20 



The shape of the tail can scarcely be considered in earnest, although its Fringilliform 

 appearance in the Drepanididae is obvious. 



The scutellation of the tarsus likewise permits of no safe generalization applicable 

 to families and not to genera only. 



Concerning the alimentary canal, the possession of a distinct ventral crop, or at least 

 an obvious dilatation of the oesophagus, by the Drepanididae is unquestionably a Fringil- 

 liform character, because of all the Oscines only some Fringillidae have hitherto been 

 known to possess a crop. The crop of the apparently strictly-insectivorous Drepanididae 

 is therefore all the more remarkable, although the Trochilidae have it likewise well- 

 developed. 



The peculiar intestinal convolutions of the Drepanididae cannot unfortunately be 

 brought into comparison, owing to want of material in other groups, notably Ccerebidae. 

 The latter are insectivorous, to judge from their tongues strictly so ; many FringillidcB 

 are granivorous or have a mixed diet : hence the resemblance between several of 

 the Drepanididae and Meliphagidae is not decisive. The bones of the palate of the 

 Drepanididse, especially of Hemignathus, and in a less degree of Vestiaria, are most 

 peculiar. The vomer is posteriorly completely fused w 7 ith the palatines, and the 

 lanceolate space between the two halves of the basal or dorsal parts of the palatines 

 is closed by a transparent plate of bone, which covers, and rests upon, the sphenoid. 

 Such a truly interpalatine plate occurs in many Fringillidae. The ventral palatine spurs 

 (Parker's interpalatine spurs) are very high and slender, and posteriorly extended 

 to such an extent that they project far beyond the level of the articulation of the 

 pterygoids. The latter articulate with the palatines, and not with the sphenoid at all, 

 by distinct cartilaginous feet, resembling in this respect again certain Fringillidae, 

 e. g. Coccothraustes, although in the latter these feet are bony and liable to fuse with the 

 palatines. The transpalatine or posterior lateral spurs are long and very slender. The 

 maxillo-palatines are long and slender, passing ventrally over and past the anterior 

 fork of the vomer and touching with their tips the anterior interpalatine spur; they 

 rather resemble the same parts of Loxioides and of Coccothraustes : we must, however, 

 bear in mind how much these little bones vary in shape, size, and position even in 

 the various Fringillidae, as a glance at the numerous illustrations in Parker's work 

 on the vEgithognathous skull will show. The whole arrangement of the bones of the 

 palate of the Drepanididae is totally different from that of the Meliphagidae and other 

 Cinnyrimorphae, and can only be compared with that which is indicated to a small 

 extent in some Fringillidae (Coccothraustes, Cardinalis, Estrilda). The configuration 

 of these bones in the Drepanididae looks as if it were derived from Fringilline conditions 

 by reduction of the width (not length) of the palatine expansions, not vice versa. This 

 may well be the case considering the lesser strength of the masticatory muscles in the 

 Drepanididae in comparison with that of typical Conirostres, and considering that the 

 elongated bill of the Drepanididae is undoubtedly not a primary feature but a secondary 

 specialization. That Loxioides and Psittacirostra differ so much in the configuration of 



