MODIFICATIONS OF MOUTH-PARTS 49 
of successively older ages to pupation and the achievement of the definitive 
imaginal condition of these parts, it is certain that the parts marked respect- 
ively imaginal mandible, imaginal maxillx, and imaginal labium, lying respect- 
ively in the larval mandibles, maxille, and labium (with homologies firmly 
based on ontogenic basis) , do develop into those definitive imaginal parts named 
mandibles, maxille, and labium. ... Fig. 2, a horizontal, frontal section 
through the head of a Simulium larva, shows also the forming imaginal maxilla 
and mandibles within corresponding larval parts.” 
Wesché has found that abnormal male Culicids occur with all the trophi 
present, as in the females, although in these cases the maxille and mandibles did 
not extend to the tip of the proboscis. Such males, to conclude from Wesché’s 
work, appear to be not uncommon. Examining a number of males of different 
species, he found an abnormal Anopheles and also a male Culex pipiens with all 
the parts present. 
It would seem that the mandibles and maxille may be absent in the female of 
certain species which do not suck blood. Such appears to be the case with the 
female of Harpagomyia splendens, recently described by De Meijere. This 
curious Javan species feeds upon honey which it obliges the ant, Cremastogaster 
difformis, to disgorge. In this mosquito the hypopharynx is free, as usual, but 
appears to lack the salivary duct. It is interesting to note that early in the 
pupal stage both mandibles and maxille can be readily seen in the developing 
imago. These gradually retrograde, until, by the end of the pupal period, they 
have become wholly atrophied, just as has been shown to occur in the develop- 
ment of the males of the blood-sucking forms. 
While the external appearance of the proboscis is frequently described, we 
know practically nothing of the modifications of structure in the different species 
or genera—Wesché states that in Anopheles the mandibles are serrated at the 
tip. The sheath of the proboscis shows much variation in length and diameter 
in the different species and it is frequently expanded towards the apex. The 
labellz also show much difference in shape and in details of structure but these 
need not be described. In Megarhinus the sheath is unusually rigid, tapering, 
becoming very slender at the tip, and the labelle are narrow and much elongated. 
In this genus, in which both sexes feed wholly upon the honey of flowers, the 
labium can not be bent as it can in the forms which suck blood. 
In the males of many species the sheath of the proboscis shows a suture out- 
wardly from the middle. Usually it is indistinct, but in the male of Deinocerites 
there are chitinous margins connected by a membranous strip as in a true joint. 
Its significance is not clear; perhaps it indicates the limits of the true labial 
structures. It may be mentioned in this connection that Meinert considered 
the sheath of the proboscis as of composite origin and formed mainly by the 
greatly produced portion of that part of the head which he considers the ventral 
plate of the first metamere. 
It has already been shown that there is a salivary duct in connection with the 
hypopharynx. Macloskie found that there were two sets of salivary and poison 
glands in connection with this duct. The glands are situated in the anterior 
ventral part of the thorax and each set consists of three glands, two of which 
