OSTEOLOGY OF BIRDS I 33 



For its upper half the tarsometatarsus is deeply excavated in the 

 longitudinal direction ; the sides of the bone are generally broadened 

 and flat ; and its hypotarsus is in one subcubical piece, which is very 

 shallowly grooved for the tendons behind. None of the basal joints 

 of the three anterior toes of pes are especially abbreviated, and in the 

 case of the third digit, this joint is longer than any of the others of 

 the same toe. 'Skeleton often markedly pneumatic. 



RELATIONSHIPS OF THE ACCIPITRES 



It will be seen from the synopses that have just been presented 

 above, and the provisional classificatory scheme, that our Accipitres 

 will classify very well upon osteological characters even through the 

 subfamilies. To this, however, some exception may be taken for 

 the Milvinae or the kites, and these birds all demand carefu^ 

 anatomical investigation. Elanus, I must believe, is more distantly 

 related with other kites, than the scheme above given would indi 

 cate. Ictinia is highly buteonine in its afifinities, and Elanoides has 

 several important characters in common with Pandion, as the form 

 of its sternum; in having a complete fibula; in having the supra- 

 orbital parts of the lacrymals much reduced; in having a similar 

 hypotarsus to its tarsometatarsus, an unusual character, and several 

 other minor points. 



Taking the group as a whole the Accipitres can be very w^ell de- 

 fined, Ibut their afifinities with other existing groups of birds are by 

 no means so easily made out. Remotely they would appear to be re- 

 lated to the Psittaci. Perhaps they may be their nearest kin among 

 existing birds. As one offshoot, the African secretary-bird fur- 

 nishes most interesting material for study and comparison, and the 

 morphology of Cariama equally good material for speculation. 



From what ancestral type these Accipitres have arisen, is now 

 most difficult to conjecture, but the opinion is growing that the 

 original stock in time started with early types from whence per- 

 haps in common also arose the Steganopodes, and certain ciconine 

 forms as the storks and their allies. 



ADDENDA 



Since the main part of the present treatise has been completed 1 

 have examined the skeletons of a number of the Accipitres, and 

 notably those of Helotarsus ecaudatus [Collec. U. S. 

 Nat. Mus. no. 17836] and of the Lammergeyer (Gypaetus 



