This page has been proofread, but needs to be validated.
908
VARIATION AND SELECTION

exclusive part in the formation of species. Wallace, whilst insisting that the range of observed and measured variation was much larger in proportion to the size of the organisms or parts of organism affected than was generally believed, leaned to the Darwinian view in excluding from the normal factors in the origin of species variations of the extremer ranges of magnitude. Later writers, and in particular W. Bateson and H. de Vries, have urged that as species are discontinuous—that is to say, marked off by structural differences of considerable magnitude—it is more probable that they have arisen from similarly discontinuous variations. De Vries gave the name “mutations” to such considerable variations (it is to be noted that a further concept, that of the mode of origin, has been added to the word mutation, and that the conception of relative size is being removed from it), and Bateson, de Vries and other writers have added many striking cases to those recorded by Darwin. It is doubtful, however, if there is any philosophical basis for distinguishing between variations merely by their magnitude. Differences which at their first appearance are very minute may result in the kind of variations which certainly would be classed as discontinuous. When the cells of the morula stage of an embryo are shaken asunder, each, instead of forming the appropriate part of a single organism, may form a complete new organism. And similarly in the development of a complicated organism, the suppression or doubling of a single cell or group of cells may bring about striking differences in the symmetry of the adult, or the reduction or increase in the number of metameric organs. A slight change in the structure or activity of a gland, by altering the internal secretion, may produce widespread alterations even in an adult organism; and we have good reason to suppose that, if compatible with viability, such minute changes would have even a greater ultimate effect if they occurred in an embryo. Even amongst the extreme advocates of the theory of mutations, the importance of magnitude is being discounted by their suggestion that some of the minute variations which have hitherto been regarded by them as insignificant “fluctuating variations” may be significant mutations. This in effect is to say that not magnitude but something else has to be sought for if we are to pick out amongst observed variations those which may be the material for the differentiation of species. So far as magnitude is concerned, the attack on the Darwinian position has failed, and it is agreed that species may be discontinuous and none the less have been produced from minute variations.

Causes of Variation.—Darwin was careful to insist that we did not know the laws of variation, and that when variation was attributed to “chance” no more should be read into the statement than an expression of our ignorance of the causation. It cannot now be doubted that a very large amount of observed variation, and especially of the indefinite variation which is sometimes spoken of as fluctuating variation, and which is usually distributed indefinitely round a mean, is directly associated with or induced by the environment. On various grounds attempts have been made to exclude such variation from the material for the making of species. The variations which de Vries has called mutations, and which were at first associated by Bateson with what he called discontinuous variations as the exclusive source of new species, are now supposed by de Vries to be distinguished from fluctuating variations by their mode of origin. Such mutations are not the product of the environment, but are an outcrop of the constitution of the germinal material of the varying organism, the result either of causes as yet undetected, or of the premutations and eliminations suggested by the work of Mendel (see Mendelism). These attempts to reject environmental variation rest on several grounds. In the first place the variations in question are “acquired characters.” When Darwin and Wallace framed their theories it was practically assumed that acquired characters were inherited, and the continuous slow action of the environment, moulding each generation to a slight extent in the same direction, was readily accepted by a generation inspired by Sir C. Lyell's doctrine of uniformitarianism in geological change, as a potent force. A. Weismann, however, from theoretical considerations and from analysis of supposed cases has at the least thrown doubt on the transmission of acquired characters. And so the newer school discard acquired characters and all the Lamarckian factors and leave the board clear for “mutations.” Analysis of any acquired character, however, shows that there are two factors involved. The organism is not a passive medium; the amount and nature of the response it makes to the action of environment depends on its own qualities, and these qualities, on any theory of inheritance, pass from generation to generation. Successful organisms, or well-adapted organisms, are those that have responded to the environment, whether by large or small variations, in suitable fashion. It is the character as acquired that affords the opportunity for selection, but the quality of responding to the environment so as to produce that character is transmitted. The conceptions of Weismann afford no ground for rejecting fluctuating variations from the materials for the production of species.

In the second place, it has been urged, particularly by de Vries, that experiment and observation have shown that the possible range of fluctuating variation is strictly limited. Breeders, he says, who try to build up qualities by the selection of the fluctuating variations that occur soon find that they reach a maximum beyond which their efforts fail, unless they turn to the more rarely occurring but heritable mutations. Something will be said later in this article as to the limitation of variation; here it is necessary only to say that de Vries is introducing no new idea. It is well known that some races and some organs in plants and animals are extremely variable, and that others are much less variable, and further, that whilst some of these differences may be due to intrinsic causes, others can be modified by experiment. As Sir W. T. Thiselton-Dyer has pointed out, what is called “specific stability” is a familiar obstacle to the producer of novelties, but one which he frequently succeeds in breaking down by cultural and other methods. In a survey of the palaeontological history of plants and animals, it is plain that extreme stability and extreme mutability both have occurred, sometimes having persisted for untold ages, sometimes having succeeded one another for varying periods. As yet no solid reason has been alleged for excluding fluctuating variations, on account of their limitation, from the materials for specific change. J. Cossar Ewart and H. M. Vernon have adduced experimental evidence as to the induction of variation by such causes as difference in the ages of the parents, in the maturity or freshness of the conjugating germ cells, and in the condition of nutrition for the embryos. Such cases show in the plainest way the co-operation of external or environmental and internal or constitutional factors.

With our present knowledge it is impossible to discriminate between variation that may or that may not be the material for the differentiation of species by scrutinizing either magnitude or probable causation. It is equally impossible to draw an exact line between variation induced by the environment and variation that may be termed intrinsic. Extrinsic and intrinsic factors are involved in every case, although there is a range from instances in which the external factor appears to be extreme to instances where the intrinsic factor is dominant. Even the results of mutilation involve an intrinsic factor, for they range, according to the organ and organism affected, from complete regeneration to the most imperfect healing. In the effects of exercise, of physiological activity and the gross results of such external agencies as food, temperature, climate, light, pressure and so forth the intrinsic factor appears to become more important. The interplay of extrinsic and intrinsic factors also differs with the age of the organism affected: the more nearly adult it may be, the more direct appears to be the influence of the environment; the more nearly embryonic the organism may be, the less direct is the result of a force impressed from without. The old organism is more stable and responds in obvious ways to direct assaults from without; the young organism is at once less stable and more profoundly