Distinction between the sterility of first crosses and of hybrids --Sterility various in degree, not universal, affected by closeinterbreeding, removed by domestication -- Laws governing the sterility ofhybrids -- Sterility not a special endowment, but incidental on otherdifferences, not accumulated by natural selection -- Causes of thesterility of first crosses and of hybrids -- Parallelism between theeffects of changed conditions of life and of crossing -- Dimorphism andtrimorphism -- Fertility of varieties when crossed and of their mongreloffspring not universal -- Hybrids and mongrels compared independently oftheir fertility -- Summary.

The view commonly entertained by naturalists is that species, whenintercrossed, have been specially endowed with sterility, in order toprevent their confusion. This view certainly seems at first highlyprobable, for species living together could hardly have been kept distincthad they been capable of freely crossing. The subject is in many waysimportant for us, more especially as the sterility of species when firstcrossed, and that of their hybrid offspring, cannot have been acquired, asI shall show, by the preservation of successive profitable degrees ofsterility. It is an incidental result of differences in the reproductivesystems of the parent-species.

In treating this subject, two classes of facts, to a large extentfundamentally different, have generally been confounded; namely, thesterility of species when first crossed, and the sterility of the hybridsproduced from them.

Pure species have of course their organs of reproduction in a perfectcondition, yet when intercrossed they produce either few or no offspring. Hybrids, on the other hand, have their reproductive organs functionallyimpotent, as may be clearly seen in the state of the male element in bothplants and animals; though the formative organs themselves are perfect instructure, as far as the microscope reveals. In the first case the twosexual elements which go to form the embryo are perfect; in the second casethey are either not at all developed, or are imperfectly developed. Thisdistinction is important, when the cause of the sterility, which is commonto the two cases, has to be considered. The distinction probably has beenslurred over, owing to the sterility in both cases being looked on as aspecial endowment, beyond the province of our reasoning powers.

The fertility of varieties, that is of the forms known or believed to bedescended from common parents, when crossed, and likewise the fertility oftheir mongrel offspring, is, with reference to my theory, of equalimportance with the sterility of species; for it seems to make a broad andclear distinction between varieties and species.

DEGREES OF STERILITY.

First, for the sterility of species when crossed and of their hybridoffspring. It is impossible to study the several memoirs and works ofthose two conscientious and admirable observers, Kolreuter and Gartner, whoalmost devoted their lives to this subject, without being deeply impressedwith the high generality of some degree of sterility. Kolreuter makes therule universal; but then he cuts the knot, for in ten cases in which hefound two forms, considered by most authors as distinct species, quitefertile together, he unhesitatingly ranks them as varieties. Gartner,also, makes the rule equally universal; and he disputes the entirefertility of Kolreuter's ten cases. But in these and in many other cases,Gartner is obliged carefully to count the seeds, in order to show thatthere is any degree of sterility. He always compares the maximum number ofseeds produced by two species when first crossed, and the maximum producedby their hybrid offspring, with the average number produced by both pureparent-species in a state of nature. But causes of serious error hereintervene: a plant, to be hybridised, must be castrated, and, what isoften more important, must be secluded in order to prevent pollen beingbrought to it by insects from other plants. Nearly all the plantsexperimented on by Gartner were potted, and were kept in a chamber in hishouse. That these processes are often injurious to the fertility of aplant cannot be doubted; for Gartner gives in his table about a score ofcases of plants which he castrated, and artificially fertilised with theirown pollen, and (excluding all cases such as the Leguminosae, in whichthere is an acknowledged difficulty in the manipulation) half of thesetwenty plants had their fertility in some degree impaired. Moreover, asGartner repeatedly crossed some forms, such as the common red and bluepimpernels (Anagallis arvensis and coerulea), which the best botanists rankas varieties, and found them absolutely sterile, we may doubt whether manyspecies are really so sterile, when intercrossed, as he believed.

It is certain, on the one hand, that the sterility of various species whencrossed is so different in degree and graduates away so insensibly, and, onthe other hand, that the fertility of pure species is so easily affected byvarious circumstances, that for all practical purposes it is most difficultto say where perfect fertility ends and sterility begins. I think nobetter evidence of this can be required than that the two most experiencedobservers who have ever lived, namely Kolreuter and Gartner, arrived atdiametrically opposite conclusions in regard to some of the very sameforms. It is also most instructive to compare--but I have not space hereto enter on details--the evidence advanced by our best botanists on thequestion whether certain doubtful forms should be ranked as species orvarieties, with the evidence from fertility adduced by differenthybridisers, or by the same observer from experiments made during differentyears. It can thus be shown that neither sterility nor fertility affordsany certain distinction between species and varieties. The evidence fromthis source graduates away, and is doubtful in the same degree as is theevidence derived from other constitutional and structural differences.

In regard to the sterility of hybrids in successive generations; thoughGartner was enabled to rear some hybrids, carefully guarding them from across with either pure parent, for six or seven, and in one case for tengenerations, yet he asserts positively that their fertility neverincreases, but generally decreases greatly and suddenly. With respect tothis decrease, it may first be noticed that when any deviation in structureor constitution is common to both parents, this is often transmitted in anaugmented degree to the offspring; and both sexual elements in hybridplants are already affected in some degree. But I believe that theirfertility has been diminished in nearly all these cases by an independentcause, namely, by too close interbreeding. I have made so many experimentsand collected so many facts, showing on the one hand that an occasionalcross with a distinct individual or variety increases the vigour andfertility of the offspring, and on the other hand that very closeinterbreeding lessens their vigour and fertility, that I cannot doubt thecorrectness of this conclusion. Hybrids are seldom raised byexperimentalists in great numbers; and as the parent-species, or otherallied hybrids, generally grow in the same garden, the visits of insectsmust be carefully prevented during the flowering season: hence hybrids, ifleft to themselves, will generally be fertilised during each generation bypollen from the same flower; and this would probably be injurious to theirfertility, already lessened by their hybrid origin. I am strengthened inthis conviction by a remarkable statement repeatedly made by Gartner,namely, that if even the less fertile hybrids be artificially fertilisedwith hybrid pollen of the same kind, their fertility, notwithstanding thefrequent ill effects from manipulation, sometimes decidedly increases, andgoes on increasing. Now, in the process of artificial fertilisation,pollen is as often taken by chance (as I know from my own experience) fromthe anthers of another flower, as from the anthers of the flower itselfwhich is to be fertilised; so that a cross between two flowers, thoughprobably often on the same plant, would be thus effected. Moreover,whenever complicated experiments are in progress, so careful an observer asGartner would have castrated his hybrids, and this would have insured ineach generation a cross with pollen from a distinct flower, either from thesame plant or from another plant of the same hybrid nature. And thus, thestrange fact of an increase of fertility in the successive generations ofARTIFICIALLY FERTILISED hybrids, in contrast with those spontaneously self-fertilised, may, as I believe, be accounted for by too close interbreedinghaving been avoided.

Now let us turn to the results arrived at by a third most experiencedhybridiser, namely, the Hon. and Rev. W. Herbert. He is as emphatic in hisconclusion that some hybrids are perfectly fertile--as fertile as the pureparent-species--as are Kolreuter and Gartner that some degree of sterilitybetween distinct species is a universal law of nature. He experimented onsome of the very same species as did Gartner. The difference in theirresults may, I think, be in part accounted for by Herbert's greathorticultural skill, and by his having hot-houses at his command. Of hismany important statements I will here give only a single one as an example,namely, that "every ovule in a pod of Crinum capense fertilised by C.revolutum produced a plant, which I never saw to occur in a case of itsnatural fecundation." So that here we have perfect, or even more thancommonly perfect fertility, in a first cross between two distinct species.

