3i« 



THE AMERICAN NATURALIST. [Vol. XXXVI 11. 



to the characters presented on the part of the insides of the 

 objects they are attempting to classify, although in saying this I 

 do not underrate the value of the external characters. 



The facts, then, that birds are a very homogeneous group and 

 the knowledge of all their characters possessed by individuals 

 who have attempted to classify them, has differed very widely in 

 amount, is the first factor that will account for the great differ- 

 ences to be seen in the various published classificatory schemes. 

 These are not the only reasons, however, and another very obvious 

 one is the attempt made by some classifiers to ignore the homo- 

 geneity of birds, and to arrange them after the manner of the other 

 great groups of animals, such as mammals or fishes. In other 

 words, the attempt is made to employ the same divisional group- 

 ings from subspecies to class in the case of birds, where perhaps no 

 greater differences can be found than exist between a thrush and 

 an ostrich, as they do in the case of mammals, where such gaps 

 exist as the one separating man and the ornithorhynchus. The 

 objection is raised here at once, however, that an order of birds, 

 for example, is a very different thing from an order of mammals. 

 This is a knotty question, and as time forbids my discussing it 

 here, I can only say that it leads directly to another very obvious 

 reason for the differences seen in the various arrangements that 

 have been proposed for birds, and that is this : — although ornith- 

 ologists, in this country at least, may be pretty well agreed as to 

 what is meant by a species, it is not clear that the same apparent 

 unanimity of opinion exists among them in regard to what is 

 meant by a genus, or rather what constitute generic characters ; 

 and so on for families, suborders and other groups, until we 

 arrive at the Class, and perhaps the Subclass, — groups, for evi- 

 dent reasons, again understood to possess the same value in all 

 avian schemes of classification wherein they have been employed. 

 The difficulty here is, no uniform laws have been drawn up set- 

 ting forth for birds exactly what characters constitute specific 

 characters, what generic characters, family characters, and so on 

 up to Class characters. The consequence is that one avian 

 classifier will employ subclasses in his scheme, which subclasses 

 are designated in the scheme of another as super orders, or even 

 as orders or some other divisional value in the scheme of a third 



