I have dwelt thus at some length upon a mere matter of definition because, as we shall now find, although it is but a matter of definition, it is fraught with consequences of no small importance to the general theory of descent. Starting from an erroneous definition of the theory of natural selection as primarily a theory of the origin of species, both friends and foes of the theory have concluded that the principle of utility must by hypothesis be of universal occurrence so far as species are concerned; whereas, if once these naturalists were to perceive that their definition of the theory is erroneous, they would likewise perceive that their conclusion cannot follow deductively from the theory itself. If such a conclusion is to be established at all, it can only be by other and independent evidence of the inductive kind—to wit, by actual observation.

Hence we see the importance of starting with an accurate definition of the theory before proceeding to examine the doctrine of utility as of universal application to species—a doctrine which, as just stated, has been habitually and expressly deduced from the theory. This doctrine occurs in two forms; or, more correctly, there are with reference to this subject two distinct doctrines, which partly coincide and partly exclude one another. First, it is held by some naturalists that all species must necessarily owe their origin to natural selection. And secondly, it is held by other naturalists, that not only all species, but likewise all specific characters must necessarily do the same. Let us consider these two doctrines separately.

The first, and less extensive doctrine, rests on the deduction that every species must owe its differentiation as a species to the evolution of at least one adaptive character, which is peculiar to that species. Although, when thus originated, a species may come to present any number of other peculiar characters of a non-adaptive kind, these merely indifferent peculiarities are supposed to hang, as it were, on the peg supplied by the one adaptive peculiarity; it is the latter which conditions the species, and so furnishes an opportunity for any number of the former to supervene. But without the evolution of at least one adaptive character there could have been no distinct species, and therefore no merely adventitious characters as belonging to that species. I will call this the Huxleyan doctrine, because Professor Huxley is its most express and most authoritative supporter.

The second and more extensive doctrine I will call, for the same reason, the Wallacean doctrine. This is, as already stated, that it follows deductively from the theory of natural selection, that not only all species, but even all the distinctive characters of every species, must necessarily be due to natural selection; and, therefore, can never be other than themselves useful, or, at the least, correlated with some other distinctive characters which are so.

Here, however, I should like to remark parenthetically, that in choosing Professor Huxley and Mr. Wallace as severally representative of the doctrines in question, I earnestly desire to avoid any appearance of discourtesy towards such high authorities.

I am persuaded—as I shall hereafter seek to show Darwin was persuaded—that the doctrine of utility as universal where species are concerned, is, in both the above forms, unsound. But it is less detrimental in its Huxleyan than in its Wallacean form, because it does not carry the erroneous deduction to so extreme a point. Therefore let us first consider the doctrine in its more restricted form, and then proceed, at considerably greater length, to deal with it in its more extended form.

The doctrine that all species must necessarily be due to natural selection, and therefore must severally present at least one adaptive character, appears to me doubly erroneous.

In the first place, it is drawn from what I have just shown to be a false premiss; and, in the second place, the conclusion does not follow even from this premiss. That the premiss—or definition of the theory as primarily a theory of the origin of species—is false, I need not wait again to argue. That the conclusion does not follow even from this erroneous premiss, a very few words will suffice to prove. For, even if it were true that natural selection is primarily a theory of the origin of species, it would not follow that it must therefore be a theory of the origin of all species. This would only follow if it were first shown that the theory is not merely a theory of the origin of species, but the theory of the origin of species—i.e. that there can be no further theory upon this subject, or any cause other than natural selection which is capable of transforming any single specific type.

Needless to say, this cannot be shown by way of deduction from the theory of natural selection itself—which, nevertheless, is the only way whereby it is alleged that the doctrine is arrived at[86].