This case of the Crinum leads me to refer to a singular fact, namely, thatindividual plants of certain species of Lobelia, Verbascum and Passiflora,can easily be fertilised by the pollen from a distinct species, but not bypollen from the same plant, though this pollen can be proved to beperfectly sound by fertilising other plants or species. In the genusHippeastrum, in Corydalis as shown by Professor Hildebrand, in variousorchids as shown by Mr. Scott and Fritz Muller, all the individuals are inthis peculiar condition. So that with some species, certain abnormalindividuals, and in other species all the individuals, can actually behybridised much more readily than they can be fertilised by pollen from thesame individual plant! To give one instance, a bulb of Hippeastrum aulicumproduced four flowers; three were fertilised by Herbert with their ownpollen, and the fourth was subsequently fertilised by the pollen of acompound hybrid descended from three distinct species: the result was that"the ovaries of the three first flowers soon ceased to grow, and after afew days perished entirely, whereas the pod impregnated by the pollen ofthe hybrid made vigorous growth and rapid progress to maturity, and boregood seed, which vegetated freely." Mr. Herbert tried similar experimentsduring many years, and always with the same result. These cases serve toshow on what slight and mysterious causes the lesser or greater fertilityof a species sometimes depends.

The practical experiments of horticulturists, though not made withscientific precision, deserve some notice. It is notorious in howcomplicated a manner the species of Pelargonium, Fuchsia, Calceolaria,Petunia, Rhododendron, etc., have been crossed, yet many of these hybridsseed freely. For instance, Herbert asserts that a hybrid from Calceolariaintegrifolia and plantaginea, species most widely dissimilar in generalhabit, "reproduces itself as perfectly as if it had been a natural speciesfrom the mountains of Chile." I have taken some pains to ascertain thedegree of fertility of some of the complex crosses of Rhododendrons, and Iam assured that many of them are perfectly fertile. Mr. C. Noble, forinstance, informs me that he raises stocks for grafting from a hybridbetween Rhod. ponticum and catawbiense, and that this hybrid "seeds asfreely as it is possible to imagine." Had hybrids, when fairly treated,always gone on decreasing in fertility in each successive generation, asGartner believed to be the case, the fact would have been notorious tonurserymen. Horticulturists raise large beds of the same hybrid, and suchalone are fairly treated, for by insect agency the several individuals areallowed to cross freely with each other, and the injurious influence ofclose interbreeding is thus prevented. Any one may readily convincehimself of the efficiency of insect agency by examining the flowers of themore sterile kinds of hybrid Rhododendrons, which produce no pollen, for hewill find on their stigmas plenty of pollen brought from other flowers.

In regard to animals, much fewer experiments have been carefully tried thanwith plants. If our systematic arrangements can be trusted, that is, ifthe genera of animals are as distinct from each other as are the genera ofplants, then we may infer that animals more widely distinct in the scale ofnature can be crossed more easily than in the case of plants; but thehybrids themselves are, I think, more sterile. It should, however, beborne in mind that, owing to few animals breeding freely under confinement,few experiments have been fairly tried: for instance, the canary-bird hasbeen crossed with nine distinct species of finches, but, as not one ofthese breeds freely in confinement, we have no right to expect that thefirst crosses between them and the canary, or that their hybrids, should beperfectly fertile. Again, with respect to the fertility in successivegenerations of the more fertile hybrid animals, I hardly know of aninstance in which two families of the same hybrid have been raised at thesame time from different parents, so as to avoid the ill effects of closeinterbreeding. On the contrary, brothers and sisters have usually beencrossed in each successive generation, in opposition to the constantlyrepeated admonition of every breeder. And in this case, it is not at allsurprising that the inherent sterility in the hybrids should have gone onincreasing.

Although I know of hardly any thoroughly well-authenticated cases ofperfectly fertile hybrid animals, I have reason to believe that the hybridsfrom Cervulus vaginalis and Reevesii, and from Phasianus colchicus with P.torquatus, are perfectly fertile. M. Quatrefages states that the hybridsfrom two moths (Bombyx cynthia and arrindia) were proved in Paris to befertile inter se for eight generations. It has lately been asserted thattwo such distinct species as the hare and rabbit, when they can be got tobreed together, produce offspring, which are highly fertile when crossedwith one of the parent-species. The hybrids from the common and Chinesegeese (A. cygnoides), species which are so different that they aregenerally ranked in distinct genera, have often bred in this country witheither pure parent, and in one single instance they have bred inter se. This was effected by Mr. Eyton, who raised two hybrids from the sameparents, but from different hatches; and from these two birds he raised noless than eight hybrids (grandchildren of the pure geese) from one nest. In India, however, these cross-bred geese must be far more fertile; for Iam assured by two eminently capable judges, namely Mr. Blyth and CaptainHutton, that whole flocks of these crossed geese are kept in various partsof the country; and as they are kept for profit, where neither pureparent-species exists, they must certainly be highly or perfectly fertile.

With our domesticated animals, the various races when crossed together arequite fertile; yet in many cases they are descended from two or more wildspecies. From this fact we must conclude either that the aboriginalparent-species at first produced perfectly fertile hybrids, or that thehybrids subsequently reared under domestication became quite fertile. Thislatter alternative, which was first propounded by Pallas, seems by far themost probable, and can, indeed, hardly be doubted. It is, for instance,almost certain that our dogs are descended from several wild stocks; yet,with perhaps the exception of certain indigenous domestic dogs of SouthAmerica, all are quite fertile together; but analogy makes me greatlydoubt, whether the several aboriginal species would at first have freelybred together and have produced quite fertile hybrids. So again I havelately acquired decisive evidence that the crossed offspring from theIndian humped and common cattle are inter se perfectly fertile; and fromthe observations by Rutimeyer on their important osteological differences,as well as from those by Mr. Blyth on their differences in habits, voice,constitution, etc., these two forms must be regarded as good and distinctspecies. The same remarks may be extended to the two chief races of thepig. We must, therefore, either give up the belief of the universalsterility of species when crossed; or we must look at this sterility inanimals, not as an indelible characteristic, but as one capable of beingremoved by domestication.

Finally, considering all the ascertained facts on the intercrossing ofplants and animals, it may be concluded that some degree of sterility, bothin first crosses and in hybrids, is an extremely general result; but thatit cannot, under our present state of knowledge, be considered asabsolutely universal.

LAWS GOVERNING THE STERILITY OF FIRST CROSSES AND OF HYBRIDS.

We will now consider a little more in detail the laws governing thesterility of first crosses and of hybrids. Our chief object will be to seewhether or not these laws indicate that species have been specially endowedwith this quality, in order to prevent their crossing and blending togetherin utter confusion. The following conclusions are drawn up chiefly fromGartner's admirable work on the hybridisation of plants. I have taken muchpains to ascertain how far they apply to animals, and, considering howscanty our knowledge is in regard to hybrid animals, I have been surprisedto find how generally the same rules apply to both kingdoms.

It has been already remarked, that the degree of fertility, both of firstcrosses and of hybrids, graduates from zero to perfect fertility. It issurprising in how many curious ways this gradation can be shown; but onlythe barest outline of the facts can here be given. When pollen from aplant of one family is placed on the stigma of a plant of a distinctfamily, it exerts no more influence than so much inorganic dust. From thisabsolute zero of fertility, the pollen of different species applied to thestigma of some one species of the same genus, yields a perfect gradation inthe number of seeds produced, up to nearly complete or even quite completefertility; and, as we have seen, in certain abnormal cases, even to anexcess of fertility, beyond that which the plant's own pollen produces. Soin hybrids themselves, there are some which never have produced, andprobably never would produce, even with the pollen of the pure parents, asingle fertile seed: but in some of these cases a first trace of fertilitymay be detected, by the pollen of one of the pure parent-species causingthe flower of the hybrid to wither earlier than it otherwise would havedone; and the early withering of the flower is well known to be a sign ofincipient fertilisation. From this extreme degree of sterility we haveself-fertilised hybrids producing a greater and greater number of seeds upto perfect fertility.

