2. The sterility which so generally attends the crossing of two specific forms is to be distinguished as of two kinds, which, although often confounded by naturalists, are in reality quite distinct. For the sterility may obtain between the two parent species when first crossed, or it may first assert itself in their hybrid progeny. In the latter case the hybrids, although possibly produced without any appearance of infertility on the part of their parent species, nevertheless prove more or less infertile among themselves, and also with members of either parent species.

3. The degree of both kinds of infertility varies in the case of different species, and in that of their hybrid progeny, from absolute sterility up to complete fertility. Thus, to take the case of plants, “when pollen from a plant of one family is placed on the stigma of a plant of a distinct family, it exerts no more influence than so much inorganic dust. From this absolute zero of fertility, the pollen of different species, applied to the stigma of some one species of the same genus, yields a perfect gradation in the number of seeds produced, up to nearly complete, or even quite complete, fertility; so, in hybrids themselves, there are some which never have produced, and probably never would produce, even with the pollen of the pure parents, a single fertile seed; but in some of these cases a first trace of fertility may be detected, by the pollen of one of the pure parent species causing the flower of the hybrid to wither earlier than it otherwise would have done; and the early withering of the flower is well known to be a sign of incipient fertilization. From this extreme degree of sterility we have self-fertilized hybrids producing a greater and greater number of seeds up to perfect fertility.”

4. Although there is, as a rule, a certain parallelism, there is no fixed relation between the degree of sterility manifested by the parent species when crossed and that which is manifested by their hybrid progeny. There are many cases in which two pure species can be crossed with unusual facility, while the resulting hybrids are remarkably sterile; and, contrariwise, there are species which can only be crossed with extreme difficulty, though the hybrids, when produced, are very fertile. Even within the limits of the same genus, these two opposite cases may occur.

5. When two species are reciprocally crossed, i.e., male A with female B, and male B with female A, the degree of sterility often differs greatly in the two cases. The sterility of the resulting hybrids may differ likewise.

6. The degree of sterility of first crosses and of hybrids runs, to a certain extent, parallel with the systematic affinity of the forms which are united. “For species belonging to distinct genera can rarely, and those belonging to distinct families can never, be crossed. The parallelism, however, is far from complete; for a multitude of closely allied species will not unite, or unite with extreme difficulty, whilst other species, widely different from each other, can be crossed with perfect facility. Nor does the difficulty depend on ordinary constitutional differences; for annual and perennial plants, deciduous and evergreen trees, plants flowering at different seasons, inhabiting different stations, and naturally living under the most opposite climates, can often be crossed with ease. The difficulty or facility apparently depends exclusively on the sexual constitution of the species which are crossed, or on their sexual elective affinity.”

Such being the principal facts of hybridism, we may next consider the relation which they bear to the theory of descent. It is obvious that the most important point of contact between the former and the latter consists in this —that, although hybridism is occasionally possible as an exception to the general infertility of species inter se, it is only, as it were, a partial exception; for, even when produced, the hybrid progeny almost invariably manifest some greater or less degree of sterility, and this not only when crossed among themselves, but even when crossed with either of their parent species. The main facts of hybridism thus at first sight seem to support the time-honoured doctrine that there are placed between all species the barriers of mutual sterility, for the purpose of preventing any admixture of specific qualities by heredity, and so for the purpose of maintaining the immutability of specific types. And the apparent support which this doctrine thus receives from the main facts of hybridism is still further strengthened when these facts are contrasted with those which are supplied by the breeding of our domestic “varieties.” For, in the latter case, and as an almost invariable rule, neither the organisms when crossed nor their resulting progeny show any indications of sterility, although the two parent varieties may differ from one another even more widely than do many natural species which are wholly infertile when crossed. This very general distinction between natural species and domestic varieties has appeared to many competent persons in the present generation so profound and significant that they deem it to be in itself sufficient to discredit, if not to negative, the whole theory of the transmutation of species.

Now, when this distinction is thus posited as an objection to the theory of descent, we must first of all remember that this theory does not require the possibility of the commingling of specific types; it requires, indeed, that specific types should not be immutably fixed, but it does not require that the causes of their mutation should depend upon their mutual crossing. The whole difficulty, therefore, which the theory of descent has here to meet is to explιin why it is that natural species are fenced about, as it were, with the mysterious barriers of sterility, while no such seeming care appears to have been taken in the case of our domestic breeds—even though in the latter case artificial selection by the breeders may have produced more visible difference between the two parent races than that which natural selection is supposed to have produced between two natural species.

