294 



NEW YORK STATE MUSEUM 



Now Clangula i s 1 a n d i c a has a sternum of an en- 

 tirely different form from the bone as I have just described it for 

 Spatula and the teals [see fig. 26]. In the first place, its body 

 is relatively much shorter for its width than it is in those ducks, 

 while in front the manubrial process has entirely disappeared. 

 Again, the costal processes are loftier and more conspicuous. The 

 xiphoidal extremity of the bone is very broad and is pierced well 

 within its hinder margin, on either side, by an elliptical foramen. 



It will also be observed that the 

 carina does not extend the entire 

 length of the sternal body, but 

 stops short at the middle point of 

 a raised line, that, being produced as 

 it is, is tangent to the posterior 

 arcs of the xiphoid fenestrae. 



The muscular lines take about 

 the same course, with the excep- 

 tion that their posterior ends are 

 inclined inward rather than out- 

 ward, as they are in Spatula. 



This form of sternum agrees in 

 many particulars with the bone as 

 we find it in Mergus, though in the 

 eider ducks, as I shall soon point 

 out, the xiphoid extremity is deeply 

 two notched [see fig. 2j]. 



Not only is the pattern of the 

 sternum essentially the same in 

 the genus Anas and Spatula, but we 



Fig. 2 S Pectoral view of the sternum find tllC Same form Poetically re- 



of Spatula ciypeata. Natural peated in D a f i 1 a a c u t a , in A i X 



size; same bone as shown in figure 24 1 , * , , 



s p o n s a , and probably in other 

 typical fresh-water ducks. In Netta rufina this pat- 

 tern is distinctly changed, and I find that in it the manubrial 

 process is aborted; there are seven haemapophysial facets upon 

 either costal border ; the bone as a whole is relatively shorter and 

 broader; and lastly, the xiphoid notches are almost converted into 

 foramina (one upon either side) of great size. This is owing to 

 the fact that the posteroexternal angles of the mid xiphoid process 

 are much produced, and very nearly meet the similarly produced 

 posterointernal angles of the lateral ones. The intervening space 

 amounts to but little, and during life is spanned by a dense bridge 



, 



