cecidomyiid^. 
581 
one segment more than is usual in dipterous larvae (i.e., 14 instead of 13 
including the head) and have a little horny hook (the ‘ 4 anchor-process ’ 5 ) 
on the underside of the front part of the body, possibly used for changing 
the position of the larva or for breaking up the substance of the gall to 
prepare it for eating. In addition to this, some of these larvae ( Miastor) 
have the extraordinary power of producing young ones in their 
interior, and that not by means of the development of testes and 
ovaries and subsequent fertilisation, but by simple growth, a kind of 
“ vegetative reproduction. 5 ’ These young larvae eat their way out 
of the parent-larva’s body, and either pupate in the ordinary way or 
themselves produce another generation after the same fashion. 
The pupae of Cecidomyiids may be either free or enclosed in cocoons. 
As regards the exact method whereby these cocoons are constructed 
there is much uncer¬ 
tainty ; some are spun 
in the usual way, but 
others appear to come 
into being without 
exertion on the part 
of the larva, which 
seems to remain quies¬ 
cent while the cocoon 
grows round it. It is 
supposed that some 
process of sublimation 
takes place (c/. p. 580). 
Fig. 376—The lower figure shows a hymenopterous para¬ 
sitic grub at the beginning of its attack on a Cecido- 
myiid pupa. The upper figure shows the great growth 
of the grub after two days, the pupa being almost 
entirely consumed. 
The pupae look very much like those of minute Lepidoptera, with 
the legs straight and free from the body at the ends. We figure a species 
found in galls on a wild plant at Pusa (fig. 2, PI. LX), and the same pupa 
attacked by the larva of one of the hymenopterous parasites to whose 
attacks the members of this family seem particularly liable (fig. 376). 
In Europe and America these flies have been a good deal studied, 
and something like a thousand species described, but in India two 
species only have been recorded ; one of them which attacks rice after 
the manner of the Hessian fly, was described by Wood-Mason under the 
name of Cecidomyia oryzce, from Bengal, but since his time nothing seems 
