SKELETON OP THE MUSK-DTJCK. 401 



ends o£ the lateral processes meet the angles of tlie truncated 

 metasternum to form fenestrse *. In the Sheldrakes and the 

 Eed-frouted Goose, for example, the union of these elements is 

 nearly complete. 



In Clangula the metasternum is of great width, converting the 

 sternal notches into fenestrge by joining with the posterior lateral 

 processes ; further, the metasterrvum is peculiar in that its 

 cartilaginous free edge has become partly ossified. In Harelda 

 this cartilaginous element is greatly increased, forming a large 

 linguiform plate, and, at the same time, by increasing the width 

 of the metasternum, it has, in very old examples, converted the 

 sternal notches into fenestrse as in Clangula. 



In Biziura the sternum is relatively longer than in other 

 Ducks, and it also ten^srto taper anteriorly. The anterior latel^l 

 processes are obliquely truncated as in Erismatura. In JErisma- 

 tura there is a broad, bifid spina externa, but in Nomonyx this 

 process is remarkably well-developed, therein agreeing with the 

 Greese. In Tachyeres it is moderately large. In all other duclvs 

 this process is either vestigial or wanting. 



The keel of the sternum in Biziura shows evident signs of 

 degeneration, being extremely low, and terminating some distance 

 before reaching the end of the sternal plate. The accompanying 

 illustration will show the relative proportions of the keel between 

 Biziura and Erismatura. In the marked length of the sternal 

 plate we have an evident adaptation to diving habits. 



It is curious to note that the degeneration of the keel in 

 Biziura is more marked than in Tachyeres, which is actually 

 flightless. 



The Mergansers only among the Ducks have the ventral border 

 of the keel produced far forwards, in a fashion recalling that of 

 the Cormorants and Grannets. In Mergus serrator this feature 

 is especially well-marked. The precise significance of this is not 

 obvious. 



The Eibs. 



The hindmost ribs of Biziura recall those of the Colymbi and 

 Alcidse, in that they are much elongated, running almost parallel 

 with the pubis. Only the first five of the thoracic ribs appear 



* Erismatura viitaia appears to form one of those puzzling exceptions which 

 seem to be inseparable from every rule, inasmuch as the notches are in this bird 

 not only converted into fenestroe, but the closing process has got>e so far as to 

 come near the obliteration of the fenestra of the left side altogether. 



