THE GENUS CHOEDEILES SWAINSON — OBERHOLSER. 17 



differences which, upon examination of proper series, are apparent 

 at a glance ; and there are also other geographic variations, not very 

 evident, except by careful study, and which may or may not have 

 special significance according to the attitude of the investigator. 

 The nominal designation of these small and apparently more or less 

 ephemeral differences may, it seems to me, properly be left until future 

 time and study shall make such a course advantageous. For the pres- 

 ent purpose, therefore, the writer has contented himself with recog- 

 nizing in nomenclature such differences of size, pattern, and color as 

 appear to maintain during the breeding season a fair average differ- 

 entiation and uniformity over an isolated or reasonably different or 

 extensive geographic area. The results, while they thus bring into 

 evidence probably all the significant geographic variations of the 

 several species, at least in so far as the available material warrants, 

 are at the same time not excessive subdivision. Fewer subspecies 

 would obscure interesting geographic variations, with little or no 

 gain in the ease of identification of given specimens; more would 

 seem, from the outlook of our present material, superfluous and 

 inadvisable. 



Identification of specimens. — The naming of subspecies is most 

 important in that it brings into prominence the facts of geographic 

 distribution and evolution and provides convenient handles for their 

 use. In the case of migratory birds it furnishes a means, often of 

 great value, sometimes indispensable, of tracing their wanderings. 



Given any certain number of races, therefore, of a Chordeiles or 

 species of any other genus, with a mass of undetermined material, 

 it naturally follows that proper identification must precede any use 

 of the facts which such material exhibits. This, however, in difficult 

 groups, like Chordeiles, is often a matter of considerable difficulty, 

 and for satisfactory results requires a good series of authentic speci- 

 mens for comparison, coupled with a thorough knowledge of differ- 

 ential characters and variations. In the present genus, so great and 

 so complicated is the variation of individuals, sex, and age, that it 

 is absolutely necessary to make comparisons with specimens of the 

 same sex and age and with typical series. On this account very full 

 comparisons are, in the following pages, made under the head of each 

 subspecies. As is always the case with subspecies, the differences 

 separating which are only average, there are specimens in nearly all 

 the races of Chordeiles^ which individually are not separable from 

 certain individuals of some other race. With breeding birds of 

 this kind identification is in most cases comparatively easy, for 

 if their respective localities are within the known range of any sub- 

 species they are, of course, identifiable as individual variants of such 



