6 



BULLETIN 86, UNITED STATES NATIONAL MUSEUM. 



pared with the primary coverts. A tabulation of these results is 

 given herewith: 



Characters. 



First primary: 



Longer than second 



Equal to second 



Shorter than second 



Longer than second on one side but shorter on the other 



Total number of birds examined 



Shortest secondary: 



Longer than primary coverts 



Equal to primary coverts 



Shorter than primary coverts 



Longer than primary coverts on one side but shorter on the other 



Total number of birds examined. 



CfhoTieilef tnr- 

 ginianu8. 



90 (85:7%) 

 7 (6.6%) 

 B (4.9%) 

 3 (2.8%) 



105 



6 (5.3%) 



3 (2.8%) 



103 (91.9%) 



(0.0%) 



112 



Chordeiles acuti- 

 pennii. 



12(11.5%) 

 11 (10.6%) 

 78 (75.0%) 

 3 (2.9%) 



104 



92 (82.9%) 



8 (7.2%) 



9 (8.1%) 

 2 (1.8%) 



111 



The above figures are based on so large a number of specimens that 

 their percentage would probably not be materially changed by the 

 examination of a greater number. From these percentages it will 

 be seen that Chordeiles virginianus is more stable in its structural 

 characters, at least in the two here considered, than is Chordeiles 

 acutipermis ; or, in other words, is farther along in its differentiation. 

 Thus, in Chordeiles virginianus the first primary is longer than the 

 second in 85.7 per cent of the specimens, while in Chordeiles acuti- 

 permis it is shorter than the second in only 75 per cent. Also in 

 Chordeiles virginiamus the shortest secondary is shorter than the pri- 

 mary coverts in 91.9 per cent ; in Chordeiles acutipemiis it is longer 

 than the primary coverts in only 82.9 per cent. Further comparisons 

 can easily be made from the table. Of the two characters here ex- 

 ploited, the length of the secondaries thus seems to be the better. 

 There does not sefem to be any difference in these structural characters 

 among the different subspecies of either Chordeiles virginianus or 

 Chordeiles acutipennis; and the above percentages of the two species 

 include representatives of various subspecies indiscriminately. 



Relationships. — The genus Chordeiles has always without ques- 

 tion been placed in the. family Caprimulgidae. A careful examina- 

 tion and comparison, however, of a series of skulls of Chordeiles 

 virginianus and the skull of Chordeiles acutipennis texensis, with 

 crania of Antrostomus carolinensis, Setochalcis ^ vocifera, Phalaenop- 

 tilus nuttallii, and Nyctidrorrms aZhicollis merriUi, and with Hux- 

 ley's figure of the skull of Caprimulgus europaeus ^ (see pi. 1, fig. 1), 

 reveal differences that make advisable, indeed even necessary, the re- 

 moval of Chordeiles from the family Caprimulgidae. So far as 

 cranial characters are concerned, Chordeiles virginianus (see pi. 1, 

 fig. 2) and Chordeiles acutipennis are generically identical, since 

 aside from some unimportant details, which are surely not of more 



1 See p. 11. 



= Proc. Zool. Sec. Lond., 1867, p. 453. 