From the doctrine of utility as advocated by Professor Huxley, we may now pass on to consider it in the much more comprehensive form advocated by Mr. Wallace. Of course it is obvious that if the doctrine is erroneous in its Huxleyan form, much more must it be so in its Wallacean; and, therefore, that having shown its erroneousness in its less extended application, there is little need to consider it further in its more extended form. Looking, however, to its importance in this more extended application, I think we ought to examine it independently as thus presented by Mr. Wallace and his school. Let us therefore consider, on its own merits, the following statement:—It follows directly from the theory of natural selection that not only all species, but likewise all specific characters, must be due to natural selection, and, therefore, must all be of use to the species which present them, or else correlated with other characters which are so.

It seems worth while to observe, in limine, that this doctrine is contradicted by that of Professor Huxley. For supposing natural selection to be the only principle concerned in the origin of all species, it by no means follows that it is the sole agency concerned in the origin of all specific characters. It is enough for the former proposition if only some of the characters distinctive of any given species—nay, as he very properly expresses it, if only one such character—has been due to natural selection; for it is clear that, as he adds, "any number of indifferent [specific] characters" may thus have been furnished with an opportunity, so to speak, of being produced by causes other than natural selection. Hence, as previously remarked, the Huxleyan doctrine, although coinciding with the Wallacean up to the point of maintaining utility as the only principle which can be concerned in the origin of species, designedly excludes the Wallacean doctrine where this proceeds to extend any similar deduction to the case of specific characters[87].

In the next place, and with special reference to the Wallacean doctrine, it is of importance to observe that, up to a certain point there is complete agreement between Darwinists of all schools. We all accept natural selection as a true cause of the origin of species (though we may not all subscribe to the Huxleyan deduction that it is necessarily a cause of the origin of all species). Moreover, we agree that specific characters are often what is called rudimentary or vestigial; and, once more, that our inability to detect the use of any given structure or instinct is no proof that such a structure or instinct is actually useless, seeing that it may very probably possess some function hitherto undetected, or possibly undetectable. Lastly, we all agree that a structure which is of use may incidentally entail the existence of some other structure which is not of use; for, in virtue of the so-called principle of correlation, the useless structure may be an indirect consequence of natural selection, since its development may be due to that of the useful structure, with the growth of which the useless one is correlated.

Nevertheless, while fully conceding all these facts and principles to the Wallacean party, those who think with Professor Huxley—and still more, of course, those few naturalists who think as I do——are unable to perceive that they constitute any grounds for holding the doctrine that all specific characters are, or formerly have been, directly or indirectly due to natural selection. My own reasons for dissenting from this Wallacean doctrine are as follows.

From what has just been said, it will be apparent that the question in debate is not merely a question of fact which can be settled by a direct appeal to observation. If this were the case, systematic naturalists could soon settle the question by their detailed knowledge of the structures which are severally distinctive of any given group of species. But so far is this from being the case, that systematic naturalists are really no better qualified to adjudicate upon the matter than are naturalists who have not devoted so much of their time to purely diagnostic work. The question is one of general principles, and as such cannot be settled by appeals to special cases. For example, suppose that the rest of this chapter were devoted to a mere enumeration of cases where it appears impossible to suggest the utility of certain specific characters, although such cases could be adduced by the thousand, how should I be met at the end of it all? Not by any one attempting to suggest the utility, past or present, of the characters named; but by being told that they must all present some hidden use, must be vestigial, or else must be due to correlation. By appealing to one or other of these assumptions, our opponents are always able to escape the necessity of justifying their doctrine in the presence of otherwise inexplicable facts. No matter how many seemingly "indifferent characters" we may thus accumulate, Mr. Wallace and his followers will always throw upon us the impossible burden of proving the negative, that these apparently useless characters do not present some hidden or former use, are not due to correlation, and therefore have not been produced by natural selection. It is in vain to retort that the burden of proof really lies the other way, or on the side of those who affirm that there is utility where no man can see it, or that there is correlation where no one can detect it. Thus, so far as any appeal to particular facts is concerned, it does not appear that there is any modus vivendi. Our opinions upon the question are really determined by the views which we severally take on matters of general principle. The issue, though it has a biological bearing, is a logical issue, not a biological one: it turns exclusively on those questions of definition and deduction with which we have just been dealing.

