PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 177 



When folded and lying close to the body their elbow -joint reaches far 

 beyond the acetahulmn. Their length is greater than that of the hand, 

 and considerably more than twice the tarsus. 



The 2)€ivis presents only few difi'erences from that of the other Ana- 

 tidce. It is, however, proportionally louger and narrower, the breadth 

 between the acetahula making only about one-fifth of the total length of 

 the ilia. The very prominent ridges, forming the internal borders of 

 the post-acetabular parts of the i/ia, run from the acetahula backward 

 nearly parallel, the hinder sacral roof being rather narrow and of equal 



Spatula clypeata. — No. 1122, Ai-my Med. Mus. Nat. size. 



breadth, whereas in the other Anatidw, the ridges converge backwards 

 from the more distant acetahula^ forming a wedge-shaped and rather 

 flat and broad roof. The foramen ischiadicnm is rather large. 



The -pelvic limbs agree in every respect with those of the typical mem- 

 bers of the family, except in the proportional length of the single bones 

 mutually. 



The number of free caudal vertebra' is seven, to which is to be added 

 the pygostyle. 



The Cygnince are more nearly related to the Anatinw than to the An- 

 serine, the Coscoroba Eeichb. being among Ducks the genus most closely 

 allied to the Swans. It has also, as stated above, usually been referred 

 to the latter group, but an accurate examination undoubtedly shows 

 that Coscoroba belongs to the Cygnince as little as does Cairina to the Anser- 

 ince, and that Sundevall was right in removing it from the Swans. 

 Proc. Nat. Mus. 82 12 July 25, 1 883. 



