I had here supplied a number of similar quotations from writers in various other departments of systematic work, but afterwards struck them out as superfluous. For it is not to be anticipated that any competent naturalist will nowadays dispute that the terms "variety," "species," and "genus" stand for merely conventional divisions, and that whether a given form shall be ranked under one or the other of them is often no more than a matter of individual taste. From the nature of the case there can be no objective, and therefore no common, standards of delimitation. This is true even as regards any one given department of systematic work; but when we compare the standards of delimitation which prevail in one department with those which prevail in another, it becomes evident that there is not so much as any attempt at agreeing upon a common measure of specific distinction.
But what, it may well be asked, is the use of thus insisting upon well-known facts, which nobody will dispute? Well, in the first place, we have already seen, in the last chapter, that it is incumbent on those who maintain that all species, or even all specific characters, must be due to natural selection, to tell us what they mean by a species, or by characters as specific. If I am told to believe that the definite quality A is a necessary attribute of B, and yet that B is "not a distinct entity," but an undefinable abstraction, I can only marvel that any one should expect me to be so simple. But, without recurring to this point, the use of insisting on the facts above stated is, in the second place, that otherwise I cannot suppose any general reader could believe them in view of what is to follow. For he cannot but feel that the cost of believing them is to render inexplicable the mental processes of those naturalists who, in the face of such facts, have deduced the following conclusions.
The school of naturalists against which I am contending maintains, as a generalization deduced from the theory of natural selection, that all species, or even all specific characters, must necessarily owe their origin to the principle of utility. Yet this same school does not maintain any such generalization, either with regard to varietal characters on the one hand, or to generic characters on the other. On the contrary, Professor Huxley, Mr. Wallace, and all other naturalists who agree with them in refusing to entertain so much as the abstract possibility of any cause other than natural selection having been productive of species, fully accept the fact of other causes having been largely concerned in the production of varieties, genera, families, and all higher groups, or of the characters severally distinctive of each. Indeed, Mr. Wallace does not question what appears to me the extravagant estimate of Professor Cope, that the non-adaptive characters distinctive of those higher groups are fully equal, in point of numbers, to the adaptive. But, surely, if the theory of evolution by natural selection is, as we all agree, a true theory of the origin of species, it must likewise be a true theory of the origin of genera; and if it be supposed essential to the integrity of the theory in its former aspect that all specific characters should be held to be useful, I fail to see how, in regard to its latter aspect, we are so readily to surrender the necessary usefulness of all generic characters. And exactly the same remark applies to the case of constant "varieties," where again the doctrine of utility as universal is not maintained. Yet, according to the general theory of evolution, constant varieties are what Darwin termed "incipient species," while species are what may be termed "incipient genera." Therefore, if the doctrine of utility as universal be conceded to fail in the case of varieties on the one hand and of genera on the other, where is the consistency in maintaining that it must "necessarily" hold as regards the intermediate division, species? Truly the shade of Darwin may exclaim, "Save me from my friends." And truly against logic of this description a follower of Darwin must find it difficult to argue. If one's opponents were believers in special creation, and therefore stood upon some definite ground while maintaining this difference between species and all other taxonomic divisions, there would at least be some issue to argue about. But when on the one hand it is conceded that species are merely arbitrary divisions, which differ in no respect as to the process of their evolution from either varieties or genera, while on the other hand it is affirmed that there is thus so great a difference in the result, all we can say is that our opponents are entangling themselves in the meshes of a sheer contradiction.
Or, otherwise stated, specific characters differ from varietal characters in being, as a rule, more pronounced and more constant: on this account advocates of utility as universal apply the doctrine to species, while they do not feel the "necessity" of applying it to varieties. But now, generic and all higher characters are even more constant and more pronounced than specific characters—not to say, in many cases, more generally diffused over a larger number of organisms usually occupying larger areas. Therefore, a fortiori, if for the reasons above stated evolutionists regard it as a necessary deduction from the theory of natural selection that all specific characters must be useful, much more ought it to be a necessary deduction from this theory that all generic, and still more all higher, characters must be useful. But, as we have seen, this is not maintained by our opponents. On the contrary, they draw the sharpest distinction between specific and all other characters in this respect, freely conceding that both those below and those above them need not—and very often do not—present any utilitarian significance.