But although it thus follows that we cannot determine in fact what proportion of apparently useless characters are or are not really useful, we may very easily determine in fact what proportion of specific characters fail to present any observable evidences of utility. Yet, even upon this question of observable fact, it is surprising to note the divergent statements which have of late years been made by competent writers; statements in fact so divergent that they can only be explained by some want of sufficient thought on the part of those naturalists who are antecedently persuaded that all specific characters must be either directly or indirectly due to natural selection. Hence they fail to give to apparently useless specific characters the attention which, apart from any such antecedent persuasion, they deserve. For example, a few years ago I incidentally stated in a paper before the Linnaean Society, that "a large proportional number of specific characters" are of a trivial and apparently unmeaning kind, to which no function admits of being assigned, and also stated that Darwin himself had expressly given utterance to the same opinion. When these statements were made, I did not anticipate that they would be challenged by anybody, except perhaps, by Mr. Wallace. And, in order now to show that my innocence at that time was not due to ignorance of contemporary thought on such matters, a sentence may here be quoted from a paper which was read at the meeting of the British Association of the same year, by a highly competent systematic naturalist, Mr. Henry Seebohm, and soon afterwards extensively republished. Criticizing adversely my then recently published paper, he said:—

"I fully admit the truth of this statement; and I presume that few naturalists would be prepared to deny that 'distinctions of specific value frequently have reference to structures which are without any utilitarian significance[88].'"

But since that time the course of Darwinian speculation has been greatly influenced by the writings of Weismann, who, among other respects in which he out-darwins Darwin, maintains the doctrine of utility as universal. In consequence of the influence which these writings have exercised, I have been more recently and extensively accused of "heresy" to Darwinian principles, for having stated that "a large proportional number of specific characters" do not admit of being proved useful, or correlated with other characters that are useful. Now, observe, we have here a simple question of fact. We are not at present concerned with the question how far the argument from ignorance may be held to apply in mitigation of such cases; but we are concerned only with the question of fact, as to what proportional number of cases actually occur where we are unable to suggest the use of specific characters, or the useful characters with which these apparently useless ones are correlated. I maintain, as a matter of fact, that the cases in question embrace "a large proportional number of specific characters." On the other hand, I am accused of betraying ignorance of species, and of the work of "species-makers," in advancing this statement; and have been told by Mr. Wallace, and others of his school, that there is absolutely no evidence to be derived from nature in support of my views. Well, in the first place, if this be the case, it is somewhat remarkable that a large body of competent naturalists, such as Bronn, Broca, Nägeli, Kerner, Sachs, De Vries, Focke, Henslow, Haeckel, Kölliker, Eimer, Giard, Pascoe, Mivart, Seebohm, Lloyd Morgan, Dixon, Beddard, Geddes Gulick, and also, as we shall presently see, Darwin himself, should have fallen into the same error. And it is further remarkable that the more a man devotes himself to systematic work in any particular department—whether as an ornithologist, a conchologist, an entomologist, and so forth—the less is he disposed to accept the dogma of specific characters as universally adaptive characters. But, in the second place, and quitting considerations of mere authority, I appeal to the facts of nature themselves; and will now proceed, as briefly as possible, to indicate the result of such an appeal.

For the following reasons, that birds and mammals seem to furnish the best field for testing the question by direct observation. First, these classes present many genera which have been more carefully worked out than is usually the case with genera of invertebrates, or even of cold-blooded vertebrates. Secondly, they comprise many genera each including a large number of species, whose habits and conditions of life are better known than is the case with species belonging to large genera of other classes. Thirdly, as birds and mammals represent the highest products of evolution in respect of organization, a more severe test is imposed than could be imposed elsewhere, when the question is as to the utility of specific characters; for if these highest products of organization fail to reveal, in a large proportional number of cases, the utility of their specific characters, much more is this likely to be the case among organic beings which stand lower in the scale of organization, and therefore, ex hypothesi, are less elaborate products of natural selection. Fourthly, and lastly, birds and mammals are the classes which Mr. Wallace has expressly chosen to constitute his ground of argument with regard to the issue on which we are now engaged.

