. In particular, while Weismann holds with us that when new characters arise in virtue of the mere prevention of intercrossing with parent forms these new characters will be of non-utilitarian kind[47], he appears to think that divergence of character under such circumstances is not likely to go on to a specific value. Now, it is of importance to observe why he arrives at this conclusion, which is not only so different from that of Delbœuf, Gulick, and myself, but apparently so inconsistent with his own recognition of the diversifying effect of "Amixia" as regards the formation of permanent varieties. For, as we have already seen while considering Darwin's views on this same principle of "Amixia," it is highly inconsistent to recognize its diversifying effect up to the stage of constituting fixed varieties, and then not to recognize that, so much divergence of character having been already secured by the isolation alone, much more must further divergence continue, and continue at an ever accelerating pace—as Delbœuf and Gulick have so well shown. What, then, is the explanation of this apparent inconsistency on Weismann's part? The explanation evidently is that, owing to his erroneous statement of the principle, he misses the real essence of it. For, in the first place, he does not perceive that this essence consists in an initial difference of average characters on the part of the isolated colony as compared with the rest of their species. On the contrary, he loses himself in a maze of speculation about all species having had what he calls "variation-periods," or eruptions of general variability alternating with periods of repose—both being as unaccountable in respect of their causation as they are hypothetical in respect of their occurrence. From these speculations he concludes, that isolation of a portion of a species will then only lead to divergence of character when the isolation happens to coincide with a "variation-period" on the part of the species as a whole, and that the divergence will cease so soon as the "variation-period" ceases. Again, in the second place as previously remarked, equally with Wagner whom he is criticizing, he fails to perceive that geographical isolation is not the only kind of isolation, or the only possible means to the prevention of free intercrossing. And the result of this oversight is, that he thinks amixia can act but comparatively seldom upon sufficiently small populations to become a factor of much importance in the differentiation of species. Lastly, in the third place, owing to his favourite hypothesis that all species pass through a "variation-period," he eventually concludes that the total amount of divergence of type producible by isolation alone (even in a small population) can never be greater than that between the extremes of variation which occur within the whole species at the date of its partition (p. 75). In other words, the possibility of change due to amixia alone is taken to be limited by the range of deviation from the general specific average, as manifested by different individual variations, before the species was divided. Thus the doctrine of amixia fails to recognize the law of Delbœuf, or the cumulative nature of divergence of type when once such divergence begins in a separated section. Therefore, in this all-important—and, indeed, essential—respect, amixia differs entirely from the principle which has been severally stated by Delbœuf, Gulick, and myself.
Upon the whole, then, we must say that although Professor Weismann was the first to recognize the diversifying influence of merely indiscriminate isolation per se (apogamy), he did so only in part. He failed to distinguish the true essence of the principle, and by overlaying it with a mass of hypothetical speculation, concealed even more of it than he revealed.
The general theory of Isolation, as independently worked out by Mr. Gulick and myself, has already been so fully explained, that it will here be sufficient merely to enumerate its more distinguishing features. These are, first, drawing the sharpest possible line between evolution as monotypic and polytypic; second, showing that while for the former the peculiar kind of isolation which is presented by natural selection suffices of itself to transform a specific type, in order to work for the latter, or to branch a specific type, natural selection must necessarily be assisted by some other kind of isolation; third, that even in the absence of natural selection, other kinds of isolation may be sufficient to effect specific divergence through independent generation alone; fourth, that, nevertheless, natural selection, where present, will always accelerate the process of divergence; fifth, that monotypic evolution by natural selection depends upon the presence of intercrossing, quite as much as polytypic evolution (whether with or without natural selection) depends upon the absence of it; sixth, that, having regard to the process of evolution throughout all taxonomic divisions of organic nature, we must deem the physiological form of isolation as the most important, with the exception only of natural selection.
The only difference between Mr. Gulick's essays and my own is, that, on the one hand, he has analyzed much more fully than I have the various forms of isolation; while, on the other hand, I have considered much more fully than he has the particular form of physiological isolation which so frequently obtains between allied species. This particular form of physiological isolation I have called "physiological selection," and claim for it so large a share in the differentiation of specific types as to find in it a satisfactory explanation of the contrast between natural species and artificial varieties in respect of cross-infertility.
Mr. Wallace, in his Darwinism, has done good service by enabling all other naturalists clearly to perceive how natural selection alone produces monotypic evolution—namely, through the free intercrossing of all individuals which have not been eliminated by the isolating process of natural selection itself. For he very lucidly shows how the law of averages must always ensure that in respect of any given specific character, half the individuals living at the same time and place will present the character above, and half below its mean in the population as a whole. Consequently, if it should ever be of advantage to a species that this character should undergo either increase or decrease of its average size, form, colour, &c., there will always be, in each succeeding generation, a sufficient number of individuals—i. e. half of the whole—which present variations in the required direction, and which will therefore furnish natural selection with abundant material for its action, without the need of any other form of isolation. It is to be regretted, however, that while thus so clearly presenting the fact that free intercrossing is the very means whereby natural selection is enabled to effect monotypic evolution, he fails to perceive that such intercrossing must always and necessarily render it impossible for natural selection to effect polytypic evolution. A little thought might have shown him that the very proof which he gives of the necessity of intercrossing where the transmutation of species is concerned, furnishes, measure for measure, as good a proof of the necessity of its absence where the multiplication of species is concerned. In justice to him, however, it may be added, that this distinction between evolution as monotypic and polytypic (with the important consequence just mentioned) still continues to be ignored also by other well-known evolutionists of the "ultra-Darwinian" school. Professor Meldola, for example, has more recently said that in his opinion the "difficulty from intercrossing" has been in large part—if not altogether—removed by Mr. Wallace's proof that natural selection alone is capable of effecting [monotypic] evolution; while he regards the distinction between monotypic and polytypic evolution as mere "verbiage[48]."
It is in relation to my presentment of the impossibility of natural selection alone causing polytypic evolution, that Mr. Wallace has been at the pains to show how the permission of intercrossing (panmixia) is necessary for natural selection in its work of causing monotypic evolution. And not only has he thus failed to perceive that the "difficulty" which intercrossing raises against the view of natural selection being of itself capable of causing polytypic evolution in no way applies to the case of monotypic; but as regards this "difficulty," where it does apply, he says:—