ON THE VARIATION OF ORGANIC BEINGS UNDER DOMESTICATION; AND ON THE PRINCIPLES OF SELECTION
The most favourable conditions for variation seem to be when organic beings are bred for many generations under domestication: one may infer this from the simple fact of the vast number of races and breeds of almost every plant and animal, which has long been domesticated. Under certain conditions organic beings even during their individual lives become slightly altered from their usual form, size, or other characters: and many of the peculiarities thus acquired are transmitted to their offspring. Thus in animals, the size and vigour of body, fatness, period of maturity, habits of body or consensual movements, habits of mind and temper, are modified or acquired during the life of the individual, and become inherited. There is reason to believe that when long exercise has given to certain muscles great development, or disuse has lessened them, that such development is also inherited. Food and climate will occasionally produce changes in the colour and texture of the external coverings of animals; and certain unknown conditions affect the horns of cattle in parts of Abyssinia; but whether these peculiarities, thus acquired during individual lives, have been inherited, I do not know. It appears certain that malconformation and lameness in horses, produced by too much work on hard roads,--that affections of the eyes in this animal probably caused by bad ventilation,--that tendencies towards many diseases in man, such as gout, caused by the course of life and ultimately producing changes of structure, and that many other diseases produced by unknown agencies, such as goitre, and the idiotcy resulting from it, all become hereditary.
 The cumulative effect of domestication is insisted on in the Origin, see e.g. Origin, Ed. i. p. 7, vi. p. 8.
 This type of variation passes into what he describes as the direct effect of conditions. Since they are due to causes acting during the adult life of the organism they might be called individual variations, but he uses this term for congenital variations, e.g. the differences discoverable in plants raised from seeds of the same pod (Origin, Ed. i. p. 45, vi. p. 53).
It is very doubtful whether the flowers and leaf-buds, annually produced from the same bulb, root, or tree, can properly be considered as parts of the same individual, though in some respects they certainly seem to be so. If they are parts of an individual, plants also are subject to considerable changes during their individual lives. Most florist-flowers if neglected degenerate, that is, they lose some of their characters; so common is this, that trueness is often stated, as greatly enhancing the value of a variety: tulips break their colours only after some years' culture; some plants become double and others single, by neglect or care: these characters can be transmitted by cuttings or grafts, and in some cases by true or seminal propagation. Occasionally a single bud on a plant assumes at once a new and widely different character: thus it is certain that nectarines have been produced on peach trees and moss roses on provence roses; white currants on red currant bushes; flowers of a different colour from that of the stock, in Chrysanthemums, Dahlias, sweet-williams, Azaleas, &c., &c.; variegated leaf-buds on many trees, and other similar cases. These new characters appearing in single buds, can, like those lesser changes affecting the whole plant, be multiplied not only by cuttings and such means, but often likewise by true seminal generation.
 [[It is not clear where the following note is meant to come]]: Case of Orchis,--most remarkable as not long cultivated by seminal propagation. Case of varieties which soon acquire, like Ęgilops and Carrot (and Maize) a certain general character and then go on varying.
The changes thus appearing during the lives of individual animals and plants are extremely rare compared with those which are congenital or which appear soon after birth. Slight differences thus arising are infinitely numerous: the proportions and form of every part of the frame, inside and outside, appear to vary in very slight degrees: anatomists dispute what is the "beau ideal" of the bones, the liver and kidneys, like painters do of the proportions of the face: the proverbial expression that no two animals or plants are born absolutely alike, is much truer when applied to those under domestication, than to those in a state of nature. Besides these slight differences, single individuals are occasionally born considerably unlike in certain parts or in their whole structure to their parents: these are called by horticulturists and breeders "sports"; and are not uncommon except when very strongly marked. Such sports are known in some cases to have been parents of some of our domestic races; and such probably have been the parents of many other races, especially of those which in some senses may be called hereditary monsters; for instance where there is an additional limb, or where all the limbs are stunted (as in the Ancon sheep), or where a part is wanting, as in rumpless fowls and tailless dogs or cats. The effects of external conditions on the size, colour and form, which can rarely and obscurely be detected during one individual life, become apparent after several generations: the slight differences, often hardly describable, which characterize the stock of different countries, and even of districts in the same country, seem to be due to such continued action.
