THE GENUS CHOEDEILES SWAINSON OBERHOLSER, 13 



a small, somewhat modified Chordeiles, is a more direct offspring of 

 the latter group. 



The real relationship and therefore the origin of the three species 

 of Chordeiles is a problem of some difficulty. It seems probable, 

 however, that Chordeiles virginianus^ to have become, in structure, 

 pattern of coloration, and color so well differentiated specifically from 

 both Chordeiles acutipennis and Chordeiles rupestris; to have estab- 

 lished itself entirely outside of South America ; to have spread over 

 and permanently occupied as a summer home so large an area ; and to 

 have become modified into so many geographic races, notwithstand- 

 ing the very brief period in each year during which the modifying in- 

 fluences could be operative, must be a relatively ancient species. On 

 the other hand, Chordeiles acutipennis, less stable than Chordeiles 

 virginicunus in structural characters, less dispersed into territory 

 where it is migratory, modified into fewer subspecific forms, and 

 these less strongly differentiated from the parent stock, and still 

 largely resident in South America, seems to be of more recent origin. 

 But Chordeiles tmpestris, which is very different from Chordeiles 

 acutipennis in pattern of coloration and in structure, and which 

 occupies approximately the same area in South America, probably, 

 therefore, developed simultaneously from the same ancestor. Fur- 

 thermore, the close superficial resemblance in color and markings 

 which Chordeiles acutipennis acutipennis from northern South 

 America bears to Chordeiles virginianus minor from Jamaica and 

 Cuba is at least suggestive, if not significant. Thus we naturally 

 arrive at the hypothesis that all three species of Chordeiles were de- 

 rived from a common ancestor, probably resembling Chordeiles vir- 

 giniamus in structure and Chordeiles acutipennis in coloration, but 

 now extinct, from which Chordeiles virginianus first became differ- 

 entiated, and afterwards both Chordeiles acutipennis and Chordeiles 

 rupestris^ the latter so greatly specialized, as it is now seen to be in color 

 pattern, by some unknown agency, possibly individual variation and 

 subsequent fortuitous, temporary segregation. The only region where 

 the two species, Chordeiles virginianus and Chordeiles acutipennis., 

 occupy a common breeding area is in the southwestern United 

 States and extreme northern Mexico. Here the former seems to have 

 immigrated first, and from the north and east; while afterwards 

 Chordeiles acutipennis., then a perfectly distinct species, extended its 

 range northward from southern Mexico until it overlapped that of 

 Chordeiles virginianus., and it seems to be still pushing northward. 



Sequence of forms. — An ideal linear sequence of species and sub- 

 species, which shall show their proper phylogenetic relationship, is, 

 however, desirable, probably out of the question in a complicated 

 group like the genus Chordeiles. As some such arrangement is, of 



