Cn ae) [xcill, xciv 
influence of sex. What explanation remains? I am strongly 
disposed to think that in the vast majority of cases the 
prevalence of dark pigmentation is adaptive, and that although 
melanism in its various degrees may certainly originate as a 
variation or a sport, its increase and establishment are to be 
attributed to some form of selection. Any one of the 
conditions mentioned may by itself, or in combination with 
others, favour melanism ; not however by direct causation, but 
by indirectly leading to the selection of melanic individuals. 
Thus, there is little doubt that the dark pigmentation is in 
many cases of advantage as aiding concealment. This is 
probably the explanation of many instances of dull or dark 
coloration in the female sex not only of insects but of other 
animals, it being well known that the female sex stands in 
special need of protection (Wallace). The writer has observed 
that the darkly-pigmented female of Selenois severina, a 
common African form, is far less conspicuous on the wing than 
the lghter-coloured male. The dark border indeed of the 
former sex is often hardly visible, and the general impression 
given is that of a much dwarfed specimen, though the average: 
female is not really smaller than her mate. An enemy attack- 
ing such forms on the wing might, it is believed, avoid the 
apparently dwarfed specimens, either because they afforded 
less prospect of a sufficient meal, or (in accordance with a 
suggestion of Mr. F. A. Heron) because, apparent size being a 
correlative of distance, the attacker might be deluded into 
supposing the black-bordered forms to be further away and so 
less easily reached than the rest. | 
“It is noticeable that in some cases of heavily-pigmented 
wet-season forms, the under-side shows no corresponding 
melanism, being often in fact far lighter in colour than the 
under-side of the same species in the dry season. This exact 
reversal of effect on the two surfaces is of itself a strong 
argument against the supposed direct operation of meteor- 
ological conditions in producing melanism. Such conditions, 
as I have elsewhere urged, and as has been amply proved by 
experiment, may act as a liberating stimulus; but only rarely, 
if at all, can they be looked upon as a direct cause of darkened 
pigmentation. I would not entirely deny their direct effect 
