958 PERNIS. 



formed me that characters could never be derived from 

 those organs, and another told a friend of mine that 

 feathers were of no value in this respect, and that, even 

 if they were, colour was much better. Yet both the 

 above statements are incorrect; for neither have the 

 buzzards a dense covering- of white feathers, over which 

 are bLick hairs, nor do their loral feathers exhibit any 

 particular affinity to those of Pernis. The latter are 

 small compact feathers without bristle- tips, the former 

 small downy feathers with the shaft prolong-ed into a 

 bristle or hair. For the accuracy of this I appeal to 

 my readers and the birds in question ; and for any er- 

 rors which I may commit I request no indulg-ence what- 

 ever, being convinced that plain truth only can advance 

 our knowledge of either buzzards or kites. 



In my opinion, the nearest approach to the honey 

 pernis is the common kite, and the nearest allied genera 

 to the present, Buteo on the one hand, and Milvus on 

 the other. The bill is so much alike in all that, so far 

 as it is concerned, the three genera might very well be 

 united. Our pernis is also allied to the kite in its feet, 

 as well as in its long wings and elongated tail, although 

 the feathers of the latter organ are differently propor- 

 tioned. 



Of all the other alleged species of this genus I can- 

 not speak with decision ; but the Pernis cristatus of 

 Cuvier seems to me to be a buzzard, as it has the bill 

 kti'ger, the loral feathers bristle-tipped, and the tarsi 

 scutelJate. 



