3. A tendency to variation exists in many races, which manifests itself rather in modi fications of the specific type presented in the course of successive generations, than in the alterations induced by external agencies in the individuals of one generation. Thus we find that the offspring of any one pair do not all precisely agree among themselves, or with their parents, in bodily conformation or in psychical character ; but that individual differ ences (as they are termed) exist among them. Now, as this tendency to variation cannot be clearly traced to any influence of external cir cumstances, it is commonly distinguished by the term spontaneous ;' but there is much to favour the belief, that such variations are attributable to agencies operating either on both parents, previous to their intercourse, or at the time of coition, or to influences acting on the female during the period of utero-gestation.* For it may be uniformly observed, that those animals exhibit the greatest tendency to this `spontaneous' va riation, which present the greatest constitu tional adaptiveness to a variety of external conditions. And there are many cases in which it seems pretty clear, that the cause of this variation must be looked for in that com bination of influences, which is known under the general term domestication. Thus it may be stated as a general fact, that the varieties of colour so remarkable in domesticated races, tend to disappear when these races return in any considerable degree towards their primi tive wild state. This has been especially noticed in the horses, cattle, sheep, hogs, and dogs, introduced by the Spaniards into South America ; and the observation has been con firmed in other parts of the globe, showing that there is nothing peculiar to the climate of that country, which brings about the altera tion. Now this has been attributed to the free intermixture of differently-coloured pa rents, which tends to confuse the breeds, and to merge those varieties which are artificially kept up among domesticated races by the matching of similar parents. But although this cause doubtless operates, yet it is far from being the whole truth ; for the converse occurrence may take place in animals which are in process of being reclaimed from their wild state, under circumstances that forbid the idea of any such. intermixture. Thus, Mr. Bell informs us that an Australian bitch, or dingo, had a litter of puppies, the father of which was also of that breed ; both parents had been taken in the wild state ; both were of the uniform reddish-brown colour which belongs to the race ; and the mother had never bred before; but the young, bred in confine ment, and in a half-domesticated state, were all more or less spotted.* Now, considering the strong evidence which exists, that the colour of the offspring, in animals in which the hue is disposed to vary, may be affected by a mental impression at the time of im pregnation t, it does not seem improbable that various differences in the general condition of a wild and of a domesticated animal, should so affect the constitution of the latter, as to occasion an amount of variation in its off spring which does not exist in the former. The same general rule holds good in Plants. The tendency to the (so called) spontaneous production of varieties, is the greatest in those species which are most susceptible of influence by cultivation ; and those which are themselves least changed by external conditions, are those which are observed to transmit their distinc tive characters most constantly and uninter ruptedly from one generation to another. Fur ther, the influence of cultivation will sometimes develop in the individual the very same de partures from the usual specific type, which are, in other cases, `spontaneously' manifested in the offspring of a common parent. Thus, it is well known that, by cultivation, the prim rose may be converted into the polyanthus, and the cowslip into the oxslip ; but the late Dean of Manchester raised all these four forms from the seeds of the same plant ; and Professor Henslow has been equally suc cessful.
Until the limits of this tendency to spon taneous variation have been determined, there fore, in each particular instance, no valid specific distinctions can be erected. It hap pens, in certain groups, that a peculiarity, in itself very trivial, is transmitted uninterrupt edly from one generation to another, with such constancy and regularity as to justify us in believing that it has been always manifested. Thus many of the reputed specific differences of Moths and Butterflies rest on no other foundation, than the constant presence of a certain spot on some part of their wings; and there are Felines, which agree so closely in the structure of their skeletons as not to bedis tinguishable osteologically, and which are only regarded as belonging to distinct species, be cause a certain stripe or spot, which uniformly shows itself on the skin of one, is as uniformly absent from that of the other. On the other hand, we see, in all our domesticated races, great diversities among the offspring of a common parentage ; and these differences are sometimes so marked, that if he had not positive evidence of this common parentage, the naturalist would undoubtedly be justified, by the importance of the diversity, in the es tablishment of numerous specific types, when he has really seen but a few of the varieties into which one and the same may pass.
Of this tendency to " spontaneous " varia tion, it may be remarked, further, that like the variation which may be traced to external conditions, it has its limits, and does not really tend to confuse the boundaries of species, although it may frequently show the necessity for their extension. Thus, notwith standing the multitude of varieties of the Apple and the Pear which we possess, and notwithstanding the apparent triviality of the specific distinction between them (this being little else than the existence of a gritty centre in the pear, which is absent in the apple), yet we never find this distinction confounded by the presence of the distinctive characters of the pear in the descendant of an apple, or by the absence of it in the descendant of a pear. So we find that, notwithstanding the multi plication of breeds of dogs and horses, sheep and oxen, pigs and poultry, they all retain the characters by which their respective kinds are distinguished from their congeners. There is no tendency to an obliteration of the dis tinctions between a dog and a fox, or between a horse and an ass ; but these distinctions are perpetuated with the same regularity, that marks the stripes of the leopard or the spots on the wing of a moth. But, on the other hand, by observation of the spontaneous changes which particular tribes of plants or animals are liable to exhibit, we are some times led to extend our ideas of the compre hensiveness of species, and to bring into the same category forms which were previously supposed to be distinct types. Thus, for example, it is the opinion of many distin guished naturalists, that not only are all the breeds of Dogs to be considered as consti tuting one species, but that this species must also include the Wolf ; or, in other words, that the wolf and the dog are to be con sidered as collateral descendants from the same original parents. We shall presently i examine more in detail the evidence on which this belief is founded ; supposing it to be correct, we are only required to enlarge our idea of the comprehensiveness of the species Canis lupus, which must then include all the breeds of C. fantiliaris as its varieties ; and none of them are in any more danger than before, of being confounded with the fox or the jackal. Until, however, the limits of variation have been clearly determined in the case of any species which is known to exhibit the tendency, it is obviously impossible to erect specific distinctions that shall possess anything but a provisional value ; and thus the naturalist may feel a confident assurance of the genuineness of one set of species in a natural group, whilst he is utterly at a loss respecting another.