 Here, as in the MS. of 1842, the author is inclined to minimise the variation occurring in nature.
 This is more strongly stated than in the Origin, Ed. i. p. 30.
On the hereditary tendency.
A volume might be filled with facts showing what a strong tendency there is to inheritance, in almost every case of the most trifling, as well as of the most remarkable congenital peculiarities. The term congenital peculiarity, I may remark, is a loose expression and can only mean a peculiarity apparent when the part affected is nearly or fully developed: in the Second Part, I shall have to discuss at what period of the embryonic life connatal peculiarities probably first appear; and I shall then be able to show from some evidence, that at whatever period of life a new peculiarity first appears, it tends hereditarily to appear at a corresponding period. Numerous though slight changes, slowly supervening in animals during mature life (often, though by no means always, taking the form of disease), are, as stated in the first paragraphs, very often hereditary. In plants, again, the buds which assume a different character from their stock likewise tend to transmit their new peculiarities. There is not sufficient reason to believe that either mutilations or changes of form produced by mechanical pressure, even if continued for hundreds of generations, or that any changes of structure quickly produced by disease, are inherited; it would appear as if the tissue of the part affected must slowly and freely grow into the new form, in order to be inheritable. There is a very great difference in the hereditary tendency of different peculiarities, and of the same peculiarity, in different individuals and species; thus twenty thousand seeds of the weeping ash have been sown and not one come up true;--out of seventeen seeds of the weeping yew, nearly all came up true. The ill-formed and almost monstrous "Niata" cattle of S. America and Ancon sheep, both when bred together and when crossed with other breeds, seem to transmit their peculiarities to their offspring as truly as the ordinary breeds. I can throw no light on these differences in the power of hereditary transmission. Breeders believe, and apparently with good cause, that a peculiarity generally becomes more firmly implanted after having passed through several generations; that is if one offspring out of twenty inherits a peculiarity from its parents, then its descendants will tend to transmit this peculiarity to a larger proportion than one in twenty; and so on in succeeding generations. I have said nothing about mental peculiarities being inheritable for I reserve this subject for a separate chapter.
 See Origin, Ed. i. p. 13.
 Origin, Ed. i. p. 86, vi. p. 105.
 It is interesting to find that though the author, like his contemporaries, believed in the inheritance of acquired characters, he excluded the case of mutilation.
Causes of Variation.
Attention must here be drawn to an important distinction in the first origin or appearance of varieties: when we see an animal highly kept producing offspring with an hereditary tendency to early maturity and fatness; when we see the wild-duck and Australian dog always becoming, when bred for one or a few generations in confinement, mottled in their colours; when we see people living in certain districts or circumstances becoming subject to an hereditary taint to certain organic diseases, as consumption or plica polonica,--we naturally attribute such changes to the direct effect of known or unknown agencies acting for one or more generations on the parents. It is probable that a multitude of peculiarities may be thus directly caused by unknown external agencies. But in breeds, characterized by an extra limb or claw, as in certain fowls and dogs; by an extra joint in the vertebrę; by the loss of a part, as the tail; by the substitution of a tuft of feathers for a comb in certain poultry; and in a multitude of other cases, we can hardly attribute these peculiarities directly to external influences, but indirectly to the laws of embryonic growth and of reproduction. When we see a multitude of varieties (as has often been the case, where a cross has been carefully guarded against) produced from seeds matured in the very same capsule, with the male and female principle nourished from the same roots and necessarily exposed to the same external influences; we cannot believe that the endless slight differences between seedling varieties thus produced, can be the effect of any corresponding difference in their exposure. We are led (as Müller has remarked) to the same conclusion, when we see in the same litter, produced by the same act of conception, animals considerably different.
 This corresponds to Origin, Ed. i. p. 10, vi. p. 9.
