i40 
SEXUAL SELECTION : BIRDS. 
Part II. 
deeper indentation at the divergent than at the con- 
vergent end. It is also manifest that if the convergence 
were strongly pronounced and the confluence complete, 
the indentation at the convergent end would tend to be 
quite obliterated. 
The tail-feathers in both species of peacock are en- 
tirely destitute of ocelli, and this apparently is related 
to their being covered up and concealed by the long 
tail-coverts. In this respect they differ remarkably from 
the tail-feathers of Polyplectron, which in most of the 
species are ornamented with larger ocelli than those on 
the tail-coverts. Hence I was led carefully to examine 
the tail-feathers of the several species of Polyplectron 
in order to discover whether the ocelli in any of tliem 
shewed any tendency to disappear, and, to my great 
satisfaction, I was successful. The central tail-feathers 
of P. Napoleonis have the two ocelli on each side of the 
shaft perfectly developed ; but the inner ocellus becomes 
less and less conspicuous on the more exterior tail- 
feathers, until a mere shadow or rudimentary vestige is 
left on the inner side of the outermost feather. Again, 
in P. malaccensej the ocelli on the tail-coverts are, as we 
have seen, confluent ; and these feathers are of unusual 
length, being two-thirds of the length of the tail-fea- 
thers, so that in both these respects they resemble the 
tail-coverts of the peacock. Now in this species the two 
central tail-feathers alone are ornamented, each with two 
brightly-coloured ocelli, the ocelli having completely 
disappeared from the inner sides of all the other tail- 
feathers. Consequently the tail-coverts and tail-feathers 
of this species of Polyplectron make a near approach 
in structure and ornamentation to the corresponding 
feathers of the peacock. 
As far, then, as the principle of gradation throws 
light on the steps by which the magnificent train of 
the peacock has been acquired, hardly anything more 