The hybrids raised from two species which are very difficult to cross, andwhich rarely produce any offspring, are generally very sterile; but theparallelism between the difficulty of making a first cross, and thesterility of the hybrids thus produced--two classes of facts which aregenerally confounded together--is by no means strict. There are manycases, in which two pure species, as in the genus Verbascum, can be unitedwith unusual facility, and produce numerous hybrid offspring, yet thesehybrids are remarkably sterile. On the other hand, there are species whichcan be crossed very rarely, or with extreme difficulty, but the hybrids,when at last produced, are very fertile. Even within the limits of thesame genus, for instance in Dianthus, these two opposite cases occur.

The fertility, both of first crosses and of hybrids, is more easilyaffected by unfavourable conditions, than is that of pure species. But thefertility of first crosses is likewise innately variable; for it is notalways the same in degree when the same two species are crossed under thesame circumstances; it depends in part upon the constitution of theindividuals which happen to have been chosen for the experiment. So it iswith hybrids, for their degree of fertility is often found to differgreatly in the several individuals raised from seed out of the same capsuleand exposed to the same conditions.

By the term systematic affinity is meant, the general resemblance betweenspecies in structure and constitution. Now the fertility of first crosses,and of the hybrids produced from them, is largely governed by theirsystematic affinity. This is clearly shown by hybrids never having beenraised between species ranked by systematists in distinct families; and onthe other hand, by very closely allied species generally uniting withfacility. But the correspondence between systematic affinity and thefacility of crossing is by no means strict. A multitude of cases could begiven of very closely allied species which will not unite, or only withextreme difficulty; and on the other hand of very distinct species whichunite with the utmost facility. In the same family there may be a genus,as Dianthus, in which very many species can most readily be crossed; andanother genus, as Silene, in which the most persevering efforts have failedto produce between extremely close species a single hybrid. Even withinthe limits of the same genus, we meet with this same difference; forinstance, the many species of Nicotiana have been more largely crossed thanthe species of almost any other genus; but Gartner found that N. acuminata,which is not a particularly distinct species, obstinately failed tofertilise, or to be fertilised, by no less than eight other species ofNicotiana. Many analogous facts could be given.

No one has been able to point out what kind or what amount of difference,in any recognisable character, is sufficient to prevent two speciescrossing. It can be shown that plants most widely different in habit andgeneral appearance, and having strongly marked differences in every part ofthe flower, even in the pollen, in the fruit, and in the cotyledons, can becrossed. Annual and perennial plants, deciduous and evergreen trees,plants inhabiting different stations and fitted for extremely differentclimates, can often be crossed with ease.

By a reciprocal cross between two species, I mean the case, for instance,of a female-ass being first crossed by a stallion, and then a mare by amale-ass: these two species may then be said to have been reciprocallycrossed. There is often the widest possible difference in the facility ofmaking reciprocal crosses. Such cases are highly important, for they provethat the capacity in any two species to cross is often completelyindependent of their systematic affinity, that is of any difference intheir structure or constitution, excepting in their reproductive systems. The diversity of the result in reciprocal crosses between the same twospecies was long ago observed by Kolreuter. To give an instance: Mirabilis jalapa can easily be fertilised by the pollen of M. longiflora,and the hybrids thus produced are sufficiently fertile; but Kolreuter triedmore than two hundred times, during eight following years, to fertilisereciprocally M. longiflora with the pollen of M. jalapa, and utterlyfailed. Several other equally striking cases could be given. Thuret hasobserved the same fact with certain sea-weeds or Fuci. Gartner, moreover,found that this difference of facility in making reciprocal crosses isextremely common in a lesser degree. He has observed it even betweenclosely related forms (as Matthiola annua and glabra) which many botanistsrank only as varieties. It is also a remarkable fact that hybrids raisedfrom reciprocal crosses, though of course compounded of the very same twospecies, the one species having first been used as the father and then asthe mother, though they rarely differ in external characters, yet generallydiffer in fertility in a small, and occasionally in a high degree.

Several other singular rules could be given from Gartner: for instance,some species have a remarkable power of crossing with other species; otherspecies of the same genus have a remarkable power of impressing theirlikeness on their hybrid offspring; but these two powers do not at allnecessarily go together. There are certain hybrids which, instead ofhaving, as is usual, an intermediate character between their two parents,always closely resemble one of them; and such hybrids, though externally solike one of their pure parent-species, are with rare exceptions extremelysterile. So again among hybrids which are usually intermediate instructure between their parents, exceptional and abnormal individualssometimes are born, which closely resemble one of their pure parents; andthese hybrids are almost always utterly sterile, even when the otherhybrids raised from seed from the same capsule have a considerable degreeof fertility. These facts show how completely the fertility of a hybridmay be independent of its external resemblance to either pure parent.

Considering the several rules now given, which govern the fertility offirst crosses and of hybrids, we see that when forms, which must beconsidered as good and distinct species, are united, their fertilitygraduates from zero to perfect fertility, or even to fertility undercertain conditions in excess; that their fertility, besides being eminentlysusceptible to favourable and unfavourable conditions, is innatelyvariable; that it is by no means always the same in degree in the firstcross and in the hybrids produced from this cross; that the fertility ofhybrids is not related to the degree in which they resemble in externalappearance either parent; and lastly, that the facility of making a firstcross between any two species is not always governed by their systematicaffinity or degree of resemblance to each other. This latter statement isclearly proved by the difference in the result of reciprocal crossesbetween the same two species, for, according as the one species or theother is used as the father or the mother, there is generally somedifference, and occasionally the widest possible difference, in thefacility of effecting an union. The hybrids, moreover, produced fromreciprocal crosses often differ in fertility.

Now do these complex and singular rules indicate that species have beenendowed with sterility simply to prevent their becoming confounded innature? I think not. For why should the sterility be so extremelydifferent in degree, when various species are crossed, all of which we mustsuppose it would be equally important to keep from blending together? Whyshould the degree of sterility be innately variable in the individuals ofthe same species? Why should some species cross with facility and yetproduce very sterile hybrids; and other species cross with extremedifficulty, and yet produce fairly fertile hybrids? Why should there oftenbe so great a difference in the result of a reciprocal cross between thesame two species? Why, it may even be asked, has the production of hybridsbeen permitted? To grant to species the special power of producinghybrids, and then to stop their further propagation by different degrees ofsterility, not strictly related to the facility of the first union betweentheir parents, seems a strange arrangement.