As variation to the degree here alluded to has been observed only in organic beings under domestication, and in plants amongst those most highly and long cultivated, we must attribute, in such cases, the varieties (although the difference between each variety cannot possibly be attributed to any corresponding difference of exposure in the parents) to the indirect effects of domestication on the action of the reproductive system. It would appear as if the reproductive powers failed in their ordinary function of producing new organic beings closely like their parents; and as if the entire organization of the embryo, under domestication, became in a slight degree plastic. We shall hereafter have occasion to show, that in organic beings, a considerable change from the natural conditions of life, affects, independently of their general state of health, in another and remarkable manner the reproductive system. I may add, judging from the vast number of new varieties of plants which have been produced in the same districts and under nearly the same routine of culture, that probably the indirect effects of domestication in making the organization plastic, is a much more efficient source of variation than any direct effect which external causes may have on the colour, texture, or form of each part. In the few instances in which, as in the Dahlia, the course of variation has been recorded, it appears that domestication produces little effect for several generations in rendering the organization plastic; but afterwards, as if by an accumulated effect, the original character of the species suddenly gives way or breaks.
 Origin, Ed. i. p. 8, vi. p. 10.
 For plasticity see Origin, Ed. i. pp. 12, 132.
 Var. under Dom., Ed. ii. I. p. 393.
We have hitherto only referred to the first appearance in individuals of new peculiarities; but to make a race or breed, something more is generally requisite than such peculiarities (except in the case of the peculiarities being the direct effect of constantly surrounding conditions) should be inheritable,--namely the principle of selection, implying separation. Even in the rare instances of sports, with the hereditary tendency very strongly implanted, crossing must be prevented with other breeds, or if not prevented the best characterized of the half-bred offspring must be carefully selected. Where the external conditions are constantly tending to give some character, a race possessing this character will be formed with far greater ease by selecting and breeding together the individuals most affected. In the case of the endless slight variations produced by the indirect effects of domestication on the action of the reproductive system, selection is indispensable to form races; and when carefully applied, wonderfully numerous and diverse races can be formed. Selection, though so simple in theory, is and has been important to a degree which can hardly be overrated. It requires extreme skill, the results of long practice, in detecting the slightest difference in the forms of animals, and it implies some distinct object in view; with these requisites and patience, the breeder has simply to watch for every the smallest approach to the desired end, to select such individuals and pair them with the most suitable forms, and so continue with succeeding generations. In most cases careful selection and the prevention of accidental crosses will be necessary for several generations, for in new breeds there is a strong tendency to vary and especially to revert to ancestral forms: but in every succeeding generation less care will be requisite for the breed will become truer; until ultimately only an occasional individual will require to be separated or destroyed. Horticulturalists in raising seeds regularly practise this, and call it "roguing," or destroying the "rogues" or false varieties. There is another and less efficient means of selection amongst animals: namely repeatedly procuring males with some desirable qualities, and allowing them and their offspring to breed freely together; and this in the course of time will affect the whole lot. These principles of selection have been methodically followed for scarcely a century; but their high importance is shown by the practical results, and is admitted in the writings of the most celebrated agriculturalists and horticulturalists;--I need only name Anderson, Marshall, Bakewell, Coke, Western, Sebright and Knight.
 Selection is here used in the sense of isolation, rather than as implying the summation of small differences. Professor Henslow in his Heredity of Acquired Characters in Plants, 1908, p. 2, quotes from Darwin's Var. under Dom., Ed. i. II. p. 271, a passage in which the author, speaking of the direct action of conditions, says:--"A new sub-variety would thus be produced without the aid of selection." Darwin certainly did not mean to imply that such varieties are freed from the action of natural selection, but merely that a new form may appear without summation of new characters. Professor Henslow is apparently unaware that the above passage is omitted in the second edition of Var. under Dom., II. p. 260.
Even in well-established breeds the individuals of which to an unpractised eye would appear absolutely similar, which would give, it might have been thought, no scope to selection, the whole appearance of the animal has been changed in a few years (as in the case of Lord Western's sheep), so that practised agriculturalists could scarcely credit that a change had not been effected by a cross with other breeds. Breeders both of plants and animals frequently give their means of selection greater scope, by crossing different breeds and selecting the offspring; but we shall have to recur to this subject again.
