SKELETOlSr OF THE MUSK-DUCK. 399 



Thus we have a parallel case to the Grrebes and Divers. The 

 former are freshwater divers, and have much smaller supra- 

 orbital glands than in the latter, which are marine. Both Grrebes 

 and Mergansers, however, are to be met with at sea, the latter 

 commonly, but it would seem that this habitat is of recent date. 



In Tachyeres and Somateria the fore part of the head lying 

 between the orbit and the base of the beak has become enor- 

 mously elongated, drawing out the nasal limb of the lachrymal to 

 an extent elsewhere unknown among the Anseres. In Siziura 

 this region is abnormally short, wherein it agrees with what 

 obtains among the Mergansers. The lachrymal of Bizmra again 

 also closely resembles that of the Mergansers. 



On the whole the skull of Biziura resembles that of JErisma- 

 tura more closely than that of any other duck ; but differs from 

 this genus in that Erismatura possesses an ossified antorbital 

 plate, and a large triangular lachrymal, an unfenestrated 

 interorbital septum, a long postorbital process, and a long 

 spatulate beak. Unfortunately I have no skeleton of any other 

 member of this group of ducks wherewdth to carry these com- 

 parisons further. 



The Veetebeal CoLUMiir. 



Only the thoracic, synsacral, and caudal vertebrae seem to 

 require notice here. 



In Biziura all the thoracic vertebrae bear long hypapophyses. 

 Erom the 2nd to the 5th the free ends of the spines expand to 

 form horizontal J.-shaped plates. 



These hypapophyses have undoubtedly been developed to 

 increase the power of diving; but whether they have been acquired 

 as the direct result of a long-sustained stimulus through a long 

 series of generations — that is to say, by the inheritance of 

 acquired characters — or whether by the selection of favourable 

 variations in this direction, I cannot pretend to say. 



It is significant to note that these processes are wanting in 

 the Geese. In non-diving ducks, e.g.^^?«5, only three are present, 

 and these lack the J.-shaped process. The Sheldrakes agree 

 with the diving Scaups, having four of these vertebrae with 

 hypapophyses. In the Scoters and Eider-Ducks, which are sea- 

 ducks and expert divers, there are 7 of these processes as in 

 Biziura. They are longer and have larger j_-shaped plates than 

 in Anas, Fuligula, or Tadorna. Erismafura, as we might expect. 



