566 Correspondence. [oct. 



tionably have been adopted, but there was not; and the diversity of 

 opinion among avian taxonomists still prevails. Furthermore it should be 

 remembered that a Check-List is by no means necessarily a classification. 

 Its very existence is for convenience, and so it is no light matter to over- 

 throw a sequence, followed by practically all writers on North American 

 birds for a quarter of a century, on the plea of being more scientific when 

 we get nothing more stable than that which we discard. 



As to the subspecies question with which Mr. Grinnell is chiefly con- 

 cerned, we hardly think that he is serious in believing that a list of the 

 binomial names in the present Check-List would answer the needs of the 

 great bulk of the membership of the A. O. U. which he classes as 'amateurs' 

 in matters of subspecific discrimination. He knows perfectly well that 

 there are very many subspecies which are more easily distinguished than 

 are certain species, and for these we must have names. For the purposes 

 of ornithological investigation along any line — life history, habits, geo- 

 graphic distribution, migration, taxonomy, economics, etc. — we must 

 have the birds of the country divided up into minor groups, species or 

 subspecies as you will. The only question is, where shall we draw the line in 

 recognizing the differentiation that nature has effected? The question is 

 a practical one, just as the whole matter of naming is practical, and when 

 we recognize by name differentiations so slight that an ornithologist cannot 

 tell what bird he has before him until he submits it to a 'specialist in 

 epeciation ' f or study, then the process has gone too far for general purposes. 

 There is however no test by which we can tell when we have gone too far. 

 The problem is one entirely of degree in which personal opinion and indi- 

 vidual ability enter into every case. As already stated the line cannot 

 be drawn between the species and the subspecies, because by our Code 

 they are distinguished not by degree of difference but by the criterion of 

 intergradation. In an effort to fix this line the A. O. U. established the 

 Committee believing that the vote of a Committee would represent the 

 nearest approach possible to the desired result. 



We do not believe that the efforts of the Committee have been so entirely 

 unsatisfactory as Mr. Grinnell implies, except of course to 'students of 

 speciation' who make a specialty of naming differentiations no matter 

 how small, regardless of whether the results of their work can be utilized 

 by specialists in the various other branches of ornithology. It was for 

 the latter we think that the Check-List was conceived. It was surely 

 never intended for such a 'specialist in speciation' as Mr. Grinnell predicts 

 who would name every finest discernible differentiation and would if pos- 

 sible make 140 races of Song Sparrows out of the 20 now recognized. In 

 European ornithology the same effort is evident in check-lists and cata- 

 logues to recognize practical subspecies but to reject those based in 

 extremely slight differentiations, and this by ornithologists who can hardly 

 be charged with catering to the amateur. 



Mr. Grinnell will perhaps understand better the attitude of the large 

 majority of ornithologists toward the subspecies if he will but consider 