The external conditions will doubtless influence and modify the results of the most careful selection; it has been found impossible to prevent certain breeds of cattle from degenerating on mountain pastures; it would probably be impossible to keep the plumage of the wild-duck in the domesticated race; in certain soils, no care has been sufficient to raise cauliflower seed true to its character; and so in many other cases. But with patience it is wonderful what man has effected. He has selected and therefore in one sense made one breed of horses to race and another to pull; he has made sheep with fleeces good for carpets and other sheep good for broadcloth; he has, in the same sense, made one dog to find game and give him notice when found, and another dog to fetch him the game when killed; he has made by selection the fat to lie mixed with the meat in one breed and in another to accumulate in the bowels for the tallow-chandler; he has made the legs of one breed of pigeons long, and the beak of another so short, that it can hardly feed itself; he has previously determined how the feathers on a bird's body shall be coloured, and how the petals of many flowers shall be streaked or fringed, and has given prizes for complete success;--by selection, he has made the leaves of one variety and the flower-buds of another variety of the cabbage good to eat, at different seasons of the year; and thus has he acted on endless varieties. I do not wish to affirm that the long-and short-wooled sheep, or that the pointer and retriever, or that the cabbage and cauliflower have certainly descended from one and the same aboriginal wild stock; if they have not so descended, though it lessens what man has effected, a large result must be left unquestioned.
 See the Essay of 1842, p. 3.
In saying as I have done that man makes a breed, let it not be confounded with saying that man makes the individuals, which are given by nature with certain desirable qualities; man only adds together and makes a permanent gift of nature's bounties. In several cases, indeed, for instance in the "Ancon" sheep, valuable from not getting over fences, and in the turnspit dog, man has probably only prevented crossing; but in many cases we positively know that he has gone on selecting, and taking advantage of successive small variations.
Selection has been methodically followed, as I have said, for barely a century; but it cannot be doubted that occasionally it has been practised from the remotest ages, in those animals completely under the dominion of man. In the earliest chapters of the Bible there are rules given for influencing the colours of breeds, and black and white sheep are spoken of as separated. In the time of Pliny the barbarians of Europe and Asia endeavoured by cross-breeding with a wild stock to improve the races of their dogs and horses. The savages of Guyana now do so with their dogs: such care shows at least that the characters of individual animals were attended to. In the rudest times of English history, there were laws to prevent the exportation of fine animals of established breeds, and in the case of horses, in Henry VIII's time, laws for the destruction of all horses under a certain size. In one of the oldest numbers of the Phil. Transactions, there are rules for selecting and improving the breeds of sheep. Sir H. Bunbury, in 1660, has given rules for selecting the finest seedling plants, with as much precision as the best recent horticulturalist could. Even in the most savage and rude nations, in the wars and famines which so frequently occur, the most useful of their animals would be preserved: the value set upon animals by savages is shown by the inhabitants of Tierra del Fuego devouring their old women before their dogs, which as they asserted are useful in otter-hunting: who can doubt but that in every case of famine and war, the best otter-hunters would be preserved, and therefore in fact selected for breeding. As the offspring so obviously take after their parents, and as we have seen that savages take pains in crossing their dogs and horses with wild stocks, we may even conclude as probable that they would sometimes pair the most useful of their animals and keep their offspring separate. As different races of men require and admire different qualities in their domesticated animals, each would thus slowly, though unconsciously, be selecting a different breed. As Pallas has remarked, who can doubt but that the ancient Russian would esteem and endeavour to preserve those sheep in his flocks which had the thickest coats. This kind of insensible selection by which new breeds are not selected and kept separate, but a peculiar character is slowly given to the whole mass of the breed, by often saving the life of animals with certain characteristics, we may feel nearly sure, from what we see has been done by the more direct method of separate selection within the last 50 years in England, would in the course of some thousand years produce a marked effect.
 See Origin, Ed. i. p. 33, vi. p. 38. The evidence is given in the present Essay rather more fully than in the Origin.
 Journal of Researches, Ed. 1860, p. 214. "Doggies catch otters, old women no."
