78 University of California PuUications in Zoology [Vol. 21 



As regards the difficulty of identifying the Unalaschcensis sub- 

 species in the winter home in California, there are the following 

 features of the problem to be considered. These races are arranged 

 linearly along the coast in their summer habitats, and the centrally 

 located subspecies intergrade, each one, with the races to the north 

 and to the south. Thus, for example, there are individuals of sinuosa 

 that approach unalaschcensis, and others that are close to annectens. 

 It might happen that a collector in California, in a limited series of 

 birds, would have the two extremes represented. Comparison with 

 sufficient material would show that his two diverse specimens must 

 both be referred to the same form, in this case sinuosa, j^et to his eye 

 they would evidently be quite as unlike as other specimens labeled as 

 of distinct subspecies. This is something that must be borne in mind 

 in the consideration of small series of specimeiLS. 



In the Schistaeea group (the thick-billed and slate-colored spar- 

 rows, so-called) there is another confusing factor. Among these 

 subspecies the most variable feature is the bill (see fig. Dj. While 

 the different subspecies are distinguished by smaller or larger bills, 

 there is also (most noticeably among the larger-billed forms) some 

 difference in this regard between young and old. For example, a 

 young stephensi might have a bill of about the size of an adult 

 mariposae, and a young mariposae one of about the size of an adult 

 monoensis (see fig. DD). It is not possible to tell, of course, wiiether 

 the bill of an early fall migrant, taken far from its summer habitat, 

 has or has not reached the maximum of growth, and in birds that are 

 otherwise alike, it is easy to perceive the source of confusion that this 

 variable feature may be. Aside from this complication there is in this 

 group again undoubted intergradation between subspecies at the 

 margins of their habitats. A thorough understanding of conditions 

 in this section of the genus requires material from many more 

 localities than are here represented. 



Osgood (1909, p. 17) in speaking of the mammalian genus Pero- 

 mysciis, makes the following comments, which, as is seen from the 

 statements in the preceding paragraphs, apply as well to Passerella. 

 "Classification becomes, then, as has been said [by Ridgwaj^, 1901, 

 p. x], like dividing the spectrum and depends largely upon the 

 standards set, for, theoretically at least, the possibilities of subdivision 

 are unlimited. It is not strange, therefore, that hundreds and even 

 thousands of specimens are intergrades almost equally resembling 



