RECOGNITION OF SUBSPECIES IN ORNITHOLOGY. 63 



a very small ratio to the whole assemblage forming the species. 

 The question therefore arises whether, if a sufficiently large 

 number of individuals of a compound species inhabiting a 

 continuous area were forthcoming, they would not present a 

 perfectly graduated series. It is enough at present to reply 

 that, as the available material increases, the evidence of sub- 

 species becomes strengthened. In some cases, indeed, instead 

 of a graduated series, it is found that examples from opposite 

 confines of the range of a species resemble one another more 

 closely than they resemble examples from intermediate regions. 

 Thus our native Long-tailed Tit and Tree-Creeper are nearest to 

 those from Japan, and Marsh-Tits from Pekin can hardly be 

 distinguished from the North Italian form.* Moreover, con- 

 tiguous forms of the same species sometimes differ considerably. 



I believe the main reason that subspecies are more generally 

 recognized by ornithologists in the United States than in the 

 British Islands is that it was in the former country that the im- 

 portance of large series was first perceived. 



Eacial differentiation being, then, a concomitant of distribu- 

 tion, it becomes necessary to find some means of designating the 

 subordinate groups if account is to be taken of them, and the 

 most convenient means yet devised consists in using trinomials. 

 The nomenclatural value of subspecies cannot be expressed 

 within the limits imposed by the binomial system. No justi- 

 fication can be found for the suppression of the middle term of 

 trinomials when describing avowed subspecies, as is the custom 

 of some ornithologists, who thereby obscure the status of these 

 forms. It is obvious that an ornithologist who designates a new 

 subspecies binomially does not encumber nomenclature and tax 

 the memory to a less degree than if he were to name it tri- 

 nomially, since in the latter case the second term is always one 

 in current use. 



It is often urged in opposition to subspecies that they are 

 frequently based upon very trifling characters, and thus lead to 

 much subdivision. A feature, however, that appears trivial in a 

 single individual, or in a few individuals, at once assumes im- 

 portance when it is found in a large series. To an accustomed 



:: I am enabled to give these instances through the kindness of Dr. Ernst 

 Hartert. 



