NATURAL SELECTION AS APPLIED TO BIRDS 333 



such a character, congenitally developed, treat that character 

 as though it were the whole individual. Any undue develop- 

 ment, or any deficiency of development, of this or that character 

 may throw the whole organism out of harmony with its environ- 

 ment, and so bring about its destruction. Hence we must 

 recognise an intra-organismal selection. 



An admirable illustration of this principle is to be found in 

 the curious development of the lower end of the trachea in the 

 males of certain Anatidse. In the true Ducks — as distinct from 

 the Geese and Swans — the lower end of the trachea, immedi- 

 ately above (cephalad) the right and left bronchi gives rise to 

 a globular bulla, seen in its simplest form in the Genus Anas. 

 In some species it arises from the right, in others from the left 

 side of the trachea, but in all cases it is formed by a fusion of 

 certain of the tracheal rings of which the windpipe is made up. 

 Herein this bulla or "ampulla" has smooth, evenly ossified 

 walls. But in many of the Diving-ducks this ampulla becomes 

 excessively developed, and losing its rounded form takes on a 

 more or less triangular shape, while at the same time the walls 

 become extensively fenestrated so that the osseous portions 

 thereof take the form of a bony skeleton or frame-work sup- 

 porting an exceedingly delicate tympanic membrane, e.g.. Mer- 

 gansers. We have here, apparently, an hypertrophied bulla on 

 the verge of dissolution ; an interpretation the more probable 

 since in some of the Diving-ducks, as in the Scoters {(Edemia), 

 it is wanting altogether, while in the allied Eider-ducks it is 

 extremely reduced. In the Mergansers, by the way, there is a 

 bulla on each side of the windpipe, that of the left side being of 

 enormous size. Figures illustrating this point will be found on 

 111. 44, p. 403. 



If a large series representing different genera and species of 

 Ducks be examined these bullae will be found to exhibit a 

 most perfectly continuous evolution and devolution ; and this 

 irrespective of external characters, and apparently also of the 

 habits of the birds ; at any rate they appear to be only very 

 roughly approximated to the bird's mode of life. As in cases 

 of other organs, and structural peculiarities, they appear, in 

 short, to be not very intimately associated with the problem of 

 the stru^le for existence. 



The skeleton affords almost innumerable instances in sup- 



