1894. PROCEEDINGS OF THE NATIONAL MUSEUM. 569 



Avill be easily seen that should this fusion increase at each end and 

 continue until the apteiiuni had disappeared, we would arrive at the 

 Caprimulgine condition. May it not be true that in this way the single 

 tract of the Capriniulgi has been formed? At any rate there is noth- 

 ing inherently improbable in the idea. If this be granted the conclu- 

 sion is inevitable that the fundamental plan of the pterylosis was 

 originally the same in both grou])s, and since in those minor points in 

 which they differ (such as the longitudinal rows on the head in the Cap- 

 riniulgi and the feathering of the tarsus in the Striges) there are numer- 

 ous intergradations, there is certainly reason to admit the possibility of 

 some relationship. It may be mentioned here that Strlv shows 

 another peculiarity which is not unlike one of the features of Chor- 

 deiles. It will be remembered that the lower cervical fork of the latter 

 genus shows a i)rominent inner branch and although this is not found 

 in any owl, yet the peculiar formation of the tract in Striv gives us a 

 hint as to its possible origin. The above hypotheses in regard to the 

 origin of the single sternal-ventral tract in the Caprimulgi ami the 

 inner cervical tooth in Chordeiles seem to indicate greater specialization 

 on the part of this group and it is worth while to bear this in mind as 

 we consider the other characters to be compared. 



Arrangement of the feathers of the wing. — Since the wing is one of 

 the most characteristic organs of a bird, and since variations in the 

 arrangement of its feathers are almost endless even within the limits 

 of well-defined groups, similarities in these points must carry con- 

 siderable weight, especially when these likenesses are in such details 

 as the comparative length of the primaries. If we compare the 

 wing of an owl with that of one of our goatsuckers, we find sub- 

 stantial agreement in several points, but some apparently important 

 differences in others. Both agree in being aquincubital, a character 

 which seems to be of the greatest im]>ortance. They agree further 

 in the number of rows and distribution of the coverts and even fairly 

 well in the comparative lengths of the primaries. Thus among the 

 Caprimulgi, the eighth and ninth are about equal and are longest; 

 then follow the seventh and tenth (about equal), and the sixth, very 

 little shorter, and then .5, 4, 3, 2, 1. Chordeiles (according to Cones) has 

 the tenth equal to the ninth and the rest in regular succession. Among 

 the owls examined there are three groups; Asia and Speotyto agreeing 

 very closely with the whip-poor-will as above given, the eighth and ninth 

 being about equal, seventh next, sixth and tenth (equal), and the rest 

 in regular order; ASVr/.r differing from these and approaching Chordeiles 

 in the greater length of the tenth primary, which is longer than the 

 seventh and almost equals the ninth; Glaucidium. and Megaseops show- 

 ing a very different arrangement with the seventh longest and the 

 tenth about equal to the first. Striges, however, possess the eleventh 

 primary in a rudimentary condition, while it has completely disappeared 

 in the Caprimulgi. The latter have only three feathers in the alula, 



