486 THE NUMBER OF PRIMARIES IN OSCINES 



ray inspection by Mr. J. A. Allen, the spurious primary is con- 

 spicuously displayed, exposed for a third of an inch.* 



* Believing thia to be au important matter deserving of furtlier investiga- 

 tion, I was not long since led to examine the general question, with satisfac- 

 tory results. I verified Professor Baird's observations in many more cases, 

 extending them to include all our North American families excepting per- 

 haps Laniidw (in Lanius) and AvijyelidcB (in Amj)elis). The clue to the 

 search for the apparently wanting primary was given by Baird (Review, pp. 

 160, 325), from which it ai)pears that in all those Vireos which seem to have 

 only nine primaries, tivo little feathers, distinct in size and ehape and some- 

 what so in position, are found at the base of the supposed first primary ; 

 while in Vireos, with obviously ten primaries, there is only one such little 

 feather. With the possible exception of Am^elis and Lanius, in which I 

 did not make out the state of the parts satisfactorily, I find that in all of 

 the numerous North American genera examined, those of ten primaries show 

 but one of these little feathers, while the rest have two. In the family J iaw- 

 didce, as in Vireonidw, some genera have ten jirimaries, others apparently 

 only nine ; and in our genus Eremophila, in which only nine are developed, 

 there are two of the little feathers just mentioned, the overlying one being 

 exactly like one of the primary coverts, the other, though not very similar, 

 more resembling an abortive primary. Alauda arvensis, which shows a 

 minute but obvious spurious quill, has but one such little feather; and in 

 Galerita cristata, with a spurious quill about two-thirds of an inch long, there 

 is likewise but one. In Clamatorial Passercs, jjerhaps without exception, 

 there are ten fully developed primaries, the first of which may equal or 

 exceed the next in length ; and in the single North American Clamatorial 

 family, Tyrannidw, I find, as before, only one of these little featliers. In a 

 Woodpecker, remarkable among Picariau birds for possessing only nine long 

 primaries, the first being short or spurious, there Is also but one. 



It thus seems to be established that among supposed nine-primaried birds, 

 the additional one, making ten in all, is normally represented by the second 

 one of these tiny quills which overlie the base of the outermost fully devel- 

 oped feather; it being this same little quill which in ten-primaried Oscines, 

 in Clamatores, and j)robably other birds, comes to the front and constitutes 

 the first regular primary, either remaining quite short, when it is the so-called 

 "spurious" primary, or lengthening to equal or exceed the other primaries 

 in extent. 



It becomes an interesting question whether hoth of these minute quills be 

 not rudimentary primaries, as one of them certainly is. I have failed to 

 detect any material difi"erence between the two in size, 8hape,ox position. 

 One overlies the other, indeed, as a covert should a primary, but the two are 

 together inserted side by side on the upper side of the first fully developed 

 quill ; both are rigid and acuminate, more like primaries than like coverts, 

 and both are abruptly shorter than the true primary coverts. So far, all 

 the evidence favors the supposition that both are rudimentary primaries. 

 On the other hand, coloration is against such hypothesis, as in the original 

 case of Vireo flavifrons, in which Baird determined the underlying one of 

 these two little feathers to be the missing primary, mainly because it was 

 colored like the primaries, the overlying one resembling the coverts in colora- 