It would take far too long to show, even in epitome, the results of this inquiry. Therefore I will only state the general upshot. Choosing genera of birds and mammals which contain a large number of species whose diagnostic characters have been worked out with most completeness, I restricted the inquiry to specific distinctions of colour, not only for the sake of having a uniform basis for comparisons, but still more because it seemed that the argument from our ignorance of possibly unknown uses could be more successfully met in the case of slight differences of colour or of shading, than in that of any differences of structure or of form. Finally, after tabulating all the differences of colour which are given as diagnostic of each species in a genus, and placing in one column those which may conceivably be useful, while placing in another column those of which it appeared inconceivable that any use could be suggested, I added up the figures in the two columns, and thus obtained a grand total of all the specific characters of the genus in respect of colours, separated into the two classes of conceivably useful and apparently useless. Now, in all cases the apparently useless characters largely preponderated over the conceivably useful ones; and therefore I abundantly satisfied myself regarding the accuracy of my previous statement, that a large proportional number—if not an actual majority—of specific characters belong to the latter category.

The following is a brief abstract of these results.

With respect to Birds, a large number of cases were collected wherein the characters of allied species differ from one another in such minute respects of colour or shading, that it seemed unreasonable to suppose them due to any selective value to the birds in question. It is needless—even if it were practicable on the present occasion—to adduce this evidence in detail, since an exceedingly good sample of it may be found in a small book which is specially devoted to considering the question in its relation to birds. I allude to an essay by Mr. Charles Dixon, entitled Evolution without Natural Selection (1885). In this work Mr. Dixon embodies the results of five years' "careful working at the geographical distribution and variations of plumage of Palaearctic birds and their allies in various other parts of the world"; and shows, by a large accumulation of facts, not only that there is no utility to be suggested in reference to the minute or trivial differences of colouration which he describes; but also that these differences are usually correlated with isolation on the one hand, or with slight differences of climate on the other. Now it will be shown later on that both these agents can be proved, by independent evidence, capable of inducing changes of specific type without reference to utility: therefore the correlation which Mr. Dixon unquestionably establishes between apparently useless (because utterly trivial) specific distinctions on the one hand, and isolation or climatic change on the other, constitutes additional evidence to show that the uselessness is not only apparent, but real. Moreover I have collected a number of cases where such minute differences of colour between allied species of birds happen to affect parts of the plumage which are concealed—as for instance, the breast and abdomen of creepers. In such cases it seems impossible to suggest how natural selection can have operated, seeing that the parts affected are not exposed to the view either of enemies or of prey.

Analogous illustrations to any amount may be drawn from Mammals. For instance, I have worked through the Marsupials with the aid of Mr. Oldfield Thomas' diagnostic description of their numerous species. Now, let us take any one of the genera, such as the kangaroos. This comprises 23 species living on an island continent of high antiquity, and not exposed to the depredations of any existing carnivorous enemies; so that there is here no present need to vary colour for purposes of protection. Moreover, in all cases the diagnostic distinctions of colour are so exceedingly trivial, that even if large carnivora were recently abundant in Australia, no one could reasonably suggest that the differences in question would then have been protective. On an average, each of the 23 species presents rather more than 20 peculiarities of shading, which are quoted as specifically diagnostic. Altogether there are 474 of these peculiarities distributed pretty evenly among the 23 species; and in no case can I conceive that utility can be suggested.