When once two or more races are formed, or if more than one race, or species fertile inter se, originally existed in a wild state, their crossing becomes a most copious source of new races. When two well-marked races are crossed the offspring in the first generation take more or less after either parent or are quite intermediate between them, or rarely assume characters in some degree new. In the second and several succeeding generations, the offspring are generally found to vary exceedingly, one compared with another, and many revert nearly to their ancestral forms. This greater variability in succeeding generations seems analogous to the breaking or variability of organic beings after having been bred for some generations under domestication. So marked is this variability in cross-bred descendants, that Pallas and some other naturalists have supposed that all variation is due to an original cross; but I conceive that the history of the potato, Dahlia, Scotch Rose, the guinea-pig, and of many trees in this country, where only one species of the genus exists, clearly shows that a species may vary where there can have been no crossing. Owing to this variability and tendency to reversion in cross-bred beings, much careful selection is requisite to make intermediate or new permanent races: nevertheless crossing has been a most powerful engine, especially with plants, where means of propagation exist by which the cross-bred varieties can be secured without incurring the risk of fresh variation from seminal propagation: with animals the most skilful agriculturalists now greatly prefer careful selection from a well-established breed, rather than from uncertain cross-bred stocks.
 The effects of crossing is much more strongly stated here than in the Origin. See Ed. i. p. 20, vi. p. 23, where indeed the opposite point of view is given. His change of opinion may be due to his work on pigeons. The whole of the discussion on crossing corresponds to Chapter VIII of the Origin, Ed. i. rather than to anything in the earlier part of the book.
 The parallelism between the effects of a cross and the effects of conditions is given from a different point of view in the Origin, Ed. i. p. 266, vi. p. 391. See the experimental evidence for this important principle in the author's work on Cross and Self-Fertilisation. Professor Bateson has suggested that the experiments should be repeated with gametically pure plants.
Although intermediate and new races may be formed by the mingling of others, yet if the two races are allowed to mingle quite freely, so that none of either parent race remain pure, then, especially if the parent races are not widely different, they will slowly blend together, and the two races will be destroyed, and one mongrel race left in its place. This will of course happen in a shorter time, if one of the parent races exists in greater number than the other. We see the effect of this mingling, in the manner in which the aboriginal breeds of dogs and pigs in the Oceanic Islands and the many breeds of our domestic animals introduced into S. America, have all been lost and absorbed in a mongrel race. It is probably owing to the freedom of crossing, that, in uncivilised countries, where inclosures do not exist, we seldom meet with more than one race of a species: it is only in enclosed countries, where the inhabitants do not migrate, and have conveniences for separating the several kinds of domestic animals, that we meet with a multitude of races. Even in civilised countries, want of care for a few years has been found to destroy the good results of far longer periods of selection and separation.
This power of crossing will affect the races of all terrestrial animals; for all terrestrial animals require for their reproduction the union of two individuals. Amongst plants, races will not cross and blend together with so much freedom as in terrestrial animals; but this crossing takes place through various curious contrivances to a surprising extent. In fact such contrivances exist in so very many hermaphrodite flowers by which an occasional cross may take place, that I cannot avoid suspecting (with Mr Knight) that the reproductive action requires, at intervals, the concurrence of distinct individuals. Most breeders of plants and animals are firmly convinced that benefit is derived from an occasional cross, not with another race, but with another family of the same race; and that, on the other hand, injurious consequences follow from long-continued close interbreeding in the same family. Of marine animals, many more, than was till lately believed, have their sexes on separate individuals; and where they are hermaphrodite, there seems very generally to be means through the water of one individual occasionally impregnating another: if individual animals can singly propagate themselves for perpetuity, it is unaccountable that no terrestrial animal, where the means of observation are more obvious, should be in this predicament of singly perpetuating its kind. I conclude, then, that races of most animals and plants, when unconfined in the same country, would tend to blend together.
 The so-called Knight-Darwin Law is often misunderstood. See Goebel in Darwin and Modern Science, 1909, p. 419; also F. Darwin, Nature, Oct. 27, 1898.
Whether our domestic races have descended from one or more wild stocks.
