40 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4TH Ser. 
Dr. Cougs (Auk, 1897, p. 314) 
Diomedeidz 
Fulmarinz 
k ee Puffininz 
Order Tubinares 4 Procellariidz Procellariine [=Thalassidromine] 
Oceanitinze 
Pelecanoididz 
Dr. Coues’s classification of these higher groups appears to 
me to be the most satisfactory, and I have therefore adopted 
it in the present paper. It is well to recall that Dr. Schlegel 
treated’ the albatrosses, petrels, and diving petrels as three 
genera, designating them respectively, Diomedea, Procellaria, 
and Halodroma. 
In genera, I have followed, with one exception (Thalasso- 
geron), the conventional genera of the monographs of Mr. 
Salvin and Dr. Godman. I feel, however, that too many have 
been recognized, especially monotypic ones founded on minor 
structural characters of species. Subgenera are given no place 
whatsoever. Where convenience ceases in a classification that 
is largely arbitrary, it is maintained that no vantage ground 
is gained by burdening the memory with a multitude of minor 
divisions and subdivisions. 
While I believe that both superficial and deep-seated charac- 
ters should be utilized in the definition of higher bird groups, 
I heartily agree with Dr. Reichenow? that the genealogy of 
birds is a subject to be considered apart from bird classi- 
fication. 
The Subspecies Question.—In theory, subspecies are incip- 
ient species; in fact, subspecies are attempts to forecast the 
future of geographic variation, which no one can foresee. 
Naturally, much difference of opinion has arisen in the appli- 
cation of the subspecies theory. Some ornithologists would 
differentiate all discernible geographic variation into sub- 
species; others would make selections and have “practical 
subspecies.’” Under the first method the separations become 
so fine that even typical examples can scarcely be determined. 
1 Mus. ee v. 6, Procell., 1863, pp. 39, 40. 
2 Die Vogel, v. 1. 
