Thus, for example, it is quite impossible that two opposing characteristics can both be represented in the fertilised ovum. One of them must be suppressed or thrown out or got rid of in some way or another. For example, the union of the male element and the female element will give rise to an embryo which may be eventually either a male or a female individual, but cannot be both. There were possibilities of it being either the one or the other at the beginning, but since the two possibilities are mutually antagonistic, one or the other must be eliminated. So again, supposing that the colour of the eyes on the one side were brown, and on the other side blue, the possibilities are that the fertilised ovum may give rise to an individual having either blue eyes or brown eyes, but, again, not both.
A variation in offspring then may be regarded as a difference between that offspring and the parents, which is due to some change in the germ-plasm, some difference, that is, between the germ-plasm from which the parent sprang and that from which the next generation arose. Such differences will, of course, be introduced at the time of fertilisation. It is important to keep clearly in mind the difference between a true variation, in the sense that we have just used the term, and a modification which is caused by the varying effects of influences affecting parents and offspring. Unless these two things are kept mentally distinct, much confusion of thought is apt to arise.
The above statement does not necessarily mean that the germ-plasm carried in the sperms and the ova respectively, cannot be affected in any way. Indeed one is forced to the conclusion that such germ-cells must be influenced by the nutrient fluids supplied to them, and by the existence of toxic or poisonous substances in the body of the parent. It is quite conceivable, and indeed inevitable, that the individual embryo resulting from the fusion of such poisoned germ-cells will show modifications, but these, however, are not to be regarded in any sense as true variations, for the simple reason that these modifications do not take their origin in the actual germ-plasm itself, but are simply the result of abnormal stimuli. Such modifications, no matter in what direction they may be, are, of course, not transmissible to the next generation, for the very obvious reason that the germ-cells which are to be concerned with the next generation have been already produced. The germ-plasm itself passes on unchanged in so far as its hereditary possibilities are concerned. We see, therefore, that in order to think clearly on this matter we must limit the meaning of the word variation to such differences or changes in germ-plasm which indicates some real change of an inheritable nature. The term should not be used to apply to a mere passing environment, in which the germ-plasm happens to be, caused by the presence of poisons, or similar factors.