Several naturalists, of whom Pallas regarding animals, and Humboldt regarding certain plants, were the first, believe that the breeds of many of our domestic animals such as of the horse, pig, dog, sheep, pigeon, and poultry, and of our plants have descended from more than one aboriginal form. They leave it doubtful, whether such forms are to be considered wild races, or true species, whose offspring are fertile when crossed inter se. The main arguments for this view consist, firstly, of the great difference between such breeds, as the Race-and Cart-Horse, or the Greyhound and Bull-dog, and of our ignorance of the steps or stages through which these could have passed from a common parent; and secondly that in the most ancient historical periods, breeds resembling some of those at present most different, existed in different countries. The wolves of N. America and of Siberia are thought to be different species; and it has been remarked that the dogs belonging to the savages in these two countries resemble the wolves of the same country; and therefore that they have probably descended from two different wild stocks. In the same manner, these naturalists believe that the horse of Arabia and of Europe have probably descended from two wild stocks both apparently now extinct. I do not think the assumed fertility of these wild stocks any very great difficulty on this view; for although in animals the offspring of most cross-bred species are infertile, it is not always remembered that the experiment is very seldom fairly tried, except when two near species both breed freely (which does not readily happen, as we shall hereafter see) when under the dominion of man. Moreover in the case of the China and common goose, the canary and siskin, the hybrids breed freely; in other cases the offspring from hybrids crossed with either pure parent are fertile, as is practically taken advantage of with the yak and cow; as far as the analogy of plants serves, it is impossible to deny that some species are quite fertile inter se; but to this subject we shall recur.
 Pallas' theory is discussed in the Origin, Ed. i. pp. 253, 254, vi. p. 374.
 See Darwin's paper on the fertility of hybrids from the common and Chinese goose in Nature, Jan. 1, 1880.
On the other hand, the upholders of the view that the several breeds of dogs, horses, &c., &c., have descended each from one stock, may aver that their view removes all difficulty about fertility, and that the main argument from the high antiquity of different breeds, somewhat similar to the present breeds, is worth little without knowing the date of the domestication of such animals, which is far from being the case. They may also with more weight aver that, knowing that organic beings under domestication do vary in some degree, the argument from the great difference between certain breeds is worth nothing, without we know the limits of variation during a long course of time, which is far from the case. They may argue that almost every county in England, and in many districts of other countries, for instance in India, there are slightly different breeds of the domestic animals; and that it is opposed to all that we know of the distribution of wild animals to suppose that these have descended from so many different wild races or species: if so, they may argue, is it not probable that countries quite separate and exposed to different climates would have breeds not slightly, but considerably, different? Taking the most favourable case, on both sides, namely that of the dog; they might urge that such breeds as the bull-dog and turnspit have been reared by man, from the ascertained fact that strictly analogous breeds (namely the Niata ox and Ancon sheep) in other quadrupeds have thus originated. Again they may say, seeing what training and careful selection has effected for the greyhound, and seeing how absolutely unfit the Italian greyhound is to maintain itself in a state of nature, is it not probable that at least all greyhounds,--from the rough deerhound, the smooth Persian, the common English, to the Italian,--have descended from one stock? If so, is it so improbable that the deerhound and long-legged shepherd dog have so descended? If we admit this, and give up the bull-dog, we can hardly dispute the probable common descent of the other breeds.
 Origin, Ed. i. p. 19, vi. p. 22.
The evidence is so conjectural and balanced on both sides that at present I conceive that no one can decide: for my own part, I lean to the probability of most of our domestic animals having descended from more than one wild stock; though from the arguments last advanced and from reflecting on the slow though inevitable effect of different races of mankind, under different circumstances, saving the lives of and therefore selecting the individuals most useful to them, I cannot doubt but that one class of naturalists have much overrated the probable number of the aboriginal wild stocks. As far as we admit the difference of our races [[to be]] due to the differences of their original stocks, so much must we give up of the amount of variation produced under domestication. But this appears to me unimportant, for we certainly know in some few cases, for instance in the Dahlia, and potato, and rabbit, that a great number of varieties have proceeded from one stock; and, in many of our domestic races, we know that man, by slowly selecting and by taking advantage of sudden sports, has considerably modified old races and produced new ones. Whether we consider our races as the descendants of one or several wild stocks, we are in far the greater number of cases equally ignorant what these stocks were.
Limits to Variation in degree and kind.