The foregoing rules and facts, on the other hand, appear to me clearly toindicate that the sterility, both of first crosses and of hybrids, issimply incidental or dependent on unknown differences in their reproductivesystems; the differences being of so peculiar and limited a nature, that,in reciprocal crosses between the same two species, the male sexual elementof the one will often freely act on the female sexual element of the other,but not in a reversed direction. It will be advisable to explain a littlemore fully, by an example, what I mean by sterility being incidental onother differences, and not a specially endowed quality. As the capacity ofone plant to be grafted or budded on another is unimportant for theirwelfare in a state of nature, I presume that no one will suppose that thiscapacity is a SPECIALLY endowed quality, but will admit that it isincidental on differences in the laws of growth of the two plants. We cansometimes see the reason why one tree will not take on another fromdifferences in their rate of growth, in the hardness of their wood, in theperiod of the flow or nature of their sap, etc.; but in a multitude ofcases we can assign no reason whatever. Great diversity in the size of twoplants, one being woody and the other herbaceous, one being evergreen andthe other deciduous, and adaptation to widely different climates, does notalways prevent the two grafting together. As in hybridisation, so withgrafting, the capacity is limited by systematic affinity, for no one hasbeen able to graft together trees belonging to quite distinct families;and, on the other hand, closely allied species and varieties of the samespecies, can usually, but not invariably, be grafted with ease. But thiscapacity, as in hybridisation, is by no means absolutely governed bysystematic affinity. Although many distinct genera within the same familyhave been grafted together, in other cases species of the same genus willnot take on each other. The pear can be grafted far more readily on thequince, which is ranked as a distinct genus, than on the apple, which is amember of the same genus. Even different varieties of the pear take withdifferent degrees of facility on the quince; so do different varieties ofthe apricot and peach on certain varieties of the plum.

As Gartner found that there was sometimes an innate difference in differentINDIVIDUALS of the same two species in crossing; so Sagaret believes thisto be the case with different individuals of the same two species in beinggrafted together. As in reciprocal crosses, the facility of effecting anunion is often very far from equal, so it sometimes is in grafting. Thecommon gooseberry, for instance, cannot be grafted on the currant, whereasthe currant will take, though with difficulty, on the gooseberry.

We have seen that the sterility of hybrids which have their reproductiveorgans in an imperfect condition, is a different case from the difficultyof uniting two pure species, which have their reproductive organs perfect;yet these two distinct classes of cases run to a large extent parallel. Something analogous occurs in grafting; for Thouin found that three speciesof Robinia, which seeded freely on their own roots, and which could begrafted with no great difficulty on a fourth species, when thus graftedwere rendered barren. On the other hand, certain species of Sorbus, whengrafted on other species, yielded twice as much fruit as when on their ownroots. We are reminded by this latter fact of the extraordinary cases ofHippeastrum, Passiflora, etc., which seed much more freely when fertilisedwith the pollen of a distinct species than when fertilised with pollen fromthe same plant.

We thus see that, although there is a clear and great difference betweenthe mere adhesion of grafted stocks and the union of the male and femaleelements in the act of reproduction, yet that there is a rude degree ofparallelism in the results of grafting and of crossing distinct species. And as we must look at the curious and complex laws governing the facilitywith which trees can be grafted on each other as incidental on unknowndifferences in their vegetative systems, so I believe that the still morecomplex laws governing the facility of first crosses are incidental onunknown differences in their reproductive systems. These differences inboth cases follow, to a certain extent, as might have been expected,systematic affinity, by which term every kind of resemblance anddissimilarity between organic beings is attempted to be expressed. Thefacts by no means seem to indicate that the greater or lesser difficulty ofeither grafting or crossing various species has been a special endowment;although in the case of crossing, the difficulty is as important for theendurance and stability of specific forms as in the case of grafting it isunimportant for their welfare.

ORIGIN AND CAUSES OF THE STERILITY OF FIRST CROSSES AND OF HYBRIDS.

At one time it appeared to me probable, as it has to others, that thesterility of first crosses and of hybrids might have been slowly acquiredthrough the natural selection of slightly lessened degrees of fertility,which, like any other variation, spontaneously appeared in certainindividuals of one variety when crossed with those of another variety. Forit would clearly be advantageous to two varieties or incipient species ifthey could be kept from blending, on the same principle that, when man isselecting at the same time two varieties, it is necessary that he shouldkeep them separate. In the first place, it may be remarked that speciesinhabiting distinct regions are often sterile when crossed; now it couldclearly have been of no advantage to such separated species to have beenrendered mutually sterile, and consequently this could not have beeneffected through natural selection; but it may perhaps be argued, that, ifa species was rendered sterile with some one compatriot, sterility withother species would follow as a necessary contingency. In the secondplace, it is almost as much opposed to the theory of natural selection asto that of special creation, that in reciprocal crosses the male element ofone form should have been rendered utterly impotent on a second form, whileat the same time the male element of this second form is enabled freely tofertilise the first form; for this peculiar state of the reproductivesystem could hardly have been advantageous to either species.

In considering the probability of natural selection having come intoaction, in rendering species mutually sterile, the greatest difficulty willbe found to lie in the existence of many graduated steps, from slightlylessened fertility to absolute sterility. It may be admitted that it wouldprofit an incipient species, if it were rendered in some slight degreesterile when crossed with its parent form or with some other variety; forthus fewer bastardised and deteriorated offspring would be produced tocommingle their blood with the new species in process of formation. But hewho will take the trouble to reflect on the steps by which this firstdegree of sterility could be increased through natural selection to thathigh degree which is common with so many species, and which is universalwith species which have been differentiated to a generic or family rank,will find the subject extraordinarily complex. After mature reflection, itseems to me that this could not have been effected through naturalselection. Take the case of any two species which, when crossed, producedfew and sterile offspring; now, what is there which could favour thesurvival of those individuals which happened to be endowed in a slightlyhigher degree with mutual infertility, and which thus approached by onesmall step towards absolute sterility? Yet an advance of this kind, if thetheory of natural selection be brought to bear, must have incessantlyoccurred with many species, for a multitude are mutually quite barren. With sterile neuter insects we have reason to believe that modifications intheir structure and fertility have been slowly accumulated by naturalselection, from an advantage having been thus indirectly given to thecommunity to which they belonged over other communities of the samespecies; but an individual animal not belonging to a social community, ifrendered slightly sterile when crossed with some other variety, would notthus itself gain any advantage or indirectly give any advantage to theother individuals of the same variety, thus leading to their preservation.

But it would be superfluous to discuss this question in detail: for withplants we have conclusive evidence that the sterility of crossed speciesmust be due to some principle, quite independent of natural selection. Both Gartner and Kolreuter have proved that in genera including numerousspecies, a series can be formed from species which when crossed yield fewerand fewer seeds, to species which never produce a single seed, but yet areaffected by the pollen of certain other species, for the germen swells. Itis here manifestly impossible to select the more sterile individuals, whichhave already ceased to yield seeds; so that this acme of sterility, whenthe germen alone is effected, cannot have been gained through selection;and from the laws governing the various grades of sterility being souniform throughout the animal and vegetable kingdoms, we may infer that thecause, whatever it may be, is the same or nearly the same in all cases.

