

21 



their palatines from the Drepanididse is another weighty argument against their 

 affinity to that family. 



Now to sum up : although these remarks are scanty, necessarily incomplete, and 

 consequently premature, on account of the want of anything like a sufficient amount of 

 suitable material, I consider that the Drepanididse form a separate family of the Fringil- 

 liformes rather than of Meliphagine birds or even of the Cinnyrimorphse, and that of the 

 Fringilliformes they are nearest allied to the Coerebidse, i. e. to the Neotropical and 

 Central-American families. Thence to the Fringillidse is a long way, but we can 

 imagine the intermediate stages. Loxioides and Psittacirostra I judge to be Fringillidse, 

 while I consider that there is no direct connection between these two genera and the 

 Drepanididse. None of these forms can be included among the Dicseidse, which are 

 an essentially Old-World family. 



If the numerous resemblances between the Drepanididse and the Meliphagidse are 

 not all merely coincidental — and they cannot be explained away at all satisfactorily — - 

 then the large group of the Cinnyrimorphse (through the Meliphagidse and possibly 

 through the genus Zosterops, unless these birds connect the Nectariniidse in another 

 direction) and that of the Fringilliformes (through the Drepanididse and Coerebidse) 

 converge to form a still larger group. How many other families will ultimately be 

 found to gravitate towards the same centre must be left to him who may be favoured 

 with an exhaustive supply of spirit-specimens, and will not shrink from devoting much 

 time and labour to their examination. Whenever we endeavour to study seriously even 

 a few different Oscines, the attempt is apt to assume enormous dimensions. The 

 examination of a small twig of the Passerine branch of the Avine tree shakes and 

 disturbs the whole branch, if not the whole top of the famous ideal tree. At any rate 

 we seem in our case to get a glimpse of one of the bigger ramifications of the Oscine 

 portion ; and although, at first sight, the idea of a Fringilli-Cinnyrimorphous branch 

 appears rather appalling, it is after all not more diversified than another branch, which 

 is composed of the Corvidse, Laniidse, and Muscicapidse. A Raven and a Flycatcher 

 do not seem to have much in common, but with the help of the Austrocoraces and 

 other tropical forms the differences fade away and vanish. 



LIST OF ILLUSTRATIONS (Plates L-IIL). 



Figs. 1-5. Ph^ornis obscura. 



1. Dorsal, 2. Lateral view of nostrils and bill. Nat. size. 



3. Tongue, dorsal view. Nat. size. 



4. Ventral view of the bones of the palate ; enlarged, pt. — pterygoids ; max.pl. — maxillo- 



palatines; a.int.pl. — anterior or interpalatine spur or process. 



5. Right-sided view of the intestinal convolutions, beginning with one near the pylorus and 



ending with eight at the anus, 



