AXATIDJi— THE SWANS, GEESE, AND DUCKS. 103 



Family ANATID^.— Thio Swans, (ieese, and Ducks. 



Chabactebs. The same as those of the Order. (See Volume I. page 45.) 



The family Amitidw, which inchides all the North Amei-icaii 

 Aiutrex, constitutes so well-mai'ked and natural a groujj of 

 birds as to need no further detinition than has already been given. 

 The species, being veiy numerous, naturally fall into several more 

 or less well-defined groups, which have been accorded the rank 

 of subfamilies. These, however, grade so insensibly into one 

 another that it is extiemely doubtful whether this rank can 

 be maintained for them.* Birds of this family are found in 

 every known part of the world; but they abound most in the 

 northern hemisphere, particular!}- in boreal regions. The North 

 American representatives may, for convenience of classification, 

 be divided into four tolerably well-defined groups, as follows: 



Cygninae. Neck e.\tremely long (as long as or longer than the body): size very large: bill 

 long as or longer than the head, the edges parallel, the nail small: tarsi shorter than 

 middle too ;lore8 naked: tail-feathers 20-24; color ehielly or entirely white (except in 

 Chenopis alratn. the Bla/.'k Swan, of Australia). 



Anserinae. Neck moderately long (shorter than the body): size variable (usually medi- 

 um, never very large) : bill not longer than the head, tapering to the end, which is chiefly 

 occupied by the large, broad nail: tarsus longer than the middle toe; lores feathered: 

 tail-feathers 14-20; color extremely variable. 



Anatinse. Neck moderately long (shorter than the body): size variable (usually small or 

 medium): bill extremely variable: tarsus shorter than the middle toe: lores usually 

 feathered ; tail-feathers 14-18; color extremely variable. 



HerginEe. Similar to the Anatino', but bill narrow, with tooth-like processes instead of 

 fine lamellir. 



The genera wiiicli liave representatives within our limits may 

 be disting-uished by the following characters: 



'"The whole family A)iatiiiir forms, as to structural features, a very homogeneous 

 group, and intermediate links ore everywhere to be found. Thus it is very diflicult to do- 

 llne the subfamilies anatomically, and to give the siructural differences by which they are 

 to bo soparalnd, so that I .'Ind it not improbable that an exact Investigation, based on a 

 more abundant material than I ean at present procure, will reduce the subfamilies to 

 groups of lower rank." Stejhioeb, in Proc. U. 8. Nat. Miis.. Vol. r,, 1882, pp. 174. 175, 



