in nature; and such students as Bateson and Morgan have suggested that that application must be made by the taxonomists if (it is implied) the taxonomists ever become convinced that there are realities in nature that deserve to be called species.
Now, with these imputations of the unreality of species, I find myself no longer in accord. In the data that follow, evidence is presented that species are realities in nature more nearly satisfying the geneticist's concepts than the conventions of current taxonomy, and that the origins of such species are more satisfactorily explained on hereditary bases than by philosophic theories that may be invoked in extenuation of the fact and the factors of organic evolution.
We may begin our analysis of species by an examination of a few individuals taken in the field. We then become impressed with the truth of the assertion that no two individuals are exactly alike. And if we extend our investigations to several dozen individuals, we shall be confused by the varying characters that enter into any population which ordinarily passes as a species.
But if, on the other hand, we extend our examination to several hundred such individuals, we shall become impressed with another opinion, namely that there are many more points of uniformity than of variation among individuals taken from a given locality and habitat. Perhaps a half-dozen characters will show appreciable variation, while the hundreds of other characters that go to make up an organism are remarkably constant. One may believe that if larger series were more often utilized in taxonomic work the current bewilderment over variation would give way to a renewed respect for a certain uniformity that exists thruout such groups of individuals.
Again, the variation that may be observed in characters that do vary occurs within narrow limits. Thus, the antennal count in species of Cynips often varies by one, but among many thousands of individuals none has shown two segments more or less than any other individual in that species. The length of the wings in relation to the length of the body of any cynipid is not altogether an invariable item, but among the many measurements I have made this wing-body ratio is always within three or four per cent of the mean for the
