Natures Night Lights, 171 
a protection against one set of enemies for a portion 
only of the period during which they are active, is 
altogether incredible. 
The current theory, which we owe to Belt, is a 
prettier one. Certain insects (also certain Batra- 
chians, reptiles, &c.) are unpalatable to the rapa- 
cious kinds ; it is therefore a direct advantage 
to these unpalatable species to be distinguishable 
from all the persecuted, and the more conspicuous 
and well-known they are, the less likely are they to 
be mistaken by birds, insectivorous mammals, &c., 
for eatable kinds and caught or injured. Hence we 
find that many such species have acquired for their 
protection very brilliant or strongly-contrasted 
colours — w^arning colours — which insect-eaters come 
to know. 
The firefly, a soft-bodied, slow-flying insect, is 
easily caught and injured, but it is not fit for food, 
and, therefore, says the theory, lest it should be 
injured or killed by mistake, it has a fiery spark to 
warn enemies— birds, bats, and rapacious insects — 
that it is uneatable. 
The theory of warning colours is an excellent 
one, but it has been pushed too far. We have 
seen that one of the most common fireflies is 
diurnal in habits, or, at any rate, that it performs 
all the important business of its life by day, when 
it has neither bright colour nor light to warn its 
bird enemies ; and out of every hundred species of 
insect-eating birds at least ninety-nine are diurnal. 
Raptorial insects, as I have said, feed freely on fire- 
flies, so that the supposed warning is not for them, 
and it would be hard to believe that the magnificent 
