BIRDS OF NORTH AND MIDDLE AMERICA. 375 



tarsi, with tlie posterior face smootii, and with indications of several 

 scutes more or less obsolete on the anterior face. The wings are of 

 moderate length, without any indication of the first or external pri- 

 mary, and with the second, third, and fourth primaries usually nearly 

 equal and longest. The bill is small and slender and without any notch 

 at the extremity of the upper mandible, but varies in structure . . . 

 The tongue is penciled at the extremity. " ^ 



It is at present not possible to give a satisfactory diagnosis of the 

 Family Coerebidse since the internal structure of more than half the 

 genera remains practically uniinown. What little is known of the 

 anatomical characters of the group pertains to the genera Cyanerpes^ 

 Cmreba, and Glossiptila; and since these maj' be regarded as the cen- 

 tral or most typical forms of the group, anj^ conclusions based on them 

 alone would be more or less open to doubt; at least until the internal 

 structure of such forms as Conirostrum, Diglossa, Oreovmnes, and 

 Xenodacnis is known, the limits of the group can not be fixed with 

 precision, nor its relationships satisfactorily determined. 



The limits of the family, as defined by Dr. Sclater,^ have already 

 been restricted by the elimination of the genus Certhidea, which Mr. 

 Lucas, by examination of its osteology and anatomy, has found to be 

 certainly not Coerebine, but probably Mniotiltine; ^ and it is by no 

 means improbable that further contraction may ultimately be required. 



So far as the tj'pical genera, Coereha, Glossiptila, and Cyanerpes are 

 concerned, Mr. Lucas finds them to represent a well-circumscribed 

 group, of uncertain affinities, though apparently more nearly related 

 to the Australasian family Meliphagidse (Honej'--eaters) than to the 

 American families Mniotiltidse and Tanagridse, usuallj^ held to be 

 the nearest relatives of the Coerebidse. The gist of Mr. Lucas's con- 

 clusions ' is as follows: 



(1) "As T'-oups of birds are constituted the CcBrebidas are certainly 

 sufficiently' distinct to stand apart, and the gap between them and the 

 Mniotiltidas seems widest,* although this may be due to a tendency on 

 my part to place considerable weight on the general pattern of the 

 palate." 



'Sclater, Cat. Birds Brit. Mus., xi, 1886, 1. 

 ^Cat. Birds, Brit. Mus., xi, 1886, xi. 

 'Proc. U. S. Nat. Mus., xvii, 1894, 309, 310. 



* Notes on the Anatomy and Affinities of the Coerebidse and other American Birds, 

 by Frederic A. Lucas, Curator of the Department of Comparative Anatomy. Proc. 

 U. 8. Nat. Mus., xvii, 1894, 299-312; illustrated by many figures. 



* It would be interesting to knowwhether Mr. Lucas's conclusions on this point 

 would have been modified by examination of Conirostrum and Ateleodacnis, genera 

 which have hitherto been referred to the Cffirebidas (the latter forming part of the 

 genus Dacnis) but which I have found it necessary to refer to the Mniotiltidse. Possi- 

 bly, as in the case of another supposedly Ccerebine, but in reality Mniotiltine, genus 

 (CeHhidea), the gap between the two groups would have been emphasized. (See 

 Lucas: The Anatomy and Affinities of Certhidea, Proc. U. S. Nat. Mus., xvii, 

 1894, 309, 310.) 



