270 NEW YORK STATE MUSEUM 



shaft of tarsometatarsus, as in many birds, but attached to a liga- 

 mentous structure stretching between the lower part of the hypo- 

 tarsus and the trochlea above mentioned. 



The hind toe which it supports is fully developed, with basal joint 

 and claw, though it is proportionately much smaller in comparison 

 than the three anterior toes with their large joints. 



These latter need no special description, they are articulated and 

 fashioned as in the anserine fowl generally, and exhibit the most 

 usual arrangement in regard to number of joints allotted to the 

 several toes. We may fancy that a certain amount of lateral com- 

 pression is present in the phalanges of these podal digits, but if it 

 is so, it is very slight, being little more in degree than is enjoyed by 

 like skeletal parts in the feet of the Anatinae. 



Before passing to the consideration of the osteology of this sub- 

 family, I would say, judging from its skeleton, that Lophodytes 

 c u c u 1 1 a t u s is more closely allied to certain ducks than is 

 M erg us serrator. It is seen in the form of the sternum 

 and pelvis ; in the increased length of scapula in the bones of the 

 shoulder girdle ; and in numerous minor points to be readily found 

 in the appendicular skeleton. 



ANATINAE 



At the time I wrote my Observations upon the Osteology of the 

 North An-erican Anseres, cited above, the account consisted in 

 the main of a description of the skeleton of Spatula cly- 

 p e a t a , comparing it more or less fully with the skeleton of 

 Clangula islandica, and in some extent with Anas 

 p 1 a t y r h 3^1 c h o s and two z^merican teal ducks — N e 1 1 i o n 

 c a r o 1 i n e n s i s and Q u e r q u e d u 1 a disc o- r s . As Spat- 

 ula, in the matter of its osteology, is a good average duck, present- 

 ing nothing peculiar beyond its unusually developed mandibles, that 

 same account will be used again in the present connection. It will, 

 however, be very much amplified, owing tO' the fact that my material 

 now includes skeletons of a number of other species, representing 

 several genera, as Oidemia ; two species of Somateria ; Polysticta; 

 Clangula; four species of Marila; Aix ; Dafila; four, species of 

 Anas ; Netta ; and others. There seems to me to be no special 

 reason why Anas p 1 a t y r h y n c h o s should be chosen as the 

 typical species to furnish, as it were, the basic description for the 

 anatine skeleton, and the only advantage that it may claim over 

 Spatula, would be the fact that the Mallard is the ancestor from 



