Page:Popular Science Monthly Volume 75.djvu/253

Rh It is not at all certain that it will be possible to reduce all the problems of the physiology of development to such categories as we have mentioned. The subject is full of unsolved problems, but so far as I can see no one has shown any real reason for assuming ultraphysical agencies in any of the events, and there is the same pragmatic reason for refusing to assent to such suggestions, which are made all too frequently, that there is in other fields of science. If we will be consistent, we are driven to the conclusion that the apparent simplicity of the germ is real, that the germ contains no gemmules, or determinants or other representative particles; that development is truly epigenetic, a natural series of events that succeed one another according to physicochemical and physiological laws; the explanation of the sequence consists simply in the discovery of each of its steps.

The problems of heredity and variation are included in a true physiological conception of the individual development; but some biological conceptions that have more or less status and reputation are inconsistent with it. Such are the inheritance of acquired characters, atavism, and the theory of unit characters. The first is a familiar problem that I shall not argue anew; the second logically implies the presence of ancestral representative particles in the germ, which is inconsistent with a physiological theory of development. But it is obvious that the facts united under the name of atavism or reversion take their place naturally in a physiological theory of development, as arrests of development, or modification of environment, or in other ways.

The theory of unit characters deserves more attention for it ia essentially a modern theory, and counts numerous adherents. This, conception has been most sharply formulated by De Vries in his Mutationstheorie. He says:

Bateson's allelomorphs constitute a similar conception. Such hypothetical elements of organization must be conceived as distinct from the germ on. They can be shuffled about from one generation to another, and can, therefore, be introduced, removed or replaced in the germ cells. It must be admitted that these conceptions fit certain facts of inheritance in many hybrids fairly well, but the progress of discovery has made necessary the installation of subsidiary hypotheses, so that the most recent conceptions of unit characters are becoming extremely