STRUCTURE OF BIRDS. 73 



with respect to the barb, similar to that of the barbs with re- 

 spect to the shaft. These smaller filaments are named barbules^ 

 Fig. 18, d. It is by means of them that the barbs are firmly- 

 kept in apposition. The manner in which this is done, is not 

 by the barbules of one barb interlocking with those of another, 

 in the manner of dove-tailing, or as the teeth of two combs 

 might be made to alternate by mutual insertion, as has gene- 

 rally been supposed. The position and direction of the barbules 

 do not admit of such union, seeing they meet each other at an 

 angle, and therefore cannot interlock, which could only happen 

 were they to meet vertically. The barbules of the side next 

 the tube are shorter and more adpressed ; those of the side next 

 the tip of the feather are longer and more patulous. The latter 

 are curved downwards at the extremity, w^hile the former are 

 curved upwards, and being placed in apposition, they form two 

 distinct and continuous edges, the incurvate or anterior series 

 of one barb overlapping and hooking into the recurvate poste- 

 rior series of the barb next to it. Although the connection of 

 the barbs may not be easily seen in the ordinary feathers, yet 

 it may in general be discovered in the quills and tail-feathers, 

 without the aid of a glass. When the barbs are pulled asunder 

 in the plane of the web, their cohesion is found to be very con- 

 siderable in most feathers. When the posterior barb is pulled 

 downwards out of the plane of the web, the cohesion is found 

 still greater ; but when the anterior barb is pulled downwards, 

 or the posterior barb upwards, there is found to be no cohesion 

 at all. The curved form of the barbules is distinctly seen by 

 the naked eye in the tail-feathers of Buceros galeatus. 



The barbules themselves frequently present an appearance 

 similar to that of barbs, giving off laterally two series of fila- 

 ments which may be termed barbie els, Fig. 19. These fila- 

 ments are much more sparse than those of the barbs, but their 

 object appears to be the same, namely, that of connecting the 

 barbules, and retaining them in apposition. They are very dis- 

 tinctly seen, with the aid of a small magnifying power in the 

 quills of Aquila Chrysaetus, Diomedea exulans, and Buceros 

 galeatus. 



It may here be remarked, that while what has been assumed, 



