in the pupe of Heterocerous Lepidoptera. 105 
transversely immediately below the eye, and stretching 
from the maxille to the antenne, and comparable in 
width with the antenne or tarsi. I suspected it to 
clothe some mouth part, but the curious thing was that 
it did not appear to unite with the mouth parts, though 
closely approaching them, but seemed to come from 
beneath the antenna. ‘The possibility that it was some 
appendage of the prothorax was thus suggested. I then 
found the same organ in the pupa of Sesiide (Bembeci- 
formis, Tipuliformis, and others), and, looking to the size 
of Bembeciformis, felt sure I should have no difficulty in 
determining what the ‘“‘ eye-collar” really was. In the 
meantime I had submitted the problem to our great 
authority on pupal morphology, Mr. E. B. Poulton, but 
he had not met with the structure, and was unable to 
suggest what it was. When Bembeciformis was ready 
to emerge, and also on emergence, I carefully examined 
this structure, and whilst disappointed in the hope of 
determining the question, found the problem much more 
interesting by the circumstance that this pupal structure 
did not contain any imaginal part, but was empty. I 
succeeded, however, shortly after, in solving the problem, 
by means of the pupa of Nepticula, in which the “ eye- 
collar”? is very well developed. It proved to be the 
case of the maxillary palpus, and the appearance of 
coming not from the mouth, but from under the antenne 
and passing inwards, was in agreement with the actual 
fact. The palpus 5- (or 6-?) jointed, on leaving the 
maxilla, passes backwards in the angle between the head 
and prothorax, until it is situated deeply beneath the 
antenna, then it turns forwards to the antenna, and only 
reaches the surface by emerging from beneath the 
antenna and turning inwards, forming the “ eye-collar,” 
which contains only its terminal joints, the others being 
concealed deeply. 
The greatest interest of this observation perhaps con- 
sists in the light it throws on the relationship of Sestide 
and others, showing that they are descended from 
ancestors with well-developed maxillary palpi, which 
are still retained in the pupal state. 
The ‘‘eye-collar’’ is also well developed in the 
Cochliopodide, which in many respects closely approach 
Nepticule, and in almost the whole Tineal section. 
As to the maxillary palpi, indeed, I do not know any 