In answer to this difficulty, the most important consideration to begin with is one that is very generally lost sight of. The consideration is that mutual sterility between organic forms has been constituted by naturalists the chief criterion of specific distinction, and therefore it is merely to argue in a circle to maintain that specific distinction is of some transcendental nature because it is so invariably associated with this mutual sterility. If it were not for the fact of their mutual sterility, this and that species would probably not have been classified as such; and therefore it is now scarcely to be considered a matter of any great significance that all species present more or less of the peculiarity in virtue of which they are recognized as species. Or, otherwise stated, on the supposition that species have had a derivative origin, whenever the modification of a specific type has proceeded sufficiently far to induce sterility with allied types, the modified type is, for this reason, classified as a distinct species; otherwise, upon the supposition, the species could, scarcely have become separated out as a distinct type. The argument which points to such sterility as evidence against this supposition is, therefore, so far inconclusive.

The case, however, ought not in fairness to be stated quite so strongly as this; for mutual sterility, although the chief, is not the only criterion of specific distinction. In forming their classifications, naturalists endeavour as much as possible to have regard to organisms in the totality of their structures and functions. It may therefore still be maintained that, although the above consideration as to mutual sterility being selected as the chief criterion of specific distinction greatly mitigates the force of the argument that natural species differ from artificial breeds in being more or less sterile with one another, still this consideration does not altogether destroy that argument. For, on the one hand, it is not mutual sterility alone which is taken as a test of specific distinction; and, on the other hand, it generally so happens that the other qualities distinctive of any given species do not differ more widely from those which are distinctive of allied species than is the case with many of our domestic breeds. It therefore still remains a significant circumstance that, along with the differences distinctive of natural species, there almost invariably goes the protective attribute of mutual sterility; while the possibly greater differences distinctive of our domestic breeds are unaccompanied by any such protective attribute. But, again, this more refined objection can be met and satisfactorily excluded by the general consideration that “as species rarely or never become modified in one character, without being at the same time modified in many, and as systematic affinity includes all visible resemblances and dissimilarities, any difference in sexual constitution between two species would naturally stand in more or less close relation with their systematic positionr.”

But we are not confined to this general consideration alone. There are several other general considerations which tend still further to mitigate the difficulty, and there are several particular facts which together prove that the alleged distinction between natural species and domestic varieties is one, not of kind, but of degree. We shall, therefore, next proceed to state these general considerations and particular facts.

Upon the theory of descent, mutual sterility between specific types is nothing more than the expression of some certain amount of modification having taken place in the reproductive system of a changing form, which up to that time, and but for the fact of this modification, would have been classified by naturalists as a mere variety. Now the causes which act upon the reproductive system, both of animals and plants, and whether in the direction of sterility or prolificness, are at present hopelessly obscure. We cannot, therefore, expect to distinguish the causes which in the case of any given species have determined sterility. Nor is it necessary, for the meeting of the present difficulty, that this should be done; it is enough for this purpose to show that the causes which thus act upon the reproductive system are much too indistinct to admit of any argument being raised upon them. And this it is most easy to show; for it is not too much to say that the reproductive system is, generally speaking, of all parts of an organism the most delicately susceptible to slight changes in the conditions of life. Mr Darwin has adduced a vast array of facts on this head in his Variation of Animals and Plants under Domestication. As a result of this delicacy, there arises an apparent capriciousness in the ways and degrees in which the reproductive system is affected by slight changes in the conditions of life, by too close interbreeding, by grafting, and by many other causes. Thus, for example, the influences of domestication produce more or less sterility in numberless species of wild animals and plants, while in other species— and this, as we shall presently see, is a matter of great importance in the present connexion—such influences are favourable to fertility. Now if we suppose, as in consistency we must suppose, that throughout the course of evolution the reproductive system has always been characterized by a sensitiveness to slight changes similar to that which we now observe, and if we remember that, in any case where these slight changes were sufficient to cause mutual sterility between the modified descendants of a common progenitor, a distinction of species must necessarily have arisen,—we shall cease to regard the present steri i ity of species inter se as anything more than what might be expected a priori, supposing the theory of descent with gradual modification to be true.

As evidence of the apparent capriciousness with which sterility may be manifested, owing to the slight and imperceptible causes on which it depends, special allusion must here be made to a highly remarkable and significant fact that has been brought to light by the direct experiments of Mr Darwin. The following is his account of these experiments :—

