90 SECTIONAL ADDRESSES. 
Carpenter showed to this Section at Birmingham in 1913, that the species 
distasteful to insectivorous animals (although not by any means entirely 
free from this danger) are specially subject to parasitic attack. At the 
same city, in 1886, I brought before this Section the theory of a compen- 
sating principle*® which would check the increase of distasteful insects ; 
for when other food became scarce they too would be devoured, and then 
their conspicuous appearance and slow movements would lead to their 
easy capture. This theory was supported by experiments which proved 
that insectivorous animals, when they are sufficiently hungry, will in fact 
eat the distasteful species, although often with signs of disgust. The 
experimental method, necessarily employed in testing the above- 
mentioned hypothesis, and also of much value when other evidence is 
wanting, was criticised by W. L. McAtee in a paper published in 1912.*8 
I was probably mistaken in not at once writing a detailed reply to these 
criticisms, which were not only directed against the conclusions drawn 
from experimental feeding, but also against other conclusions on which 
the theory of mimicry is founded. On the other hand, there was much to 
be said for waiting until far more evidence had been collected, and now, 
after nearly twenty years, it may be fairly maintained that such evidence 
has been forthcoming. 
In the first place it may be granted that, apart from its special value 
as a test, the experimental method is, in this investigation, very inferior 
to the direct observation of attacks made upon insects by birds and other 
enemies in their natural surroundings and undisturbed. It is impossible 
on this occasion to attempt to give any account of the great number of 
such records which have accumulated since the appearance of McAtee’s 
criticisms. I will, however, mention two sets of observations. In 1927 
Dr. Hale Carpenter kindly sent me the wings of Uganda hawkmoths— 
twenty-one specimens and seven species—found on the floor of a rest-house 
where they had been dropped by bats hanging in the roof. This interesting 
observation suggested an examination of moths’ wings dropped by British 
bats—an ideal means for discovering their true preferences. Wings 
representing 1,328 moths were collected in sheltered places frequented 
by bats—probably always by the Long-eared Bat (Plecotus awritus). All 
the specimens except sixteen belonged to species with protective (Pro- 
cryptic) colours and habits. The exceptions included relatively con- 
spicuous species shown by experiments on other animals to be rather 
distasteful (sometimes accepted, sometimes refused), also species of which 
the palatability is unknown. Not a single specimen with a striking warning 
pattern was present.‘’ ; 
45 Considered in detail in a paper published in the following year: Proc. Zool. Soc. 
Lond., 1887, p. 191. 
46 «The experimental method of testing the efficiency of warning and cryptic 
coloration in protecting animals from their enemies.’ Proc. Acad. Nat. Sci. Phila- 
delphia, 1912, pp. 281-364. 
47 Proc. Zool. Soc. Lond., Pt. 2, 1929, p. 277. The interesting plates I-III in 
Proc. Ent. Soc. Lond., vol. vi., 1931, provide evidence of the same value as that furnished 
by the rejected wings. They show young cuckoos being fed by fosterers with ‘ Small 
Garden White’ butterflies, in Sussex, photographed by Mr. H. F. Chittenden, and 
Cumberland, by Mr. A. G. Britten. Dr. J. G. Myers’ observations on the insect food 
of the Coati (Nasua) were also in large part made under natural conditions (Jbid., 
vol. v., 1930, p. 69). See also Capt. C. R. S. Pitman’s experiments on an African 
Lemur (Perodicticus) on p. 91, and in vol. iv., 1929, p. 90. 
