124 Mr W. Macgillivray on the Covering of' Birds. 



the different species of Aptenodytes. The crest feathers of Pa- 

 vo japonicus are almost destitute of these soft barbs at the base ; 

 and this is, in general, the case with all those elongated feathers 

 which, by the French, are termed Plumes de luxe, on whatever 

 part of the body they grow. In quills, there is, in general, 

 scarcely any downy part. In the downy barb, the filament is 

 nearly equal in all its diameters, and is extremely attenuated. 

 The barbules are also elongated, in many of the gallinaceous 

 birds, for example, being twenty times the length of the bar- 

 bules of the apicial part of the feather. These barbules are, in 

 all cases, biserial, like the others, but very frequently they as- 

 sume a direction the reverse of these, coming off from the fila- 

 ment, not in the plane of the web, but at right angles to it, or, 

 in other words, from the face and back of the web, so as to pre- 

 sent on these surfaces a layer of minute silky filaments. This 

 arrangement is especially remarkable in the gallinaceae. Fre- 

 quently the filament becomes spirally twisted ; in which case the 

 barbs seem to have a circular arrangement, although they are 

 still biserial. 



With respect to relative magnitude, the following is an ac- 

 count of the ordinary distribution of feathers in birds. From 

 the head, backwards to the tail, they increase in length and 

 size ; those on the face, or around the base of the beak, being 

 smallest, the tail-coverts longest. The wing-feathers are much 

 shorter than those of the body, and also increase backwards. 

 Those of the upper or dorsal half of the body are almost always 

 shorter than those of the under or abdominal ; and the dispro- 

 portion seems to have reference to the degree of obliquity of the 

 body in its ordinary posture ; for, in those birds which have a 

 nearly vertical position, such as penguins, auks, guillemots, the 

 feathers of the under surface are scarcely longer than those of 

 the upper. The feathers of the upper parts are also more com- 

 pact than those of the lower. 



There is at least as great a difference as to size among fea- 

 thers, as there is among the hairs of [quadrupeds. The margi- 

 nirostral feather of Trochilus moschitus is about one-sixteenth of 

 an inch, while the middle caudal feathers of the Argus are three 

 feet in length. In the same bird, also, the disproportion is of- 

 ten extremely great. For example, the frontal feathers of Pavo 



