THE GBNTJS OHOBDBILBB SWAINSON — OBEBHOLSEB. 7 



than specific value, the skulls of these two species are the same. 

 Both Chordeiles virginianus and Chordeiles acutipermis are per- 

 fectly desmognathous,^ as the figure of the former" and the pre- 

 viously published figure of the latter* clearly show, which charac- 

 ter alone in such a complete exemplification should be sufficient 

 as a basis for recognition of a distinct family. In both species the 

 maxillo-palatines meet on the median line and become ankylosed, 

 at least in old birds, with each other and with the vomer. There is 

 thus no doubt about their complete desmognathism. Besides this, 

 however, Chordeiles differs from all the above-mentioned genera of 

 Caprimulgidae in many other respects, as follows : Premaxillary (to 

 cranio- facial suture) relatively as well as actually much shorter, only 

 37 per cent of the total length of the skull, whereas in Setochdlcis 

 vocifera it is 49 per cent; this, together with the actually as well as 

 relatively much higher braincase, giving to the skull in general a 

 curious humpbacked, pug-nosed appearance; interorbital portion of 

 frontal region very much broader, medially depressed and fur- 

 rowed, but laterally rising in a crest to the edge of the orbits, while 

 in the other genera examined just the reverse is the case, the frontal 

 region being rather inflated, with the orbital edges more or less 

 rounded; vomer relatively narrow and slender; maxillo-palatines 

 subtriangular instead of more or less strap-like or shoe-shaped ; mid- 

 dle portion of palatines much broader, the lateral outline thus less 

 constricted medially, more nearly straight, giving to the posterior 

 portion a decidedly less flaring effect; antero-internal processes of 

 palatines reaching to the maxillo-palatines ; palatal opening between 

 the halves of the premaxillary much smaller, narrower, and somewhat 

 narrowly triangular instead of oval in shape ; gall bladder absent ; * 

 rictal bristles wanting. 



Unfortunately I have been unable to examine the skulls of the 

 South American genera commonly referred to the Caprimulgidae, so 

 that it is impossible to determine with exactness which of the above 

 characters are to be regarded as super'generic. Doubtless some of 

 them will prove, upon examination of all the remaining groups, to 

 be merely generic. The same lack of material renders doubtful the 

 precise limits of the family group which includes Chordeiles. The 

 nearest ally of Chordeiles is without much doubt Nannochordeiles 

 Hartert, and this certainly belongs to the same family, while Nycti- 

 progne is probably also hot distantly related. From external char- 

 acters it would perhaps be better to include both Lurocalis and 



iThe statement of Beddard (Structure and Classification of Birds, 1898, p. 239) that 

 In Ohordeilea virginianus the maxillo-palatines do not meet across the vomer is doubtless 

 a slip or the result of examining an imperfect, abnormal, or immature skull, for in every 

 one of the half dozen skulls we have examined these bones meet just as In Chordeiles 

 acutipennis. 



' See plate 1, flg. 2. 



' Shufeldt, Proc. Zool. Soc. Lond., 1885, pi. 59, flg. i. 



• According to Beddard, Structure and Classification of Birds, 1898, p. 242. 