Hitherto we have been considering the question of fact, as to whether "a large proportional number of specific characters" do or do not admit of having their utility demonstrated, or even so much as plausibly suggested. In the result, I can only conclude that this question of fact is really not an open one, seeing that it admits of an abundantly conclusive answer by any naturalist who will take the trouble to work through the species of any considerable number of genera in the way above indicated. But although the question of fact is thus really closed, there remains a more ultimate question as to its theoretical interpretation. For, as already pointed out, no matter how great an accumulation of such facts may be collected, our opponents are always able to brush them aside by their a priori appeal to the argument from ignorance. In effect they say—We do not care for any number of thousands of such facts; it makes no difference to us what "proportional number" of specific characters fail to show evidence of utility; you are merely beating the air by adducing them, for we are already persuaded, on antecedent grounds, that all specific characters must be either themselves useful, or correlated with others that are, whether or not we can perceive the utility, or suggest the correlation.

To this question of theoretical interpretation, therefore, we must next address ourselves. And here, first of all, I should like to point out how sturdy must be the antecedent conviction of our opponents, if they are to maintain it in the face of such facts as have just been adduced. It must be remembered that this antecedent conviction is of a most uncompromising kind. By its own premisses it is committed to the doctrine that all specific characters, without a single exception, must be either useful, vestigial, or correlated. Well, if such be the case, is it not somewhat astonishing that out of 474 differences of colour which are distinctive of the 23 species of the genus Macropus, no single one appears capable of having any utility demonstrated, or indeed so much as suggested? For even the recent theory that slight differences of colour, which cannot be conceived as serving any other purpose, may enable the sexes of the same species quickly to recognize each other, is not here available. The species of the genus Macropus are more conspicuously distinguished by differences of size and form than by these minute differences of colour; and therefore no such use can be attributed to the latter. And, as previously stated, even within the order Marsupialia the genus Macropus is not at all exceptional in this respect; so that by including other genera of the order it would be easy to gather such apparently indifferent specific characters by the hundred, without any one of them presenting evidence—or even suggestion—of utility. How robust therefore is the faith of an a priori conviction which can stand against such facts as these! What, then, are the a priori grounds on which it stands? Mr. Wallace, the great leader of this school of thought, says:—

"It is a necessary deduction from the theory of natural selection, that none of the definite facts of organic nature, no special organ, no characteristic form or marking, no peculiarities of instinct or of habit, no relations between species or between groups of species, can exist, but which must now be, or once have been, useful to the individuals or the races which possess them[89]."

Here, then, we have in brief compass the whole essence of our opponents' argument. It is confessedly an argument a priori, a deduction from the theory of natural selection, a supposed consequence of that theory which is alleged to be so necessary that to dispute the consequence is tantamount to denying the theory from which it is derived. In short, as before stated, it is a question of theory, not a question of fact: our difference of opinion is logical, not biological: it depends on our interpretation of principles, not on our observation of species. It will therefore be my endeavour to show that the reasoning in question is fallacious: that it is not a necessary deduction from the theory of natural selection that no characteristic form or marking, no peculiarities of instinct or of habit, can exist, but which must now be, or once have been, useful, or correlated with some other peculiarity that is useful.

"The tuft of hair on the breast of a wild turkey-cock cannot be of any use, and it is doubtful whether it can be ornamental in the eyes of the female bird;—indeed, had the tuft appeared under domestication, it would have been called a monstrosity[90]."

As a matter of common sense, unprejudiced by dogma, this appears to be a perfectly sound judgement; but if Wallace had asked Darwin to prove such a negative, Darwin could only have replied that it was for Wallace to prove the affirmative—and thus the issue would have been thrown back upon a discussion of general principles. Then Wallace would have said—"The assertion of inutility in the case of any organ or peculiarity which is not a rudiment or a correlation is not, and can never be, the statement of a fact, but merely an expression of our ignorance of its purpose or origin[91]." Darwin, however, would have replied:—"Our ignorance of the laws of variation is profound"; and while, on this account, we ought "to be extremely cautious in pretending to decide what structures are now, or have formerly been, of use to each species," in point of fact "there can be little doubt that the tendency to vary in the same manner has often been so strong, that all individuals of the same species have been similarly modified without the aid of any form of selection[92]."