Although it appears to me that this doctrine is self-contradictory, and on this ground alone might be summarily dismissed, as it is now held in one or other of its forms by many naturalists, I will give it a more detailed consideration in both its parts—namely, first with respect to the distinction between varieties and species, and next with respect to the distinction between species and genera.
Until it can be shown that species are something more than merely arbitrary divisions, due to the disappearance of intermediate varietal links; that in some way or another they are "definite entities," which admit of being delineated by the application of some uniform or general principles of definition; that, in short, species have only then been classified as such when it has been shown that the origin of each has been due to the operation of causes which have not been concerned in the production of varieties;—until these things are shown, it clearly remains a gratuitous dogma to maintain that forms which have been called species differ from forms which have been called varieties in the important respect, that they (let alone each of all their distinctive characters) must necessarily have been due to the principle of utility. Yet, as we have seen, even Mr. Wallace allows that a species is "not a distinct entity," but "an assemblage of individuals which have become somewhat modified in structure, form, and constitution"; while estimates of the kinds and degrees of modification which are to be taken as of specific value are conceded to be undefinable, fluctuating, and in not a few cases almost ludicrously divergent.
Perhaps one cannot more forcibly present the rational value of this position than by noting the following consequences of it. Mr. Gulick writes me that while studying the land-shells of the Sandwich Islands, and finding there a rich profusion of unique varieties, in cases where the intermediate varieties were rare he could himself have created a number of species by simply throwing these intermediate varieties into his fire. Now it follows from the dogma which we are considering, that, by so doing, not only would he have created new species, but at the same time he would have proved them due to natural selection, and endowed the diagnostic characters of each with a "necessarily" adaptive meaning, which previously it was not necessary that they should present. Before his destruction of these intermediate varieties, he need have felt himself under no obligation to assume that any given character at either end of the series was of utilitarian significance: but, after his destruction of the intermediate forms, he could no longer entertain any question upon the matter, under pain of being denounced as a Darwinian heretic.
Now the application is self-evident. It is a general fact, which admits of no denial, that the more our knowledge of any flora or fauna increases, the greater is the number of intermediate forms which are brought to light, either as still existing or as having once existed. Consequently, the more that such knowledge increases, the more does our catalogue of "species" diminish. As Kerner says, "bad species" are always multiplying at the expense of "good species"; or, as Oscar Schmidt (following Häckel) similarly remarks, if we could know as much about the latter as we do about the former, "all species, without any exception, would become what species-makers understand by 'bad species'[129]." Hence we see that, just as Mr. Gulick could have created good species by secretly destroying his intermediate varieties, so has Nature produced her "good species" for the delectation of systematists. And just as Mr. Gulick, by first hiding and afterwards revealing his intermediate forms, could have made the self-same characters in the first instance necessarily useful, but ever afterwards presumably useless, so has Nature caused the utility of diagnostic characters to vary with our knowledge of her intermediate forms. It belongs to the essence of our theory of descent, that in all cases these intermediate forms must either be now existing or have once existed; and, therefore, that the work of species-makers consists in nothing more than marking out the lacunae in our knowledge of them. Yet we are bound to believe that wherever these lacunae in our knowledge occur, there occurs also the objective necessity of causation as utilitarian—a necessity, however, which vanishes so soon as our advancing information supplies the intermediate forms in question. It may indeed appear strange that the utility or non-utility of organic structures should thus depend on the accidents of human knowledge; but this is the Darwinian faith, and he who doubts the dogma is to be anathema.
Turning next to the similar distinction which it is sought to draw between species and genera, here it will probably be urged, as I understand it to be urged by Mr. Wallace, that generic characters (and still more characters of families, orders, &c.) refer back to so remote a state of things that utility may have been present at their birth which has disappeared in their maturity. In other words, it is held that all generic characters were originally specific characters; that as such they were all originally of use; but that, after having been rendered stable by heredity, many of them may have ceased to be of service to the descendants of those species in which they originated, and whose extinction has now made it impossible to divine what that service may have been.