Man's power in making races deends, in the first instance, on the stock on which he works being variable; but his labours are modified and limited, as we have seen, by the direct effects of the external conditions,--by the deficient or imperfect hereditariness of new peculiarities,--and by the tendency to continual variation and especially to reversion to ancestral forms. If the stock is not variable under domestication, of course he can do nothing; and it appears that species differ considerably in this tendency to variation, in the same way as even sub-varieties from the same variety differ greatly in this respect, and transmit to their offspring this difference in tendency. Whether the absence of a tendency to vary is an unalterable quality in certain species, or depends on some deficient condition of the particular state of domestication to which they are exposed, there is no evidence. When the organization is rendered variable, or plastic, as I have expressed it, under domestication, different parts of the frame vary more or less in different species: thus in the breeds of cattle it has been remarked that the horns are the most constant or least variable character, for these often remain constant, whilst the colour, size, proportions of the body, tendency to fatten &c., vary; in sheep, I believe, the horns are much more variable. As a general rule the less important parts of the organization seem to vary most, but I think there is sufficient evidence that every part occasionally varies in a slight degree. Even when man has the primary requisite variability he is necessarily checked by the health and life of the stock he is working on: thus he has already made pigeons with such small beaks that they can hardly eat and will not rear their own young; he has made families of sheep with so strong a tendency to early maturity and to fatten, that in certain pastures they cannot live from their extreme liability to inflammation; he has made (i.e. selected) sub-varieties of plants with a tendency to such early growth that they are frequently killed by the spring frosts; he has made a breed of cows having calves with such large hinder quarters that they are born with great difficulty, often to the death of their mothers; the breeders were compelled to remedy this by the selection of a breeding stock with smaller hinder quarters; in such a case, however, it is possible by long patience and great loss, a remedy might have been found in selecting cows capable of giving birth to calves with large hinder quarters, for in human kind there [[are]] no doubt hereditary bad and good confinements. Besides the limits already specified, there can be little doubt that the variation of different parts of the frame are connected together by many laws: thus the two sides of the body, in health and disease, seem almost always to vary together: it has been asserted by breeders that if the head is much elongated, the bones of the extremities will likewise be so; in seedling-apples large leaves and fruit generally go together, and serve the horticulturalist as some guide in his selection; we can here see the reason, as the fruit is only a metamorphosed leaf. In animals the teeth and hair seem connected, for the hairless Chinese dog is almost toothless. Breeders believe that one part of the frame or function being increased causes other parts to decrease: they dislike great horns and great bones as so much flesh lost; in hornless breeds of cattle certain bones of the head become more developed: it is said that fat accumulating in one part checks its accumulation in another, and likewise checks the action of the udder. The whole organization is so connected that it is probable there are many conditions determining the variation of each part, and causing other parts to vary with it; and man in making new races must be limited and ruled by all such laws.
 Var. under Dom., Ed. ii. vol. II. p. 211.
 This discussion corresponds to the Origin, Ed. i. pp. 11 and 143, vi. pp. 13 and 177.
In what consists Domestication.
In this chapter we have treated of variation under domestication, and it now remains to consider in what does this power of domestication consist, a subject of considerable difficulty. Observing that organic beings of almost every class, in all climates, countries, and times, have varied when long bred under domestication, we must conclude that the influence is of some very general nature. Mr Knight alone, as far as I know, has tried to define it; he believes it consists of an excess of food, together with transport to a more genial climate, or protection from its severities. I think we cannot admit this latter proposition, for we know how many vegetable products, aborigines of this country, here vary, when cultivated without any protection from the weather; and some of our variable trees, as apricots, peaches, have undoubtedly been derived from a more genial climate. There appears to be much more truth in the doctrine of excess of food being the cause, though I much doubt whether this is the sole cause, although it may well be requisite for the kind of variation desired by man, namely increase of size and vigour. No doubt horticulturalists, when they wish to raise new seedlings, often pluck off all the flower-buds, except a few, or remove the whole during one season, so that a great stock of nutriment may be thrown into the flowers which are to seed. When plants are transported from high-lands, forests, marshes, heaths, into our gardens and greenhouses, there must be a considerable change of food, but it would be hard to prove that there was in every case an excess of the kind proper to the plant. If it be an excess of food, compared with that which the being obtained in its natural state, the effects continue for an improbably long time; during how many ages has wheat been cultivated, and cattle and sheep reclaimed, and we cannot suppose their amount of food has gone on increasing, nevertheless these are amongst the most variable of our domestic productions. It has been remarked (Marshall) that some of the most highly kept breeds of sheep and cattle are truer or less variable than the straggling animals of the poor, which subsist on commons, and pick up a bare subsistence. In the case of forest-trees raised in nurseries, which vary more than the same trees do in their aboriginal forests, the cause would seem simply to lie in their not having to struggle against other trees and weeds, which in their natural state doubtless would limit the conditions of their existence. It appears to me that the power of domestication resolves itself into the accumulated effects of a change of all or some of the natural conditions of the life of the species, often associated with excess of food. These conditions moreover, I may add, can seldom remain, owing to the mutability of the affairs, habits, migrations, and knowledge of man, for very long periods the same. I am the more inclined to come to this conclusion from finding, as we shall hereafter show, that changes of the natural conditions of existence seem peculiarly to affect the action of the reproductive system. As we see that hybrids and mongrels, after the first generation, are apt to vary much, we may at least conclude that variability does not altogether depend on excess of food.
