nsectivorous birds abound. But in open country
the dark form would be quite as conspicuous as the yellow form, if not
more so, so that the resemblance to an inedible species would be there
more needed.[103]

The only probable case of mimicry in this country is that of the moth,
Diaphora mendica, whose female only is white, while the larva is of
protective colours, and therefore almost certainly edible. A much more
abundant moth, of about the same size and appearing about the same time,
is Spilosoma menthrasti, also white, but in this case both it and its
larva have been proved to be inedible. The white colour of the female
Diaphora, although it must be very conspicuous at night, may, therefore,
have been acquired in order to resemble the uneatable Spilosoma, and
thus gain some protection.[104]

_Mimicry among Protected (Uneatable) Genera._

Before giving some account of the numerous other cases of warning
colours and of mimicry that occur in the animal kingdom, it will be well
to notice a curious phenomenon which long puzzled entomologists, but
which has at length received a satisfactory explanation.

We have hitherto considered, that mimicry could only occur when a
comparatively scarce and much persecuted species obtained protection by
its close external resemblance to a much more abundant uneatable species
inhabiting its own district; and this rule undoubtedly prevails among
the great majority of mimicking species all over the world. But Mr.
Bates also found a number of pairs of species of different genera of
Heliconidae, which resembled each other quite as closely as did the
other mimicking species he has described; and since all these insects
appear to be equally protected by their inedibility, and to be equally
free from persecution, it was not easy to see why this curious
resemblance existed, or how it had been brought about. That it is not
due to close affinity is shown by the fact that the resemblance occurs
most frequently between the two distinct sub-families into which (as Mr.
Bates first pointed out) the Heliconidae are naturally divided on
account of very important structural differences. One of these
sub-families (the true Heliconinae) consists of two genera only,
Heliconius and Eueides, the other (the Danaoid Heliconinae) of no less
than sixteen genera; and, in the instances of mimicry we are now
discussing, one of the pairs or triplets that resemble each other is
usually a species of the large and