220 THE CONDOR Vol. XXV 



of North American Birds. To scatter closely connected, sometimes scarcely- 

 distinguishable forms, among others most remotely related, is to obscure the 

 facts. All recognizable subspecies are not equidistant from one another, and 

 they should not be treated as if they were. As more and more material ac- 

 cumulates, permitting careful study of variation in one species after another, 

 facts eomie to light that modify our conceptions of racial differences and our 

 standard books should reflect such advances in knowledge. The application 

 of this in the present instance is to urge, in any new edition of the A. 0. U. 

 Chech-List, or any other general work on North American birds, a modifica- 

 tion of the present mode of listing the local races of such birds as Melospiza 

 melodia. In place of a string of heterogeneous subspecies, listed according 

 to a chronological system true enough, but absolutely at haphazard as regards 

 relationship, degree of difference, or geographical position, I would urge that, 

 where possible, stress be placed upon the main divisions of the species with 

 some subordination of the differences within each such section. 



It is perhaps permissible to speculate somewhat as to the relationship 

 borne by rufina and morphna toward caurina, immediately to the northward. 

 R'ufina in certain characters occupies an intermediate position between morphna 

 and caurina, but geographically morphna (as here defined) imipinges more 

 directly upon the habitat of caurina than rufina does. It may seem startling 

 to affirm the close resemblance of the Queen Charlotte Mands song sparrow 

 to that of Yakutat Bay (see Brooks, Auk, xl, 1923, p. 223), but there is no 

 gainsaying the similarity in size and, to some extent, in color and in shape of 

 bill. Study of the map will demonstrate that the close relationship implied by 

 such a claim is not so unreasonable as it first appears to be. Directly south of 

 Glacier Bay extend the western islands of the Alexander Archipelago, and 

 south of them the Queen Charlotte Islands. In many respects the animal life 

 of the western islands differs markedly from that of the inner islands and the 

 mainland, and it is easy to conceive of the dispersal of song sparrows in such 

 a way as to produce different lines of variation in the two regions, west and 

 east. Thus there might be produced in dispetsal southward from the main- 

 land, on the western islands and on the Queen Charlottes, a modification of the 

 caurina type in which size remained the same, bill remained as long and 

 nearly as slender, and color became darker and more brown. Morphna, in 

 similarly hypothetical dispersal northward along the mainland coast, became 

 slightly darker colored but otherwise almost unchanged. Something of the 

 sort is what I believe really happened, with the resulting implication that 

 caurina and rufina may be more nearly related, than are morphna and 

 rufina. 



There are two specimens at hand that have a decided bearing upon this 

 theory, two breeding birds from Glacier Bay. One of these (Mus. Vert. Zool., 

 no. 514, female, July 16, 1907), is an example of caurina. It may be taken as 

 indicative of the southern extremity of the breeding range of that subspecies. 

 The other (Mus. Vert. Zool., no. 512, male, July 2, 1907), is typical, not of 

 rufina, but of morphna. In small size, ruddy color, and stubby bill, it is like 

 breeding birds from Vancouver Island, rather than like those from the western 

 islands of the Alexander Archipelago. These, the only adiilts at hand from 

 Glacier Bay, are too few specimens to permit of any finality in the deductions 

 made therefrom (especially in view of the variation exhibited in series from 