It will be my endeavour in the following discussion to show that Darwin would have had an immeasurable advantage in this imaginary debate.

To begin with, Wallace's deductive argument is a clear case of circular reasoning. We set out by inferring that natural selection is a cause from numberless cases of observed utility as an effect: yet, when "in a large proportional number" of cases we fail to perceive any imaginable utility, it is argued that nevertheless utility must be there, since otherwise natural selection could not have been the cause.

Be it observed, in any given case we may properly anticipate utility as probable, even where it is not perceived; because there are already so enormous a number of cases where it is perceived, that, if the principle of natural selection be accepted at all, we must conclude with Darwin that it is "the main means of modification." Therefore, in particular cases of unperceived utility we may take this antecedent probability as a guide in our biological researches—as has been done with such brilliant success both by Darwin and Wallace, as well as by many of their followers. But this is a very different thing from laying down the universal maxim, that in all cases utility must be present, whether or not we shall ever be able to detect it[93]. For this universal maxim amounts to an assumption that natural selection has been the "exclusive means of modification." That it has been "the main means of modification" is proved by the generality of the observed facts of adaptation. That it has been "the exclusive means of modification," with the result that these facts are universal, cannot be thus proved by observation. Why, then, is it alleged? Confessedly it is alleged by way of deduction from the theory of natural selection itself. Or, as above stated, after having deduced the theory from the facts, it is sought to deduce the facts from the theory.

Thus far I have been endeavouring to show that the universality of adaptation cannot be inferred from its generality, or from the theory of natural selection itself. But, of course, the case would be quite different if there were any independent evidence—or rather, let us say, any logical argument—to show that natural selection is "the exclusive means of modification." For in this event it would no longer involve circular reasoning to maintain that all specific characters are likewise adaptive characters. It might indeed appear antecedently improbable that no other principle than natural selection can possibly have been concerned in the differentiation of those relatively permanent varieties which we call species—that in all the realm of organic nature, and in all the complexities of living processes, there is no room for any other influence in the production of change, even of the most trivial and apparently unmeaning kind. But if there were any good evidence or logical argument to the contrary, this antecedent presumption would have to give way; and the certainty that all specific characters are likewise adaptive characters would be determined by the cogency of such evidence or argument as could be adduced. In short, we are not entitled to conclude—and still less does it follow "as a necessary deduction from the theory of natural selection"—that all the details of specific differentiation must in every case be either useful, vestigial, or correlated, unless it has been previously shown, by independent evidence, or accurate reasoning, that there is no room for any other principle of specific change.

This, apparently, is the central core of the question. Therefore I will now proceed to consider such arguments as have been adduced to prove that, other than natural selection, there can have been no "means of modification." And, after having exhibited the worthlessness of these arguments, I will devote the next chapter to showing that, as a matter of observable fact, there are a considerable number of other principles, which can be proved to be capable of producing such minute differences of form and colour as "in a large proportional number" of cases constitute diagnostic distinctions between species and species.

First, then, for the reasons a priori—and they are confessedly a priori—which have been adduced to prove that natural selection has been what in Darwin's opinion it has not been,—"the exclusive means of modification." Disregarding the Lamarckian factors—which, even if valid, have but little relation to the present question, seeing that they are concerned, almost exclusively, with the evolution of adaptive characters—it is alleged that natural selection must occupy the whole field, because no other principle of change can be allowed to operate in the presence of natural selection. Now, I fully agree that this statement may hold as regards any principle of change which is deleterious; but clearly it does not hold as regards any principle which is merely neutral. If any one were to allege that specific characters are frequently detrimental to the species presenting them, he would no doubt lay himself open to the retort that natural selection could not allow such characters to persist; or, which amounts to the same thing, that it does "necessarily follow from the theory of natural selection" that specific characters can never be in any large number, or in any large measure, harmful to the species presenting them. But where the statement is that specific characters are frequently indifferent—again to use Professor Huxley's term—the retort loses all its relevancy. No reason has ever been shown why natural selection should interfere with merely indifferent characters, supposing such to have been produced by any of the agencies which we shall presently have to consider. Therefore this argument—or rather assertion—goes for nothing.