 See Origin, Ed. i. p. 7, vi. p. 7.
 [[Note in the original.]] "Isidore G. St Hilaire insists that breeding in captivity essential element. Schleiden on alkalies. [[See Var. under Dom., Ed. ii. vol. II. p. 244, note 10.]] What is it in domestication which causes variation?"
 [[Note in the original.]] "It appears that slight changes of condition [[are]] good for health; that more change affects the generative system, so that variation results in the offspring; that still more change checks or destroys fertility not of the offspring." Compare the Origin, Ed. i. p. 9, vi. p. 11. What the meaning of "not of the offspring" may be is not clear.
 In the Origin, Ed. i. p. 41, vi. p. 46 the question is differently treated; it is pointed out that a large stock of individuals gives a better chance of available variations occurring. Darwin quotes from Marshall that sheep in small lots can never be improved. This comes from Marshall's Review of the Reports to the Board of Agriculture, 1808, p. 406. In this Essay the name Marshall occurs in the margin. Probably this refers to loc. cit. p. 200, where unshepherded sheep in many parts of England are said to be similar owing to mixed breeding not being avoided.
 See Origin, Ed. i. p. 8, vi. p. 8.
After these views, it may be asked how it comes that certain animals and plants, which have been domesticated for a considerable length of time, and transported from very different conditions of existence, have not varied much, or scarcely at all; for instance, the ass, peacock, guinea-fowl, asparagus, Jerusalem artichoke. I have already said that probably different species, like different sub-varieties, possess different degrees of tendency to vary; but I am inclined to attribute in these cases the want of numerous races less to want of variability than to selection not having been practised on them. No one will take the pains to select without some corresponding object, either of use or amusement; the individuals raised must be tolerably numerous, and not so precious, but that he may freely destroy those not answering to his wishes. If guinea-fowls or peacocks became "fancy" birds, I cannot doubt that after some generations several breeds would be raised. Asses have not been worked on from mere neglect; but they differ in some degree in different countries. The insensible selection, due to different races of mankind preserving those individuals most useful to them in their different circumstances, will apply only to the oldest and most widely domesticated animals. In the case of plants, we must put entirely out of the case those exclusively (or almost so) propagated by cuttings, layers or tubers, such as the Jerusalem artichoke and laurel; and if we put on one side plants of little ornament or use, and those which are used at so early a period of their growth that no especial characters signify, as asparagus and seakale, I can think of none long cultivated which have not varied. In no case ought we to expect to find as much variation in a race when it alone has been formed, as when several have been formed, for their crossing and recrossing will greatly increase their variability.
 See Origin, Ed. i. p. 42, vi. p. 48.
Summary of first Chapter.
To sum up this chapter. Races are made under domestication: 1st, by the direct effects of the external conditions to which the species is exposed: 2nd, by the indirect effects of the exposure to new conditions, often aided by excess of food, rendering the organization plastic, and by man's selecting and separately breeding certain individuals, or introducing to his stock selected males, or often preserving with care the life of the individuals best adapted to his purposes: 3rd, by crossing and recrossing races already made, and selecting their offspring. After some generations man may relax his care in selection: for the tendency to vary and to revert to ancestral forms will decrease, so that he will have only occasionally to remove or destroy one of the yearly offspring which departs from its type. Ultimately, with a large stock, the effects of free crossing would keep, even without this care, his breed true. By these means man can produce infinitely numerous races, curiously adapted to ends, both most important and most frivolous; at the same time that the effects of the surrounding conditions, the laws of inheritance, of growth, and of variation, will modify and limit his labours.