OSTEOLOGY OF BIRDS 289 



The postpubis is a slender rod as it passes beneath the obturator 

 space, but after its articulation with the ischium posteriorly it has 

 its width nearly doubled, and in Clangula the hinder ends are 

 slightly enlarged. This latter duck departs from the above 

 description principally in such a minor detail as having a relatively 

 much larger ischiac foramen and longer obturator space. 



In all of these species we find the pelvic basin upon the ventral 

 aspect very capacious, both as to its depth and width. 



Be it noted here as a matter of interest and comparison that the 

 pelvis in Olor has a very different fomi from that bone as we 

 find it in the ducks. It assumes a shape that at once brings to 

 our mind the mergine pattern, with its greater length as compared 

 with its width; the almost entire disappearance of the interdia- 



I'"iG. 22 Left lateral aspect of the pelvis of Spatula c 1 y p c a t"a . Natural size; 

 same specimen as figure 21 



pophysial foramina, and the broad, paddle-shaped extremities of 

 the postpubic elements. This model sees its extreme modification 

 in the Pygopodes ; and if we remove the intrasternal chamber for 

 the accommodation of the tracheal loop, we find in the sternum, 

 too, of the swan a great deal to remind us of that bone in Gavia. 

 In the genus Anas I find that in all the species the pelvis has 

 much the same form that it has in Spatula'clypeata. So 

 it may be said also of Dafila acuta, and of Aix sponsa. 

 A departure is seen in N e 1 1 a r u f i n a , however, for in this 

 species the pelvis is long and narrow, and although it possesses 

 the common anatine characters, it is distinguished by the unusual 

 size and number of the interdiapophysial foramina in the post- 

 acetabular region, and by the elongation of the pubic elements 

 posteriorly, where they curve downward and inward, and have 

 their extremities expanded somewhat, as is the case in the swans. 



