288 MISS NEWBiGiN ON THE [Feb. 18, 



colour, it may be noticed as a fact for future investigators that, as 

 is indicated in the iigures, the laminae in the feather described show 

 raore or less distinctly an arrangement of cross-bars (or compart- 

 ments). This is common in the case of the Humming-birds, but it 

 is also sometimes distinctly visible in the laminss of the barbules of 

 Bun-birds (see PI. XI. fig. 3 and PI. XII. fig. 18), so that its signi- 

 ficance is somewhat doubtful. One other fact in connection with 

 the metallic colour may be noted. Contrary to the usual rule, the 

 featherof the Humming-bird mentioned under certain circumstances 

 exhibits metallic tints by transmitted light even under 5" objective. 

 When this occurs, it is clearly seen that the colour is confined to the 

 dark infolded part of the barbule, which alone under ordinary circum- 

 stances is visible (see PI. XI. fig. 10). This infolded part cannot, 

 however, be itself the cause of the metallic colour, for it is absent 

 from the metallic feathers in some cases (see PI. XII. fig. 19). 



The ridging of the surface of the feather produced by the way 

 in which the barbules are inserted, though apparently a factor in 

 colour production, cannot be absolutely essential, for it is some- 

 times absent. Thus, in Calothorax lucifer in the brilliant throat- 

 patch the feathers in the middle line are very distinctly ridged and 

 have the usual plate-like structure, while those at the sides of the 

 throat are quite devoid of ridges but do not show diminished 

 brightness. Similarly, in CyanolesUa gonjo the brilliant green 

 feathers of the crest are perfectly smooth, and have a certain hard 

 gloss which is absent in the ridged feathers. 



The fact that in Humming-birds it is the proximal, and in Sun- 

 birds the distal region of the barbules which is metallic has more 

 effect upon the general coloration than might at first sight be 

 supposed. It is, in the first place, now obvious that there is no- 

 thing in the structure of the metallic feathers of Humming-birds 

 which is likely to affect their efficiency even in quill-feathers. We 

 can thus understand how it is that many of the tail-quills in 

 Humming-birds, and even the wing-quills in Eulampis jxujularis, 

 may exliibit metallic colour and yet perform their usual function. 

 But this is not the only effect produced by the difference. In Sun- 

 birds the lamina is rudimentary in the metallic barbules ; now the 

 lamina tends to become rudimentary in the barbules at the bases 

 of the barbs, that is in downy barbules, and in the barbules at 

 the apices of the barbs. Downy barbules uever seem to show a 

 tendency to become metallic, and we thus find that in Sun-birds the 

 metallic barbules occur only at the ends of the barbs. A little re- 

 flection will show that the consequence of this is, that the metallic 

 barbules will form_ a transverse band on contour-feathers, where 

 the ends of the barbs stand much on the same level, and a longi- 

 tudinal band on quill-feathers where the racliis is much elongated. 

 The position of the bands of metallic colour on the feathers of Sun-» 

 birds is thus a consequence of the kind of modification to which their 

 metallic barbules are subject. (See PI. XI. fig. 1 and PI. XII. fig. 17.) 



In Humming-birds there is an almost complete reversal of this, 

 state of affairs. In them the metallic barbules have an extremely 



