1896.] MBTALMO COLOURS OF BIRDS. 287 



gradual to show that the stalk of the metallic barbule is the 

 rudimentary lamina, and the club-shaped body a modification of the 

 distal region of an ordinary barbule. As was observed by Gadow, 

 the cilia are totally suppressed. 



To contrast with this feather, we may take one from the bril- 

 liant green gorget of the Humming-bird Basilinna leucotis (PI. XI. 

 fig. 7). Such a feather is comparatively short, and the especial bril- 

 liancy is confined to a darkly pigmented apical band. The apex of the 

 feather is very abruptly rounded and the barbs are closely connected 

 so as to give a plate-like or scaly appearance. The surface is 

 strongly marked with furrows, which when examined with a lens 

 are seen to correspond to the Ijarbs, each of which lies as it were 

 at the bottom of a trough. The sides of this trough are formed by 

 the two rows of barbules, which are inserted at a distinct angle. 

 The sides of the trough have an unequal slope, as the proximal 

 barbules are inserted at a larger angle than the distal. A further 

 point of interest is that the naked barbs are prolonged beyond the 

 apex of the feather, producing the appearance of a very delicate 

 fringe. This prolongation of the naked barb was noticed many 

 years ago (see the Introduction to Gould's Monograph of the 

 Humming-birds), and its meaning will be explained later on. 



Examined microscopically, the downy barbules of this feather 

 show nothing particularly worthy of notice. The remaining bar- 

 bules have all a well-developed lamina, and a distal region usually 

 well provided with cilia. Those nearest the apex of the barb, that 

 is those which are very brilliantly metalUc, are very darkly pig- 

 mented (PL XI. fig. 9). Those near the base of the upper barbs are 

 much less strongly pigmented, and show very little metallic colour. 

 Between the two extremes there are also otiier marked differences 

 (PI. XI. figs. 8 and 9). Thus in the basal barbules there is no marked 

 angle between the proximal flattened region and the distal cilia- 

 bearing region. In the apical barbules the proximal region is widened 

 and has a well-marked infolding, while the distal region is inclined 

 to it at such an angle as to be almost invisible in surface view 

 (PI. XI. fig. 10). 



Microscopically, the naked tip of the barb is seen to be furnished 

 with rudiments of barbules. 



As this type of metallic feather occurs in all the Humming- 

 birds of which the feathers were examined, it is obvious that Gadow's 

 statement that all metallic barbules are devoid of cilia is too 

 universal. It is certainly true for a great number of cases (Sun- 

 birds, Peacock, &c.), because in these it is the region which ordi- 

 narily bears the cilia — the distal or filamentous region — which is 

 modified into the colour-producing structure, and in the course of 

 the modification the cilia arc lost. In the Humming-birds, on the 

 other band, it is the proximal region of the barbule which is the 

 colour-producing structure, and in consequence the filamentous 

 region is not affected and may bear cilia as usual. In the feathei? 

 described above, the cilia are exceedingly weU-developed in the 

 metallic barbules. 



Though this paper is not concerned with the physical cause of 



