THE INADEQUACY OF EXPERIMENT 173 



So acute does our observation become that when we have once seen 

 and noted a face we are usually able to distinguish it again even after 

 the passage of months or years. Here our power of discrimination 

 far surpasses our power of expression in words. Nevertheless, so 

 confusing is the effect of modifications and spontaneous variations 

 that, even in the members of our own families, we are unable to 

 note clearly the effect produced by the crossing of parental 

 differences, though these are often considerable. No experimental 

 observer, no matter to what an extent a recluse and a student, is as 

 familiar with any type of animal or plant as he is with human 

 beings, for, at least, he is always more intimate with himself, a 

 human being, than with anything else. Much less, therefore, is 

 his power of observing the inheritance of small differences in other 

 types. When, therefore, relying on experimental evidence, he 

 states that the inheritance of the smaller mutations, like that of 

 ' sports,' is alternative, or when he declares that all true variations 

 are mutations, and therefore extremely stable, he is guessing. He 

 has, as I say, no data on which to found a judgment. Unless we, 

 also, are content merely to guess, we must test his hypotheses by 

 the ordinary scientific procedure of making a rigorous deductive 

 inference of consequences followed by an appeal to nature, to 

 ascertain whether these consequences accord with reality. The 

 reader must bear in mind that the experimental evidence is not 

 neglected when more is added to it, and the thinking not the 

 less likely to be correct because we use our best endeavours to 

 ensure its accuracy. I do not think it will be possible to contest a 

 single fact of this added evidence which I shall adduce ; and, if it 

 is incontestible, then, as far as I am able to judge, the conclusions 

 we shall reach follow as necessary consequences. 



290. When we cross two varieties of fowls or any other 

 domestic animals or plants, then, as regards the differences we are 

 able to note the larger differences the development is often 

 alternative. There is usually some blending observable both in 

 the first and in the succeeding hybrid generations ; but it is not 

 the conspicuous feature. The conspicuous feature is alternative 

 reproduction ; the resemblance to one parent type is much closer 

 than to the other; there is seldom such blending that both the 

 opposing parental characters are equally, or nearly equally repre- 

 sented. Now races of men differ among themselves in size, 

 shape, colour, and so on, almost as much as do varieties of poultry. 

 If human varieties cross, and almost every human variety has 

 crossed repeatedly with almost every other, the inheritance of 



