on the Bionomics of Southern Nigerian Insects. 341 
or two parasitic Chalcids next mail. I'll be so vexed if 
submarines get this lot. Vm really nearly afraid to send 
at all, but the wet season is setting in and things are apt 
to spoil if kept. I will answer all the points you raise in 
detail next mail, which will be in about a week, I think, 
for I cannot get time just now. It was indeed very odd 
that I should find the Teratoneura first and re-discover it 
even to its larva. I may say that the larvae that I had 
feeding were all definitely of the species, no question 
whatever. 
I am especially pleased about the Teraloneura, because 
I really knew what I was after and managed to do it. 
Of course it would never have happened but for Lamborn, 
and it is so very kind of you to take so much trouble 
over it all. 
2. The Larva of Teratoneura. 
[The Lymantrid appearance of the larva was very evident 
in the spirit specimens and is shown in Dr. Eltringham’s 
figure (Pl. XH, fig. 8; and the description, p. 477; see 
also p. 342). It is also the fact that Teratoneura is the 
only butterfly larva yet known which has the urticating 
type of spicules, although their effect has not been 
observed. Mr. W. A. Lamborn happened to be at home 
when the larvae arrived; he recognised them directly, but 
said that he had always mistaken them for moth larvae. 
On Dec. 25, 1917, Farquharson bred the Lymantrid 
moth Naroma signifera Walk. from a larva on the Apocy- 
naceous tree Alstonia congensis, on which Hewitsonia, 
Teratoneura, and other Liptenine larvae were found. The 
specimen was referred to on Jan. 26, 1918, in the following 
passage, in which ‘* Hewitsonia > was probably written for 
“ Teratoneura,’ although the former also appears on the 
notes accompanying the specimen: “ The moth with the 
Hewitsonia-like larva is rather interesting. The larva is 
very like a Hewitsonia larva and lives on ant-trees (I have 
only, however, found two, and of these one was accidentally 
destroyed). It has two glands, in the mid-dorsal line on 
segments 9 and 10, protected by spines. Hardly any 
cocoon is spun—simply a few threads—and the cast skin 
is left hanging near the pupa.” Further specially directed 
observations will be required in order to test the conclusion 
that these larvae always feed on ant-trees and are the 
models of Hewitsonia, Teratoneura, or other Liptenine 
larvae. Naroma signifera is an extremely abundant and 
