290 MISS NEWBIGIN ON THH [Feb. 18, 
tion of the ordinary metallic contour-feathers of most Humming- 
birds has been accompanied by a reduction of this terminal band 
in width and in the individual barbules forming it, a large increase 
of pigment in the lamine of the barbules forming the blackish- 
grey part of the feather in Phaethornis, and a specialization of the 
barbules of this region, of which the most obvious result is the 
shortening of the filamentous portion and its inclination to the 
lamina. In the feathers of the patches of especial brilliancy these 
changes have been carried further, and have been accompanied by 
a shortening of the feather and rounding of its tip, and a change 
in the angle of insertion of the barbules. The changes in the 
wing-quills seem to have been of a simpler description, and to have 
been chiefly accompanied by a specialization of the lamina. 
As to the meaning to the individual of this progressive change, 
one suggestion may be hazarded. The Humming-birds are espe- 
cially characterized by their power of flight. Now of the many 
correlated variations which must occur during the gradual improve- 
ment of the power of flight, an increased development of the 
lamina and of cilia and hooklets is likely to be important. It is 
therefore perhaps not unreasonable to suppose that the metallic 
colours of Humming-birds are due to a persistence in the same line 
of variation which produced their powers of flight. If Kolliker 
(“ Die Entstehung des Pigments,” Zeitsch. f. wiss. Zool. vol. xlv.) is 
right in his belief that the formation of pigment is connected with 
the blood-system, it is quite comprehensible that an increase in 
structural specialization should be accompanied by an increase in 
the amount of pigment. It might be objected that the Swifts, 
which are probably nearly allied to the Humming-birds, have also 
great powers of flight and yet do not show metallic colours. In 
reply to this objection, it may be said that it is generally admitted 
that Humming-birds have few enemies, and that therefore variations 
might occur in them unchecked which would possibly lead to 
elimination in other forms. A more important objection is that 
the wings do not usually show metallic colour : it seems impossible 
to suggest a reason for this beyond the simple fact that the wings 
seem to be slow to vary incolour. It will be noticed that, in the 
especially brilliant patches, the extreme closeness of the connection 
between the adjacent barbs is a variation in the direction of the 
ordinary condition of the feathers of flight. 
The course of the development of metallic tints in Sun-birds offers 
many points of contrast to that just described for Humming-birds. 
Tf we take (PI. XII. fig. 22) one of the ordinary contour-feathers 
of a non-metallie form, such as a female of a species of Cinnyris, 
probably C.jugularis, it will be found that it presents considerable 
resemblance to a corresponding feather from a “‘ hermit” Humming- 
bird. Thus it consists of a basal downy region, a mid-region 
pigmented with brownish black, and an apical region with discon- 
nected diverging barbs, usually of a dull olive colour. Beginning 
with such a feather, the development of metallic colour is 
accompanied by an increasing predominance and pigmentation of 