We will now look a little closer at the probable nature of the differencesbetween species which induce sterility in first crosses and in hybrids. Inthe case of first crosses, the greater or less difficulty in effecting aunion and in obtaining offspring apparently depends on several distinctcauses. There must sometimes be a physical impossibility in the maleelement reaching the ovule, as would be the case with a plant having apistil too long for the pollen-tubes to reach the ovarium. It has alsobeen observed that when the pollen of one species is placed on the stigmaof a distantly allied species, though the pollen-tubes protrude, they donot penetrate the stigmatic surface. Again, the male element may reach thefemale element, but be incapable of causing an embryo to be developed, asseems to have been the case with some of Thuret's experiments on Fuci. Noexplanation can be given of these facts, any more than why certain treescannot be grafted on others. Lastly, an embryo may be developed, and thenperish at an early period. This latter alternative has not beensufficiently attended to; but I believe, from observations communicated tome by Mr. Hewitt, who has had great experience in hybridising pheasants andfowls, that the early death of the embryo is a very frequent cause ofsterility in first crosses. Mr. Salter has recently given the results ofan examination of about 500 eggs produced from various crosses betweenthree species of Gallus and their hybrids; the majority of these eggs hadbeen fertilised; and in the majority of the fertilised eggs, the embryoshad either been partially developed and had then perished, or had becomenearly mature, but the young chickens had been unable to break through theshell. Of the chickens which were born, more than four-fifths died withinthe first few days, or at latest weeks, "without any obvious cause,apparently from mere inability to live;" so that from the 500 eggs onlytwelve chickens were reared. With plants, hybridized embryos probablyoften perish in a like manner; at least it is known that hybrids raisedfrom very distinct species are sometimes weak and dwarfed, and perish at anearly age; of which fact Max Wichura has recently given some striking caseswith hybrid willows. It may be here worth noticing that in some cases ofparthenogenesis, the embryos within the eggs of silk moths which had notbeen fertilised, pass through their early stages of development and thenperish like the embryos produced by a cross between distinct species. Until becoming acquainted with these facts, I was unwilling to believe inthe frequent early death of hybrid embryos; for hybrids, when once born,are generally healthy and long-lived, as we see in the case of the commonmule. Hybrids, however, are differently circumstanced before and afterbirth: when born and living in a country where their two parents live,they are generally placed under suitable conditions of life. But a hybridpartakes of only half of the nature and constitution of its mother; it maytherefore, before birth, as long as it is nourished within its mother'swomb, or within the egg or seed produced by the mother, be exposed toconditions in some degree unsuitable, and consequently be liable to perishat an early period; more especially as all very young beings are eminentlysensitive to injurious or unnatural conditions of life. But after all, thecause more probably lies in some imperfection in the original act ofimpregnation, causing the embryo to be imperfectly developed, rather thanin the conditions to which it is subsequently exposed.

In regard to the sterility of hybrids, in which the sexual elements areimperfectly developed, the case is somewhat different. I have more thanonce alluded to a large body of facts showing that, when animals and plantsare removed from their natural conditions, they are extremely liable tohave their reproductive systems seriously affected. This, in fact, is thegreat bar to the domestication of animals. Between the sterility thussuperinduced and that of hybrids, there are many points of similarity. Inboth cases the sterility is independent of general health, and is oftenaccompanied by excess of size or great luxuriance. In both cases thesterility occurs in various degrees; in both, the male element is the mostliable to be affected; but sometimes the female more than the male. Inboth, the tendency goes to a certain extent with systematic affinity, forwhole groups of animals and plants are rendered impotent by the sameunnatural conditions; and whole groups of species tend to produce sterilehybrids. On the other hand, one species in a group will sometimes resistgreat changes of conditions with unimpaired fertility; and certain speciesin a group will produce unusually fertile hybrids. No one can tell till hetries, whether any particular animal will breed under confinement, or anyexotic plant seed freely under culture; nor can he tell till he tries,whether any two species of a genus will produce more or less sterilehybrids. Lastly, when organic beings are placed during several generationsunder conditions not natural to them, they are extremely liable to vary,which seems to be partly due to their reproductive systems having beenspecially affected, though in a lesser degree than when sterility ensues. So it is with hybrids, for their offspring in successive generations areeminently liable to vary, as every experimentalist has observed.

Thus we see that when organic beings are placed under new and unnaturalconditions, and when hybrids are produced by the unnatural crossing of twospecies, the reproductive system, independently of the general state ofhealth, is affected in a very similar manner. In the one case, theconditions of life have been disturbed, though often in so slight a degreeas to be inappreciable by us; in the other case, or that of hybrids, theexternal conditions have remained the same, but the organisation has beendisturbed by two distinct structures and constitutions, including of coursethe reproductive systems, having been blended into one. For it is scarcelypossible that two organisations should be compounded into one, without somedisturbance occurring in the development, or periodical action, or mutualrelations of the different parts and organs one to another or to theconditions of life. When hybrids are able to breed inter se, they transmitto their offspring from generation to generation the same compoundedorganisation, and hence we need not be surprised that their sterility,though in some degree variable, does not diminish; it is even apt toincrease, this being generally the result, as before explained, of tooclose interbreeding. The above view of the sterility of hybrids beingcaused by two constitutions being compounded into one has been stronglymaintained by Max Wichura.

It must, however, be owned that we cannot understand, on the above or anyother view, several facts with respect to the sterility of hybrids; forinstance, the unequal fertility of hybrids produced from reciprocalcrosses; or the increased sterility in those hybrids which occasionally andexceptionally resemble closely either pure parent. Nor do I pretend thatthe foregoing remarks go to the root of the matter: no explanation isoffered why an organism, when placed under unnatural conditions, isrendered sterile. All that I have attempted to show is, that in two cases,in some respects allied, sterility is the common result--in the one casefrom the conditions of life having been disturbed, in the other case fromthe organisation having been disturbed by two organisations beingcompounded into one.

A similar parallelism holds good with an allied yet very different class offacts. It is an old and almost universal belief, founded on a considerablebody of evidence, which I have elsewhere given, that slight changes in theconditions of life are beneficial to all living things. We see this actedon by farmers and gardeners in their frequent exchanges of seed, tubers,etc., from one soil or climate to another, and back again. During theconvalescence of animals, great benefit is derived from almost any changein their habits of life. Again, both with plants and animals, there is theclearest evidence that a cross between individuals of the same species,which differ to a certain extent, gives vigour and fertility to theoffspring; and that close interbreeding continued during severalgenerations between the nearest relations, if these be kept under the sameconditions of life, almost always leads to decreased size, weakness, orsterility.

Hence it seems that, on the one hand, slight changes in the conditions oflife benefit all organic beings, and on the other hand, that slightcrosses, that is, crosses between the males and females of the samespecies, which have been subjected to slightly different conditions, orwhich have slightly varied, give vigour and fertility to the offspring. But, as we have seen, organic beings long habituated to certain uniformconditions under a state of nature, when subjected, as under confinement,to a considerable change in their conditions, very frequently are renderedmore or less sterile; and we know that a cross between two forms that havebecome widely or specifically different, produce hybrids which are almostalways in some degree sterile. I am fully persuaded that this doubleparallelism is by no means an accident or an illusion. He who is able toexplain why the elephant, and a multitude of other animals, are incapableof breeding when kept under only partial confinement in their nativecountry, will be able to explain the primary cause of hybrids being sogenerally sterile. He will at the same time be able to explain how it isthat the races of some of our domesticated animals, which have often beensubjected to new and not uniform conditions, are quite fertile together,although they are descended from distinct species, which would probablyhave been sterile if aboriginally crossed. The above two parallel seriesof facts seem to be connected together by some common but unknown bond,which is essentially related to the principle of life; this principle,according to Mr. Herbert Spencer, being that life depends on, or consistsin, the incessant action and reaction of various forces, which, asthroughout nature, are always tending towards an equilibrium; and when thistendency is slightly disturbed by any change, the vital forces gain inpower.

RECIPROCAL DIMORPHISM AND TRIMORPHISM.

This subject may be here briefly discussed, and will be found to throw somelight on hybridism. Several plants belonging to distinct orders presenttwo forms, which exist in about equal numbers and which differ in norespect except in their reproductive organs; one form having a long pistilwith short stamens, the other a short pistil with long stamens; the twohaving differently sized pollen-grains. With trimorphic plants there arethree forms likewise differing in the lengths 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, the threeforms possess altogether six sets of stamens and three kinds of pistils. These organs are so proportioned in length to each other that half thestamens in two of the forms stand on a level with the stigma of the thirdform. Now I have shown, and the result has been confirmed by otherobservers, that in order to obtain full fertility with these plants, it isnecessary that the stigma of the one form should be fertilised by pollentaken from the stamens of corresponding height in another form. So thatwith dimorphic species two unions, which may be called legitimate, arefully fertile; and two, which may be called illegitimate, are more or lessinfertile. With trimorphic species six unions are legitimate, or fullyfertile, and twelve are illegitimate, or more or less infertile.