The only other argument I have met with on this side of the question is one that has recently been adduced by Mr. Wallace. He says:—

"One very weighty objection to the theory that specific characters can ever be wholly useless appears to have been overlooked by those who have maintained the frequency of such characters, and that is, their almost necessary instability[94]."

This argument he proceeds to elaborate at considerable length, but fails to perceive what appears to me the obvious answer. Provided that the cause of the useless character is constant, there is no difficulty in understanding why the character is stable. Utility is not the only principle that can lead to stability: any other principle must do the same, provided that it acts for a sufficient length of time, and with a sufficient degree of uniformity, on all the individuals of a species. This is a consideration the cogency of which was clearly recognized by Darwin, as the following quotations will show. Speaking of unadaptive characters, he says they may arise as merely

"fluctuating variations, which sooner or later become constant through the nature of the organism and of surrounding conditions, but not through natural selection[95]."

Elsewhere we read:—

"Each of the endless variations which we see in the plumage of our fowls must have had some efficient cause; and if the same cause were to act uniformly during a long series of generations on many individuals, all probably would be modified in the same manner."

As special illustrations of this fact I may quote the following cases from Darwin's works.

"Dr. Bachman states that he has seen turkeys raised from the eggs of wild species, lose their metallic tints, and become spotted in the third generation. Mr. Yarrell many years ago informed me that the wild ducks bred in St James' Park lost their true plumage after a few generations. An excellent observer (Mr. Hewitt) ... found that he could not breed wild ducks true for more than five or six generations, as they proved so much less beautiful. The white collar round the neck of the mallard became broader and more irregular, and white feathers appeared in the duckling's wings &c.[96]"

Now, such cases—to which numberless others might be added—prove that even the subtle and inconspicuous causes incidental to domestication are capable of inducing changes of specific character quite as great, and quite as "stable," as any that in a state of nature are taken to constitute specific distinctions. Yet there can here be no suggestion of utility, inasmuch as the change takes place in the course of a few generations, and therefore without leaving time for natural selection to come into play—even if it ever could come into play among the sundry domesticated birds in question.

But the facts of domestication also make for the same conclusion in another way—namely, by proving that when time enough has been allowed for the production of useless changes of greater magnitude, such changes are not infrequently produced. And the value of this line of evidence is that, great as are the changes, it is impossible that either natural or artificial selection can have been concerned in their production. It will be sufficient to give two examples—both with regard to structure.

The first I will render in the words whereby it has already been stated in my own paper on Physiological Selection, because I should like to take this opportunity of answering Mr. Wallace's objection to it.

"Elsewhere (Origin of Species, p. 158) Mr. Darwin points out that modifications which appear to present obvious utility are often found on further examination to be really useless. This latter consideration, therefore, may be said to act as a foil to the one against which I am arguing, namely, that modifications which appear to be useless may nevertheless be useful. But here is a still more suggestive consideration, also derived from Mr. Darwin's writings. Among our domesticated productions changes of structure—or even structures wholly new—not unfrequently arise, which are in every way analogous to the apparently useless distinctions between wild species. Take, for example, the following most instructive case:—

Old Irish Pig, showing jaw-appendages Fig. 2.—Old Irish Pig, showing jaw-appendages (after Richardson).