“Several plants belonging to distinct orders present two forms, which exist in about equal numbers, and which differ in no respect except in their reproductive organs,—one form having a long pistil with short stamens, the other a short pistil with long stamens; both with differently-sized pollen grains. With trimorphic plants there are three forms likewise differing in the length of their pistils and stamens, in the size and colour of the pollen grains, and in some other respects; and, as in each of the three forms there are two sets of stamens, there are altogether six sets of stamens and three kinds of pistils. These organs are so proportioned in length to each other that, in any two of the forms, half the stamens in each stand on a level with the stigma of the third form. Now I have shown, and the result has been confirmed by other observers, that in order to obtain full fertility with other plants, it is necessary that the stigma of the one form should be fertilized by pollen taken from the stamens of corresponding height in the other form. So that with dimorphic species two unions, which may be called legitimate, are fully fertile, and two which may be called illegitimate are more or less infertile. With trimorphic species six unions are legitimate or fully fertile, and twelve are illegitimate or more or less infertile. The infertility which may be observed in various dimorphic and trimorphic plants when they are illegitimately fertilized, that is, by pollen taken from stamens not corresponding in height with the pistil, differs much in degree up to absolute and utter sterility,— just in the same manner as occurs in crossing distinct species.” In this case we appear to have actual evidence of different stages of increasing sterility in transitu, and this even within the limits of the same natural species. And if even such evidence as this can be resisted, there still remains one very important fact, which directly affects the whole alleged distinction between the sterility of natural species and the fertility of domestic breeds. This fact is that plants belonging to several species of the genus Passiflora have been amply proved, not only to be completely fertile with plants belonging to other species, but even to be as completely infertile with plants belonging to their own. Thus fruit could not be obtained from P. alata and P. racemosa except by reciprocally fertilizing them with each other’s pollen; and similar facts have been observed by several experimentalists with regard to four or five other species of this genus. The fullest details on the subject are those given by Mr Scott in the Journal of the Liηnean Society, vol. viii. p. 168. Plants belonging to three species of the genus, viz., P. racemosa, coerulea, and alata, flowered for many years in Edinburgh, but, though repeatedly fertilized by Mr Scott and others with their own pollen, never produced seed. But when mutually crossed in various ways they all produced seed. After quoting this case Mr Darwin adds :—

“Returning to Ρ. αlαtα, I have received (1866) some interesting details from Mr Robinson Munro. Three plants, including one in England, have already been mentioned which were inveterately self-sterile, and Mr Munro informs me of several others which, after repeated trials during many years, have been found in the same predicament. At some other places, however, this species fruits readily when fertilized with its own pollen. At Taymouth Castle there is a plant which was formerly grafted by Mr Donaldson on a distinct species, name unknown, and ever since the operation it has produced fruit in abundance by its own pollen, so that this small and unnatural .change in the state of this plant has restored its selffertility. Some of the seedlings from the Taymouth Castle plant were found to be not only sterile with their own pollen, but with each other’s pollen and with the pollen of distinct species. Pollen from the Taymouth plant failed to fertilize certain plants of the same species, but was successful on one plant in the Edinburgh Botanic Gardens. Seedlings were raised from this latter union, and some of their flowers were fertilized by Mr Munro with their own pollen; but they were found to be as self-impotent as the mother plant had always proved, except when fertilized by the grafted Taymouth plant, and except, as we shall see, when fertilized by her own seedlings. Yet Mr Munro fertilized eighteen flowers on the self-impotent mother plant with pollen from these her own self-impotent seedlings, and obtained, remarkable as the fact is, eighteen fine capsules full of excellent seed. I have met with no case in regard to plants which shows so well as this of P. alata on what small and mysterious causes complete fertility or complete sterility depends.”

These cases in the genus Passiflora, although so highly remarkable, are not wholly unique. There is not, indeed, any other case of a natural species, the members of which are only fertile with members of another species; but there are several cases of natural species, the members of which are self-impotent, though freely fertile either with other plants of the same species, or with plants of different though allied species. This may perhaps be regarded as a transitional stage between the ordinary condition of plants and the extraordinary condition that obtains among species of the genus Passiflora. It occurs in individual plants of certain species of Lobelia and Verbascum, and among several genera of orchids. The cases of the latter are particularly remarkable, inasmuch as Fritz Müller found from numerous experiments, not only that individual plants belonging to the several species were not fertilized by their own pollen, while freely so by pollen taken from distinct species, and even from distinct genera, but that the plant’s own pollen was positively deleterious to its stigma, and acted as a poison to the destruction of the flower.

So much,, then, for the facts which go to prove on what slight constitutional differences sterility may depend, and the consequent probability there is that it should generally be found to accompany a change of organization which is sufficiently great to be regarded by naturalists as a specific distinction. But the pleading must not end here. For there still remains to be adduced the fact, already mentioned as one of the general facts of hybridism, that “the degree of both kinds of fertility varies in the case of different species, and in that of their hybrid progeny, from absolute sterility up to complete fertility.” As a matter of fact, the distinction between natural species and domestic varieties, upon which the whole discussion has hitherto proceeded, is in itself untenable; the infertility of natural species when crossed, although without question the general rule, is nevertheless not the invariable rule. We need not point to the highly anomalous case of Passiflora recently mentioned in another connexion, and probably to be explained as the result of cultivation; for we appear to have sufficient evidence without it. It is true that a great deal of negative evidence has been published upon this point by very...