The infertility which may be observed in various dimorphic and trimorphicplants, when they are illegitimately fertilised, that is by pollen takenfrom stamens not corresponding in height with the pistil, differs much indegree, up to absolute and utter sterility; just in the same manner asoccurs in crossing distinct species. As the degree of sterility in thelatter case depends in an eminent degree on the conditions of life beingmore or less favourable, so I have found it with illegitimate unions. Itis well known that if pollen of a distinct species be placed on the stigmaof a flower, and its own pollen be afterwards, even after a considerableinterval of time, placed on the same stigma, its action is so stronglyprepotent that it generally annihilates the effect of the foreign pollen;so it is with the pollen of the several forms of the same species, forlegitimate pollen is strongly prepotent over illegitimate pollen, when bothare placed on the same stigma. I ascertained this by fertilising severalflowers, first illegitimately, and twenty-four hours afterwardslegitimately, with pollen taken from a peculiarly coloured variety, and allthe seedlings were similarly coloured; this shows that the legitimatepollen, though applied twenty-four hours subsequently, had wholly destroyedor prevented the action of the previously applied illegitimate pollen. Again, as in making reciprocal crosses between the same two species, thereis occasionally a great difference in the result, so the same thing occurswith trimorphic plants; for instance, the mid-styled form of Lythrumsalicaria was illegitimately fertilised with the greatest ease by pollenfrom the longer stamens of the short-styled form, and yielded many seeds;but the latter form did not yield a single seed when fertilised by thelonger stamens of the mid-styled form.

In all these respects, and in others which might be added, the forms of thesame undoubted species, when illegitimately united, behave in exactly thesame manner as do two distinct species when crossed. This led me carefullyto observe during four years many seedlings, raised from severalillegitimate unions. The chief result is that these illegitimate plants,as they may be called, are not fully fertile. It is possible to raise fromdimorphic species, both long-styled and short-styled illegitimate plants,and from trimorphic plants all three illegitimate forms. These can then beproperly united in a legitimate manner. When this is done, there is noapparent reason why they should not yield as many seeds as did theirparents when legitimately fertilised. But such is not the case. They areall infertile, in various degrees; some being so utterly and incurablysterile that they did not yield during four seasons a single seed or evenseed-capsule. The sterility of these illegitimate plants, when united witheach other in a legitimate manner, may be strictly compared with that ofhybrids when crossed inter se. If, on the other hand, a hybrid is crossedwith either pure parent-species, the sterility is usually much lessened: and so it is when an illegitimate plant is fertilised by a legitimateplant. In the same manner as the sterility of hybrids does not always runparallel with the difficulty of making the first cross between the twoparent-species, so that sterility of certain illegitimate plants wasunusually great, while the sterility of the union from which they werederived was by no means great. With hybrids raised from the same seed-capsule the degree of sterility is innately variable, so it is in a markedmanner with illegitimate plants. Lastly, many hybrids are profuse andpersistent flowerers, while other and more sterile hybrids produce fewflowers, and are weak, miserable dwarfs; exactly similar cases occur withthe illegitimate offspring of various dimorphic and trimorphic plants.

Altogether there is the closest identity in character and behaviour betweenillegitimate plants and hybrids. It is hardly an exaggeration to maintainthat illegitimate plants are hybrids, produced within the limits of thesame species by the improper union of certain forms, while ordinary hybridsare produced from an improper union between so-called distinct species. Wehave also already seen that there is the closest similarity in all respectsbetween first illegitimate unions and first crosses between distinctspecies. This will perhaps be made more fully apparent by an illustration;we may suppose that a botanist found two well-marked varieties (and suchoccur) of the long-styled form of the trimorphic Lythrum salicaria, andthat he determined to try by crossing whether they were specificallydistinct. He would find that they yielded only about one-fifth of theproper number of seed, and that they behaved in all the other abovespecified respects as if they had been two distinct species. But to makethe case sure, he would raise plants from his supposed hybridised seed, andhe would find that the seedlings were miserably dwarfed and utterlysterile, and that they behaved in all other respects like ordinary hybrids. He might then maintain that he had actually proved, in accordance with thecommon view, that his two varieties were as good and as distinct species asany in the world; but he would be completely mistaken.

The facts now given on dimorphic and trimorphic plants are important,because they show us, first, that the physiological test of lessenedfertility, both in first crosses and in hybrids, is no safe criterion ofspecific distinction; secondly, because we may conclude that there is someunknown bond which connects the infertility of illegitimate unions withthat of their illegitimate offspring, and we are led to extend the sameview to first crosses and hybrids; thirdly, because we find, and this seemsto me of especial importance, that two or three forms of the same speciesmay exist and may differ in no respect whatever, either in structure or inconstitution, relatively to external conditions, and yet be sterile whenunited in certain ways. For we must remember that it is the union of thesexual elements of individuals of the same form, for instance, of two long-styled forms, which results in sterility; while it is the union of thesexual elements proper to two distinct forms which is fertile. Hence thecase appears at first sight exactly the reverse of what occurs, in theordinary unions of the individuals of the same species and with crossesbetween distinct species. It is, however, doubtful whether this is reallyso; but I will not enlarge on this obscure subject.

We may, however, infer as probable from the consideration of dimorphic andtrimorphic plants, that the sterility of distinct species when crossed andof their hybrid progeny, depends exclusively on the nature of their sexualelements, and not on any difference in their structure or generalconstitution. We are also led to this same conclusion by consideringreciprocal crosses, in which the male of one species cannot be united, orcan be united with great difficulty, with the female of a second species,while the converse cross can be effected with perfect facility. Thatexcellent observer, Gartner, likewise concluded that species when crossedare sterile owing to differences confined to their reproductive systems.

FERTILITY OF VARIETIES WHEN CROSSED, AND OF THEIR MONGREL OFFSPRING, NOTUNIVERSAL.

It may be urged as an overwhelming argument that there must be someessential distinction between species and varieties inasmuch as the latter,however much they may differ from each other in external appearance, crosswith perfect facility, and yield perfectly fertile offspring. With someexceptions, presently to be given, I fully admit that this is the rule. But the subject is surrounded by difficulties, for, looking to varietiesproduced under nature, if two forms hitherto reputed to be varieties befound in any degree sterile together, they are at once ranked by mostnaturalists as species. For instance, the blue and red pimpernel, whichare considered by most botanists as varieties, are said by Gartner to bequite sterile when crossed, and he consequently ranks them as undoubtedspecies. If we thus argue in a circle, the fertility of all varietiesproduced under nature will assuredly have to be granted.

If we turn to varieties, produced, or supposed to have been produced, underdomestication, we are still involved in some doubt. For when it is stated,for instance, that certain South American indigenous domestic dogs do notreadily unite with European dogs, the explanation which will occur toeveryone, and probably the true one, is that they are descended fromaboriginally distinct species. Nevertheless the perfect fertility of somany domestic races, differing widely from each other in appearance, forinstance, those of the pigeon, or of the cabbage, is a remarkable fact;more especially when we reflect how many species there are, which, thoughresembling each other most closely, are utterly sterile when intercrossed. Several considerations, however, render the fertility of domestic varietiesless remarkable. In the first place, it may be observed that the amount ofexternal difference between two species is no sure guide to their degree ofmutual sterility, so that similar differences in the case of varietieswould be no sure guide. It is certain that with species the cause liesexclusively in differences in their sexual constitution. Now the varyingconditions to which domesticated animals and cultivated plants have beensubjected, have had so little tendency towards modifying the reproductivesystem in a manner leading to mutual sterility, that we have good groundsfor admitting the directly opposite doctrine of Pallas, namely, that suchconditions generally eliminate this tendency; so that the domesticateddescendants of species, which in their natural state probably would havebeen in some degree sterile when crossed, become perfectly fertiletogether. With plants, so far is cultivation from giving a tendencytowards sterility between distinct species, that in several well-authenticated cases already alluded to, certain plants have been affectedin an opposite manner, for they have become self-impotent, while stillretaining the capacity of fertilising, and being fertilised by, otherspecies. If the Pallasian doctrine of the elimination of sterility throughlong-continued domestication be admitted, and it can hardly be rejected, itbecomes in the highest degree improbable that similar conditions long-continued should likewise induce this tendency; though in certain cases,with species having a peculiar constitution, sterility might occasionallybe thus caused. Thus, as I believe, we can understand why, withdomesticated animals, varieties have not been produced which are mutuallysterile; and why with plants only a few such cases, immediately to begiven, have been observed.

