whose larvxe are external feeders. 349 
reaches the 2nd abdominal segment ; a very remarkable 
structure that exists in other families, but nowhere else so 
well developed as here, and which I have called an eye- 
flange. Where, in most pupe, the eye abuts against the 
antenne, it is here rather separate, and a flat tlange-like 
margin, with sharp edge, and in some marked with 
radiating lines, surrounds the eye without quite joining 
the antenna. 
In Lagoa crispata the antennz of the male are avery 
marked feature of the pupa. Unfortunately, I had only 
this one pupa, and so failed to obtain eggs. ‘The cocoon 
is very like the others except in one very important point, 
it has a practicable lid. 
In Zygena the egg is of an ovoid form, with a delicate 
colourless and apparently structureless shell. ‘The con- 
tents divide into two portions—a yellow at one end, and 
a nearly colourless at the other, and till one is familar 
with them one is persuaded they are addled at least. 
The pupa (I have examined filipendulx, lonicere, 
trifolui, eeulans) is very different from that of Limacodes 
in form and colour, but in the most essential points the 
number of free segments, the looseness of attachment of 
the appendages, and in the dehiscense, it is practically 
identical. The maxillary palpus is nearly or quite obso- 
lete, and in some other points the appendage-cases have 
a structure differing from Limacodes. The dorsal head- 
cover is still well-developed. 
I may note that Syntomis, placed in Zygenidex or left 
close by in Syntomidx, not only has, as has been often 
remarked, a very Arctioid larva, but, as is demonstrated 
by the pupa, is really an Arctia, with no affinities what- 
ever with Zygxna. The ovum is also an Arctia ovum, 
not indeed very far from Caja, but differing from Zygena 
toto ccelo, 
I am indebted to Dr. W. G. Clements for the privilege 
of examining a cocoon and pupa of Huchromia fulvida, 
a species also credited with Aygenid affinities. The 
pupa in no way resembles Zygena, but though very 
delicate and transparent, affords no characters that I can 
certainly seize to separate it, much more than generically, 
from Lubricipeda. 
It presented a curious Arctioid peculiarity, viz., the 
adherence of the cocoon to the pupa. Caja cocoon, for 