"'Another curious anomaly is offered by the appendages described by M. Eudes-Deslongchamps as often characterizing the Normandy pigs. These appendages are always attached to the same spot, to the corners of the jaws; they are cylindrical, about three inches in length, covered with bristles, and with a pencil of bristles rising out of a sinus on one side; they have a cartilaginous centre with two small longitudinal muscles; they occur either symmetrically on both sides of the face, or on one side alone. Richardson figures them on the gaunt old Irish Greyhound pig; and Nathusius states that they occasionally appear in all the long-eared races, but are not strictly inherited, for they occur or fail in the animals of the same litter. As no wild pigs are known to have analogous appendages, we have at present no reason to suppose that their appearance is due to reversion; and if this be so, we are forced to admit that a somewhat complex, though apparently useless, structure may be suddenly developed without the aid of selection[97].'"

To this case Mr. Wallace objects:—

"But it is expressly stated that they are not constant; they appear 'frequently' or 'occasionally,' they are 'not strictly inherited, for they occur or fail in animals of the same litter'; and they are not always symmetrical, sometimes appearing on one side of the face alone. Now, whatever may be the cause or explanation of these anomalous appendages, they cannot be classed with 'specific characters,' the most essential features of which are, that they are symmetrical, that they are inherited, and that they are constant[98]."

But, to begin with, I have not classed these appendages with "specific characters," nor maintained that Normandy pigs ought to be regarded as specifically distinct on account of them. What I said was:—

"Now, if any such structure as this occurred in a wild species, and if any one were to ask what is the use of it, those who rely on the argument from ignorance would have a much stronger case than they usually have; for they might point to the cartilage supplied with muscles, and supporting a curious arrangement of bristles, as much too specialized a structure to be wholly meaningless. Yet we happen to know that this particular structure is wholly meaningless[99]."

In the next place, is it either fair or reasonable to expect that a varietal character of presumably very recent origin should be as strongly inherited—and therefore as constant both in occurrence and symmetry—as a true specific character, say, of a thousand times its age? Even characters of so-called "constant varieties" in a state of nature are usually less constant than specific characters; while, again, as Darwin says, "it is notorious that specific characters are more variable than generic,"—the reason in both cases being, as he proceeds to show, that the less constant characters are characters of more recent origin, and therefore less firmly fixed by heredity[100]. Hence I do not understand how Mr. Wallace can conclude, as he does, "that, admitting that this peculiar appendage is wholly useless and meaningless, the fact would be rather an argument against specific characters being also meaningless, because the latter never have the characteristics [i.e. inconstancy of occurrence, form, and transmission] which this particular variation possesses[101]." Mr. Wallace can scarcely suppose that when specific characters first arise, they present the three-fold kind of constancy to which he here alludes. But, if not, can it be denied that these peculiar appendages appear to be passing through a phase of development which all "specific characters" must have passed through, before they have had time enough to be firmly fixed by heredity[102]?

If, however, even this should be denied, what will be said of the second case, that of the niata cattle?

"I saw two herds on the northern bank of the Plata.... The forehead is very short and broad, with the nasal end of the skull, together with the whole plane of the upper molar-teeth, curved upwards. The lower jaw projects beyond the upper, and has a corresponding upward curvature.... The skull which I presented to the College of Surgeons has been thus described by Professor Owen. 'It is remarkable from the stunted development of the nasals, premaxillaries, and fore part of the lower jaw, which is unusually curved upwards to come into contact with the premaxillaries. The nasal bones are about one-third the ordinary length, but retain almost their normal breadth. The triangular vacuity is left between them and the frontal and lachrymal, which latter bone articulates with the premaxillary, and thus excludes the maxillary from any junction with the nasal.' So that even the connexion of some of the bones is changed. Other differences might be added: thus the plane of the condyles is somewhat modified, and the terminal edge of the premaxillaries forms an arch. In fact, on comparison with the skull of a common ox, scarcely a single bone presents the same exact shape, and the whole skull has a wonderfully different appearance[103]."