EXPERIMENTAL VARIATION The Types of Variation.—Not everything differentiating an offspring from its parents can be handed down to later generations (see above). Only the alterations occurring in the hereditary de terminers, or genes (q.v.), can furnish actually new materials for the permanent fabric of a race, and these changes in the genes, called gene-mutations, are remarkably rare. However, after the mutations have occurred, the changed or mutant genes enter into varying combinations with each other and with the old genes, in successive generations, and so the actual degree of variation be- . comes increased. Individuals having assortments of characteristics different from their parents are thus formed, by recombination, in somewhat the same way as, by drawing some cards out of each of- two given hands, we might obtain new hands, having one or another new assortment of cards. Besides the above two kinds of variation, which involve the genes, there are also many important variations not due to the genes, but to the environment. These variations may be grouped together under the term modifications. For detailed consideration we may best treat these three classes of variations in an order the reverse of the above.
Modifications Caused by Environment.—The common question, "Which is more important, heredity or environment ?" permits of no answer, except that both are equally and absolutely essential. We are related to our heredity and environment more nearly as an arithmetical product of the two than as a sum, and, with either annulled, the product must be zero. The genes provide the egg, not with the guarantees of developing certain characters, but only with the possibilities of reacting in certain ways so as to produce given adult structures and peculiarities, provided food and other conditions are appropriate. Change the environment (it cannot really be abolished completely) and these genes may exercise their possibilities of reacting in very different ways, to produce a different type of development, or disintegration.
It is, however, more pertinent to put our question in the form, "Which set of factors are ordinarily responsible for a larger part of the variation which actually occurs, differences in heredity or differences in environment?" If, now, we take variation in the larger sense, to include differences between all living things, it is obvious that, all in all, differences in environment are an insig nificant cause of the existing variation as compared with hered itary differences, and this is usually true even of two individuals chosen from closely related yet different species. On the other
hand, within a given population of one species, the variation due to environment is sometimes comparable to, or even much greater than, that dependent on the genes. Which sort of variation predominates, and to what extent, will depend on many citcum stances, and to answer our question will therefore require sep arate study, and often elaborate experimentation and measure ment, in the case of each population considered and even in the case of different characteristics in the same population. The unravelling of any intimate combination of hereditary and en vironmental effects will usually demand either that we hold the effective features of the environment constant (often an impos sible operation), or else that we secure a set of individuals iden tical in their genes (as a result of inbreeding, twinning or asexual reproduction) ; under such circumstances the residual variation, due to the influence that was not held constant, may then be as certained.
It is especially desirable to discover not only the extent of variation due to environmental differences, but the principles governing the production of this variation. This quest, however, leads to a study of all the intricacies of embryonic development, physiology and mechanistic biology in general. Although the de tails of the chemistry and physics comprised in these phenomena are for the most part scarcely guessed, we at least know, from our observations on the grosser visible occurrences, that the proc esses whereby most adult organs and characteristics are formed are highly involved, interdependent in a complicated way and composed of many successive reactions. Make one change, and effect follows effect until the final result bears no resemblance to the initial one. So, for example, a quantitative change in some primary process may appear in the adult as a qualitative change, and vice versa. It should also be observed that there are often reg ulatory mechanisms which automatically compensate for effects produced, as in regeneration of a lost limb or in the accelerated growth of a child following temporary stunting.