1890.] Allen on the Recognition of Geographical Forms. *1 



series of specimens of any species from distant points are com- 

 pared, in cases where the environment is more or less diverse, 

 we are accustomed to find appreciable differences — in some cases 

 slight, in others so well-marked as to be obvious at a glance. In 

 many instances, however, the differences are apparent only when 

 large series are available for comparison ; the differences being 

 merely average differences ; a greater or less proportion of the 

 specimens of the two series are practically indistinguishable, 

 the range of individual variation in either series overlapping 

 the difference characterizing the two slightly differentiated forms. 

 In other cases the occurrence of specimens that cannot be easily 

 referred, without knowing their origin, to one or to the other, is 

 exceptional. 



These being the general facts in the case we are at once con- 

 fronted with a serious question and a grave danger. The splitters 

 of an earlier time regarded every form, however slightly differ- 

 entiated, as a species. We arbitrarily define a species as a group of 

 individuals standing out distinct and disconnected from any similar 

 group, within which, though occupying different parts of a com- 

 mon habitat, we recognize other forms characteristic of, and re- 

 stricted to particular areas. These reach a maximum degree of 

 differentiation at some point in the habitat, and thence gradually 

 shade into other conspecific forms geographically contiguous. 



The distinction we thus make between species and subspecies, 

 though a purely conventional one, forms an indispensable basis 

 for the convenient recognition of the various minor stages in the 

 evolution of organized beings. The serious question is where to 

 draw the line in recognizing local forms in nomenclature. While 

 it is important to discover, and in some way record, even the very 

 slight differences due to peculiarities of environment, there is ob- 

 viously a reasonable limit to the naming of such forms by the use 

 of the trinominal system of nomenclature. How well-marked 

 then, must be a set of intergrades to entitle them to recognition r 

 On this point no arbitrary hard-and-fast line can be laid down. 

 Much, at least for the present, must be left to the discretion of the 

 investigator. We are still groping in the dark ; our steps are, in 

 the main, tentative and provisional. We cannot act decisively in 

 respect to the bird life of North America, or of any large area 

 till we know thoroughly the phases of variation throughout even- 

 nook and corner of the area in question. At present new forms 