The real difficulty in our present subject is not, as it appears to me, whydomestic varieties have not become mutually infertile when crossed, but whythis has so generally occurred with natural varieties, as soon as they havebeen permanently modified in a sufficient degree to take rank as species. We are far from precisely knowing the cause; nor is this surprising, seeinghow profoundly ignorant we are in regard to the normal and abnormal actionof the reproductive system. But we can see that species, owing to theirstruggle for existence with numerous competitors, will have been exposedduring long periods of time to more uniform conditions, than have domesticvarieties; and this may well make a wide difference in the result. For weknow how commonly wild animals and plants, when taken from their naturalconditions and subjected to captivity, are rendered sterile; and thereproductive functions of organic beings which have always lived undernatural conditions would probably in like manner be eminently sensitive tothe influence of an unnatural cross. Domesticated productions, on theother hand, which, as shown by the mere fact of their domestication, werenot originally highly sensitive to changes in their conditions of life, andwhich can now generally resist with undiminished fertility repeated changesof conditions, might be expected to produce varieties, which would belittle liable to have their reproductive powers injuriously affected by theact of crossing with other varieties which had originated in a like manner.

I have as yet spoken as if the varieties of the same species wereinvariably fertile when intercrossed. But it is impossible to resist theevidence of the existence of a certain amount of sterility in the fewfollowing cases, which I will briefly abstract. The evidence is at leastas good as that from which we believe in the sterility of a multitude ofspecies. The evidence is also derived from hostile witnesses, who in allother cases consider fertility and sterility as safe criterions of specificdistinction. Gartner kept, during several years, a dwarf kind of maizewith yellow seeds, and a tall variety with red seeds growing near eachother in his garden; and although these plants have separated sexes, theynever naturally crossed. He then fertilised thirteen flowers of the onekind with pollen of the other; but only a single head produced any seed,and this one head produced only five grains. Manipulation in this casecould not have been injurious, as the plants have separated sexes. No one,I believe, has suspected that these varieties of maize are distinctspecies; and it is important to notice that the hybrid plants thus raisedwere themselves PERFECTLY fertile; so that even Gartner did not venture toconsider the two varieties as specifically distinct.

Girou de Buzareingues crossed three varieties of gourd, which like themaize has separated sexes, and he asserts that their mutual fertilisationis by so much the less easy as their differences are greater. How farthese experiments may be trusted, I know not; but the forms experimented onare ranked by Sagaret, who mainly founds his classification by the test ofinfertility, as varieties, and Naudin has come to the same conclusion.

The following case is far more remarkable, and seems at first incredible;but it is the result of an astonishing number of experiments made duringmany years on nine species of Verbascum, by so good an observer and sohostile a witness as Gartner: namely, that the yellow and white varietieswhen crossed produce less seed than the similarly coloured varieties of thesame species. Moreover, he asserts that, when yellow and white varietiesof one species are crossed with yellow and white varieties of a DISTINCTspecies, more seed is produced by the crosses between the similarlycoloured flowers, than between those which are differently coloured. Mr.Scott also has experimented on the species and varieties of Verbascum; andalthough unable to confirm Gartner's results on the crossing of thedistinct species, he finds that the dissimilarly coloured varieties of thesame species yield fewer seeds, in the proportion of eighty-six to 100,than the similarly coloured varieties. Yet these varieties differ in norespect, except in the colour of their flowers; and one variety cansometimes be raised from the seed of another.

Kolreuter, whose accuracy has been confirmed by every subsequent observer,has proved the remarkable fact that one particular variety of the commontobacco was more fertile than the other varieties, when crossed with awidely distinct species. He experimented on five forms which are commonlyreputed to be varieties, and which he tested by the severest trial, namely,by reciprocal crosses, and he found their mongrel offspring perfectlyfertile. But one of these five varieties, when used either as the fatheror mother, and crossed with the Nicotiana glutinosa, always yielded hybridsnot so sterile as those which were produced from the four other varietieswhen crossed with N. glutinosa. Hence the reproductive system of this onevariety must have been in some manner and in some degree modified.

>From these facts it can no longer be maintained that varieties when crossedare invariably quite fertile. From the great difficulty of ascertainingthe infertility of varieties in a state of nature, for a supposed variety,if proved to be infertile in any degree, would almost universally be rankedas a species; from man attending only to external characters in hisdomestic varieties, and from such varieties not having been exposed forvery long periods to uniform conditions of life; from these severalconsiderations we may conclude that fertility does not constitute afundamental distinction between varieties and species when crossed. Thegeneral sterility of crossed species may safely be looked at, not as aspecial acquirement or endowment, but as incidental on changes of anunknown nature in their sexual elements.

HYBRIDS AND MONGRELS COMPARED, INDEPENDENTLY OF THEIR FERTILITY.

Independently of the question of fertility, the offspring of species and ofvarieties when crossed may be compared in several other respects. Gartner,whose strong wish it was to draw a distinct line between species andvarieties, could find very few, and, as it seems to me, quite unimportantdifferences between the so-called hybrid offspring of species, and theso-called mongrel offspring of varieties. And, on the other hand, theyagree most closely in many important respects.

I shall here discuss this subject with extreme brevity. The most importantdistinction is, that in the first generation mongrels are more variablethan hybrids; but Gartner admits that hybrids from species which have longbeen cultivated are often variable in the first generation; and I havemyself seen striking instances of this fact. Gartner further admits thathybrids between very closely allied species are more variable than thosefrom very distinct species; and this shows that the difference in thedegree of variability graduates away. When mongrels and the more fertilehybrids are propagated for several generations, an extreme amount ofvariability in the offspring in both cases is notorious; but some fewinstances of both hybrids and mongrels long retaining a uniform charactercould be given. The variability, however, in the successive generations ofmongrels is, perhaps, greater than in hybrids.

This greater variability in mongrels than in hybrids does not seem at allsurprising. For the parents of mongrels are varieties, and mostly domesticvarieties (very few experiments having been tried on natural varieties),and this implies that there has been recent variability; which would oftencontinue and would augment that arising from the act of crossing. Theslight variability of hybrids in the first generation, in contrast withthat in the succeeding generations, is a curious fact and deservesattention. For it bears on the view which I have taken of one of thecauses of ordinary variability; namely, that the reproductive system, frombeing eminently sensitive to changed conditions of life, fails under thesecircumstances to perform its proper function of producing offspring closelysimilar in all respects to the parent-form. Now, hybrids in the firstgeneration are descended from species (excluding those long cultivated)which have not had their reproductive systems in any way affected, and theyare not variable; but hybrids themselves have their reproductive systemsseriously affected, and their descendants are highly variable.

