STRUCTURAL CHARACTERS AS MEANS OF 



IDENTIFICATION 



[W. P. PYCRAPT] 

 INTRODUCTORY 



WHETHER we are discussing the Class Aves as a whole, or merely such species as are 

 regarded as " British," some definite plan must be adopted. One must either follow the 

 dictionary system, or some scheme of classification. To the former there are many objections ; 

 and the choice of the latter is by no means easy, as those who are even superficially acquainted 

 with this subject must know. It becomes necessary, then, to choose between a purely arbitrary 

 system, as into land birds and water birds, and so on, or into birds with hard beaks and soft 

 beaks, and so on, after the fashion of the older systematists, or a system based, as nearly as 

 we can, on natural affinities, on blood-relationship and not superficial likeness a phylogenetic 

 system, in short. 



There can be no doubt but that this is the only true system of classification, even though, 

 at present, opinions differ widely as to the degrees of relationship, thereby leading to more 

 or less strikingly different sequences in the order of different groups, large and small. It 

 was therefore decided to follow, in these volumes, in the main that formulated by Dr. Gadow 

 and employed by Mr. Evans in his most invaluable work on "Birds," published in the 

 Cambridge Natural History. For all departures from this arrangement I am myself respon- 

 sible, the editor having alleged me this privilege. In this, however, I am referring to the 

 major divisions Orders, Suborders, and Families. In the matter of Genera the case is far 

 otherwise. Here it became impossible for me to impose my own judgment, owing to the 

 breach of continuity it would have occasioned with all other similar works, and the consequent 

 inconvenience to those using these volumes. I have had to adopt genera which no one 

 with even an elementary grasp of the meaning and principles of classification would dream 

 of making. 



There are few better illustrations of this lack of appreciation of what should constitute 

 a genus than are to be found in Howard Saunders Manual of British Birds. Many of the 

 characters there given are meaningless, and more are utterly inconsequent, as when he sagely 

 remarks in the case of some genera of Ducks that the toes are webbed, or in the case of some 

 of the Alcidse, when we are told that there are " three toes, all in front, and webbed " ! Let any 

 one analyse the formidable array of generic characters he gives as distinctive of Passer, 

 Fringilla, and Linota, and then estimate how many of these are really points of difference, and 

 what is their value. 



No attempt is made to furnish diagnostic characters for the Orders, Suborders, Families, 

 and Subfamilies, though such divisions are indicated by headings. He depended mainly, as 

 most ornithologists depend for systematic characters, on colour, and this is a factor which 

 should be ignored, if classification is to be framed on sound, scientific lines. When groups 

 of species display a common type of coloration they should be made sections (a), (6), (c), etc., 

 of one genus based on structural characters. 



558 



