88 SECTIONAL ADDRESSES. 
collects into crowded groups on bare twigs in the evening, as was first 
recorded by Philip Gosse in Jamaica in 1851 and since then by numerous 
observers. H. charithonia, which belongs to the distasteful Heliconines 
referred to on pp. 16,17, and is itself mimicked by other butterflies,?* has 
been carefully studied by Dr. F. M. Jones, and its gregarious habits 
described in detail in his paper ‘The Sleeping Heliconias of Florida.’#® He 
here suggests that the warning characters may be rendered more effective 
at night ‘ by the close proximity of large numbers, under these conditions 
readily recognisable by form, colour, or scent, as identical in kind and 
inedible ; for thus the injury or destruction of one of the group might 
conceivably work for the protection of the many.’ It may be added 
that the choice of leafless twigs for a resting-place obviously enhances the 
conspicuousness of the assemblage. 
We must now return to one of the African Pseudacreas, a wide-ranging 
species (the Linnean ewrytus) which subdivides into a number of local 
forms mimicking the local Acreeine models. This species is represented in 
Uganda by a race (hobleyz) so significant in its bearing on evolution by 
selection that it is necessary to give a little time to it. Hurytus hobleyi 
appears in three forms—two, with male and female alike, mimicking two 
Acreine butterflies (Planema) differing in colour but also with male and 
female alike. The third, with male and female very different, mimics a 
third Planema, the sexes resembling the corresponding sexes of the model. 
Now these four mimetic forms—for the male and female of the last were 
believed to be of different species—have all been described and named as 
distinct, and there was great astonishment and even some incredulity 
when Dr. Karl Jordan, relying on structural features, pronounced them to 
be one. After many efforts to test this conclusion by breeding, a cable 
was received from Dr. Hale Carpenter on Bugalla Island (N.W. Victoria 
Nyanza), giving the information which proved that Dr. Jordan was right. 
Many other families were then bred by Dr. Carpenter, and these, with 
his captured specimens, showed that, in the islands, the three forms run 
into each other, being connected by an abundance of transitional varieties 
which are extremely rare on the adjacent mainland of Uganda. The 
significance of this is obvious when it is realised that the models are for 
some unknown reason comparatively scarce on the islands.‘° 
The same conclusion is enforced by the wonderful families of Papilio 
dardanus, bred by Dr. V. G. L. van Someren and Canon K. St. Aubyn Rogers 
from localities near Nairobi. Now in the families of this butterfly that have 
been bred in other parts of Africa—by Carpenter in Uganda, by Lamborn on 
the W. and E. coasts, by Swynnerton in 8.E. Rhodesia, and by Leigh in 
Natal, the mimetic forms of the females are sharply separated—a fact 
which led to the mistaken conclusion that these patterns appeared fully 
formed and complete, each as a single variation. But in the Nairobi 
families, as in the Pseudacreeas of the Uganda islands, all kinds of transi- 
38 W. J. Kaye in Proc. Ent. Soc. Lond., vol. v., 1930, p. 89. 
89 ‘ Natural History,’ Journ. American Mus. Nat. Hist., vol. xxx., p. 635. A full 
abstract, with references to other observations, in Proc. Ent. Soc. Lond., vol. Vi., 
1931, p. 4. 
40 Trane. Ent. Soc. Lond., 1912, p. 706 ; 1913, p. 606; 1920, p. 84; 1923, p. 469. 