But to return to our comparison of mongrels and hybrids: Gartner statesthat mongrels are more liable than hybrids to revert to either parent form;but this, if it be true, is certainly only a difference in degree. Moreover, Gartner expressly states that the hybrids from long cultivatedplants are more subject to reversion than hybrids from species in theirnatural state; and this probably explains the singular difference in theresults arrived at by different observers. Thus Max Wichura doubts whetherhybrids ever revert to their parent forms, and he experimented onuncultivated species of willows, while Naudin, on the other hand, insistsin the strongest terms on the almost universal tendency to reversion inhybrids, and he experimented chiefly on cultivated plants. Gartner furtherstates that when any two species, although most closely allied to eachother, are crossed with a third species, the hybrids are widely differentfrom each other; whereas if two very distinct varieties of one species arecrossed with another species, the hybrids do not differ much. But thisconclusion, as far as I can make out, is founded on a single experiment;and seems directly opposed to the results of several experiments made byKolreuter.

Such alone are the unimportant differences which Gartner is able to pointout between hybrid and mongrel plants. On the other hand, the degrees andkinds of resemblance in mongrels and in hybrids to their respectiveparents, more especially in hybrids produced from nearly related species,follow, according to Gartner the same laws. When two species are crossed,one has sometimes a prepotent power of impressing its likeness on thehybrid. So I believe it to be with varieties of plants; and with animals,one variety certainly often has this prepotent power over another variety. Hybrid plants produced from a reciprocal cross generally resemble eachother closely, and so it is with mongrel plants from a reciprocal cross. Both hybrids and mongrels can be reduced to either pure parent form, byrepeated crosses in successive generations with either parent.

These several remarks are apparently applicable to animals; but the subjectis here much complicated, partly owing to the existence of secondary sexualcharacters; but more especially owing to prepotency in transmittinglikeness running more strongly in one sex than in the other, both when onespecies is crossed with another and when one variety is crossed withanother variety. For instance, I think those authors are right whomaintain that the ass has a prepotent power over the horse, so that boththe mule and the hinny resemble more closely the ass than the horse; butthat the prepotency runs more strongly in the male than in the female ass,so that the mule, which is an offspring of the male ass and mare, is morelike an ass than is the hinny, which is the offspring of the female-ass andstallion.

Much stress has been laid by some authors on the supposed fact, that it isonly with mongrels that the offspring are not intermediate in character,but closely resemble one of their parents; but this does sometimes occurwith hybrids, yet I grant much less frequently than with mongrels. Lookingto the cases which I have collected of cross-bred animals closelyresembling one parent, the resemblances seem chiefly confined to charactersalmost monstrous in their nature, and which have suddenly appeared--such asalbinism, melanism, deficiency of tail or horns, or additional fingers andtoes; and do not relate to characters which have been slowly acquiredthrough selection. A tendency to sudden reversions to the perfectcharacter of either parent would, also, be much more likely to occur withmongrels, which are descended from varieties often suddenly produced andsemi-monstrous in character, than with hybrids, which are descended fromspecies slowly and naturally produced. On the whole, I entirely agree withDr. Prosper Lucas, who, after arranging an enormous body of facts withrespect to animals, comes to the conclusion that the laws of resemblance ofthe child to its parents are the same, whether the two parents differlittle or much from each other, namely, in the union of individuals of thesame variety, or of different varieties, or of distinct species.

Independently of the question of fertility and sterility, in all otherrespects there seems to be a general and close similarity in the offspringof crossed species, and of crossed varieties. If we look at species ashaving been specially created, and at varieties as having been produced bysecondary laws, this similarity would be an astonishing fact. But itharmonises perfectly with the view that there is no essential distinctionbetween species and varieties.

SUMMARY OF CHAPTER.

First crosses between forms, sufficiently distinct to be ranked as species,and their hybrids, are very generally, but not universally, sterile. Thesterility is of all degrees, and is often so slight that the most carefulexperimentalists have arrived at diametrically opposite conclusions inranking forms by this test. The sterility is innately variable inindividuals of the same species, and is eminently susceptible to action offavourable and unfavourable conditions. The degree of sterility does notstrictly follow systematic affinity, but is governed by several curious andcomplex laws. It is generally different, and sometimes widely different inreciprocal crosses between the same two species. It is not always equal indegree in a first cross and in the hybrids produced from this cross.

In the same manner as in grafting trees, the capacity in one species orvariety to take on another, is incidental on differences, generally of anunknown nature, in their vegetative systems, so in crossing, the greater orless facility of one species to unite with another is incidental on unknowndifferences in their reproductive systems. There is no more reason tothink that species have been specially endowed with various degrees ofsterility to prevent their crossing and blending in nature, than to thinkthat trees have been specially endowed with various and somewhat analogousdegrees of difficulty in being grafted together in order to prevent theirinarching in our forests.

The sterility of first crosses and of their hybrid progeny has not beenacquired through natural selection. In the case of first crosses it seemsto depend on several circumstances; in some instances in chief part on theearly death of the embryo. In the case of hybrids, it apparently dependson their whole organisation having been disturbed by being compounded fromtwo distinct forms; the sterility being closely allied to that which sofrequently affects pure species, when exposed to new and unnaturalconditions of life. He who will explain these latter cases will be able toexplain the sterility of hybrids. This view is strongly supported by aparallelism of another kind: namely, that, firstly, slight changes in theconditions of life add to the vigour and fertility of all organic beings;and secondly, that the crossing of forms, which have been exposed toslightly different conditions of life, or which have varied, favours thesize, vigour and fertility of their offspring. The facts given on thesterility of the illegitimate unions of dimorphic and trimorphic plants andof their illegitimate progeny, perhaps render it probable that some unknownbond in all cases connects the degree of fertility of first unions withthat of their offspring. The consideration of these facts on dimorphism,as well as of the results of reciprocal crosses, clearly leads to theconclusion that the primary cause of the sterility of crossed species isconfined to differences in their sexual elements. But why, in the case ofdistinct species, the sexual elements should so generally have become moreor less modified, leading to their mutual infertility, we do not know; butit seems to stand in some close relation to species having been exposed forlong periods of time to nearly uniform conditions of life.

It is not surprising that the difficulty in crossing any two species, andthe sterility of their hybrid offspring, should in most cases correspond,even if due to distinct causes: for both depend on the amount ofdifference between the species which are crossed. Nor is it surprisingthat the facility of effecting a first cross, and the fertility of thehybrids thus produced, and the capacity of being grafted together--thoughthis latter capacity evidently depends on widely differentcircumstances--should all run, to a certain extent, parallel with thesystematic affinity of the forms subjected to experiment; for systematicaffinity includes resemblances of all kinds.

First crosses between forms known to be varieties, or sufficiently alike tobe considered as varieties, and their mongrel offspring, are verygenerally, but not, as is so often stated, invariably fertile. Nor is thisalmost universal and perfect fertility surprising, when it is rememberedhow liable we are to argue in a circle with respect to varieties in a stateof nature; and when we remember that the greater number of varieties havebeen produced under domestication by the selection of mere externaldifferences, and that they have not been long exposed to uniform conditionsof life. It should also be especially kept in mind, that long-continueddomestication tends to eliminate sterility, and is therefore little likelyto induce this same quality. Independently of the question of fertility,in all other respects there is the closest general resemblance betweenhybrids and mongrels, in their variability, in their power of absorbingeach other by repeated crosses, and in their inheritance of characters fromboth parent-forms. Finally, then, although we are as ignorant of theprecise cause of the sterility of first crosses and of hybrids as we arewhy animals and plants removed from their natural conditions becomesterile, yet the facts given in this chapter do not seem to me opposed tothe belief that species aboriginally existed as varieties.