374 TROPICAL NATURE v; 
white companions? ‘This seems a more probable supposition 
than the altogether hypothetical choice of the female, some- 
times exercised in favour of and sometimes against every new 
variety of colour in her partner. 
A strictly analogous case is that of the glow-worm, whose 
light, as originally suggested by Mr. Belt, is admitted to 
be a warning of its uneatability to insectivorous nocturnal 
animals. The male, having wings, does not require this 
protection. In the tropics the number of nocturnal insect- 
ivorous birds and bats is very much greater, hence winged 
species possess the light, as they would otherwise be eaten by 
mistake for more savoury insects; and it may be that the 
luminous Elateride of the tropics really mimic the true fire- 
flies (Lampyride), which are uneatable. This is the more 
probable, as the Elateride, in the great majority of species, 
have brown or protective colours, and are therefore certainly 
palatable to insectivorous animals. 
Origin of the Ornamental Plumage of Male Birds 
We now come to such wonderful developments of plum- 
age and colour as are exhibited by the peacock and the 
Argus-pheasant; and I may here mention that it was the case 
of the latter bird, as fully discussed by Mr. Darwin, which 
first shook my belief in “sexual,” or more properly “female” 
selection. The long series of gradations by which the beauti- 
fully shaded ocelli on the secondary wing-feathers of this bird 
have been produced, are clearly traced out, the result being a 
set of markings so exquisitely shaded as to represent “balls 
lying loose within sockets ”—purely artificial objects of which 
these birds could have no possible experience. That this 
result should have been attained, through thousands and tens 
of thousands of female birds all preferring those males whose 
markings varied slightly in this one direction, this uniformity 
of choice continuing through thousands and tens of thousands 
of generations, is to me absolutely incredible. And when, 
further, we remember that those which did not so vary 
would also, according to all the evidence, find mates and leave 
offspring, the actual result seems quite impossible of attain- 
ment by such means, 
