NEMATOCERA. 47 
brown flat chrysalis, known as a “ flax-seed.” ‘The larvae of Cecids 
do not undergo many moults, and eventually form a cocoon. 
The history of this cocoon formation, if correct, is curious. Win- 
nertz absolutely denies that this cocoon is spun: he says it is 
generally formed by an excretion and not by threads. He observed 
one larva that in twenty-four hours was surrounded by a halo of 
white thread-like particles, resembling the spicules of crystals, the 
larva remaining motionless the whole time. There are often no signs 
of a thread in these cocoons, which are formed in a few days. 
Others have resinous cocoons, and some spin undoubtedly a silken 
thread. The larval skin is not always shed during transformation. 
This change from the larval to pupal state is marked by a change of 
colour in the maggot, the anterior segment at the same time 
becoming distended, and by degrees the rudiments of the wings, 
legs, etc., become developed, and eventually the full pupal state is 
reached. 
The pupz (Fig. 9) of the Cectdomyide resemble in some respects 
those of Scéara, and are usually brownish flattened. bodies of small 
size, resembling seeds. The bases of the antennz are often produced 
into points, resembling horns. Behind these horns bristle-like pro- 
jections may be seen in most pupze—one pair on the head, and 
another on the thorax. Spinose processes may also be developed on 
the abdominal segments, all these processes helping the insect to 
work its way out of the gall or earth at the time of emergence. 
The length of pupation is very variable, warmth greatly hastening 
the development. Under unfavourable conditions the pupz may 
remain unhatched until the next year. Those of the Hessian-fly 
may be retarded by cold, but are not killed by a low temperature ; 
_ dampness is essential to their hatching, a dry cold being injurious 
to the ‘‘ flax-seeds.” 
Mention has been made of the injurious nature of the Cecido- 
myide. Of this there is no doubt; they, nevertheless, are beneficial 
as well as injurious. By their means a great many plants are 
fertilized ; owing to their small size they can enter plants that very 
few insects can enter. 
No better example of the fertilization of a plant can be taken than 
that of Aristolochia clematis, which is largely done by Cecidomyide. 
I have myself seen as many as five Cecids in a single blossom of this 
plant ; one, however, is the general rule. 
Fertilization takes place in the following way: This plant, which is 
a protogynous dichogamous plant, belongs to the family Arzsfolo- 
chiacee, and is widely spread over the globe, several species of the 
