THE GENUS CHOEDEILES SWAINSON — OBERHOLSER. 5 



the tip of the sixth primary ; the secondaries as in Chordeiles acuti- 

 pennis; the tail-feathers as in Chordeiles virginiarms; skull not 

 examined. 



It is thus apparent that Chordeiles virginianus and Chordeiles 

 acutipennis differ in structural characters, which, were they trenchant 

 and constant, instead of only average, would clearly warrant generic 

 separation. But Chordeiles rupestris appears to be practically inter- 

 mediate between these two species, or, to be more exact, partakes of 

 the structural characters of both, though in pattern of coloration it 

 is widely different from either. In the small series examined the 

 characters of Chordeiles rupestris, as above given, are pretty con- 

 stant, but in a larger series greater range of variation would doubt- 

 less be evident. The present case, therefore, might appear at first 

 sight to be an ideal one for the profitable employment of sub- 

 genera ; since if anything more than personal caprice is to determine 

 what is to be considered a genus and what a subgenus, it is probably 

 to be found in the principle of intergradation, a principle of general 

 acceptance as applied to species and subspecies; by which a full 

 genus would be a group trenchantly defined by structural char- 

 acters, even though these differences be relatively slight; and a sub- 

 genus one the structural diagnostic characters of which intergrade 

 through intermediate species with some other group, even if the 

 extreme differences be relatively great. In the present case, however, 

 we have a rather peculiar and unusual condition, for while there is 

 more or less connection between the three species which would, by 

 the above criterion, possibly form three subgenera, the intergrada- 

 tion is complicated by the fact that all the structural characters 

 separating the species are, in Chordeiles virginianus and Chordeiles 

 acutipennis at least, decidedly inconstant. To consider instable struc- 

 tural characters of this sort as the basis for subgeneric separation 

 seems to the writer certainly undesirable. 



These birds have, however, a decided interest for the student of 

 evolution, because it is very evident that we have in these three species 

 a case of genera in process of development. Unfortunately we have 

 not sufficient osteological material to determine the amount of varia- 

 tion in cranial characters; but to show, as far as possible by actual 

 figures, the constancy of the external structural characters separating 

 Chordeiles virginianus and Chordeiles acutipennis, and thus to indi- 

 cate the state of their relative evolutionary divergence, the writer ex- 

 amined a large number of specimens of each species and counted 

 the individuals showing the different variations of the two most 

 tangible characters, i. e., the comparative length of the outermost 

 primary, and the relative length of the shortest secondary as com- 



