134 
SEXUAL SELECTION : BIRDS. 
p & 
development of a perfect ocellus does not req 
uiw 
long course of variation and selection. 
f# 
With birds and many other animals it sceins, ^ 
the comparison of allied species, to follow, th< 1 ^ ^ 
cular spots are often generated by the breakup, 
and contraction of stripes. In the Tragopan pk^k.vi 
faint white lines in the female represent the bea® ^ 
white spots in the male ; 
and something ^ 
same kind may be observed in the two sexes 0 6 ; 
Argus pheasant. However this may be, appe® 1 ’’^ 
strongly favour the belief that, on the one hand, 11 
spot is often formed by the colouring-matter 
drawn towards a central point from a surro ®® 1 ^ 
zone, which is thus rendered lighter. And, on the * 
hand, that a white spot is often formed by the , ir 
7 ---------- x „ . , • i ( ' ' 
being driven away from a central point, so that n ^ 
mulates in a surrounding darker zone. In eith* 3 ^ 
an ocellus is the result. The colouring matter ^,j, 
to be a nearly constant quantity, but is redfeto ti i 
either centripetally or centrifugaUy. The fea^.jjjt 1 ' 
the common guinea-fowl offer a good instance 1,1 $ 
spots surrounded by darker zones; and where' 1 ® (! i 
white spots are large and stand near each oth e ®’^ 
surrounding dark zones become confluent. In th e 
wing-feather of the Argus pheasant dark sp° ts 
be seen surrounded by a pale zone, and whit '- 1 || lF 
o r e 
by a dark zone. Thus the formation of an “ ^jj 
B y what further steps the more complex ocelli- ' 
in its simplest state appears to be a simple 
wonderful amount of variation in the coloration and shape 
of this butterfly, in his ‘Rhopalocera Africse Australis,’ P- 
J 
ofl<: 
also an interesting paper by the Rev. H. H. Higgins, on 1 
of the ocelli in the Lepidoptera in the ‘Quarterly Journal 
July, 1808, p. 825. 
«• Jerdon, ‘Birds of India,’ vol. iii. p, 517. 
