408 Mr. 0. H. Latter on the secretion of 
very intolerant of the unnatural conditions attending 
captivity. 
Mr. Edmonds of Windsor informs me that they can 
only be safely removed from their own cocoons just 
before pupating, otherwise the larvae endeavour to 
construct a fresh cocoon, and so exhaust themselves. 
He also tells me that in casting the last larval skin it 
presses very hard on the sides of the cocoon and 
accomplishes the moult with remarkable rapidity. If, 
as is not improbable, strong pressure is applied to the 
sides of the cocoon when the pupal skin is about to be 
cast oflP, it can be readily seen that advantage would 
result in the pupa being so armed with a pointed boss 
that the exertion of bursting the pupa case should at 
the same time drive the boss into the line of junction 
between the lid and body of the cocoon. If this is so, 
then, probably, the very slight alkaline effusion is of no 
avail in softening the walls. 
Halias prasinana. 
I am indebted to the Rev. Gr. Chilton for pupae by 
means of which I ascertained that this species dis- 
charges an alkaline liquid when escaping from the 
cocoon. The amount produced is not great, but very 
decided and more copious than that discharged by 
B. lanestris or L. testudo. I could discover no definite 
organ adapted for opening the cocoon, unless such be 
represented by a pair of short sharp projections from 
the anterior ventral margin of the rim of the eyes 
(Fig. 14, s). It should be mentioned, however, that no 
part of the cocoon is removed by the emerging imago. 
The anterior portion of the cocoon is formed by the 
drawing together of the two sides and the inter- 
weaving of their component fibres, so that the imago 
has merely to burst this suture and thrust the softened 
walls apart right and left, the walls returning into 
position immediately after. The mouth appendages are 
undoubtedly inadequate for the work — the antlise are 
very long and flexible, lacking the necessary rigidity ; 
the palps are soft and terminate in remarkable sucker- 
like discs (Fig. 13, p), they are curved upwards so as 
to present their ventral surfaces to the front, the dorsal 
surfaces being correspondingly concave; the former 
