102 
To get the eggs and the subsequent stages it is only necessary to 
turn a lot of the midges loose on potted white clover under a lamp 
chimney (the narrow cylindrical kind) covered at the top with net- 
ting, making sure that the plant bears some young unopened leaves. 
It is best, also, to cut off most of the old leaves, especially in order to 
facilitate observation. 
These little flies mate readily under such conditions, and lay eggs 
freely, as I have several times observed. After coition, which may 
not last longer than a minute, oviposition occurs within an hour or so. 
The female either drops to the ground or alights on the plant and 
walks downward. In either event she finds, after more or less explora- 
tion, a young leaflet still folded in halves. Standing at the base of 
this, she wriggles her long flexible ovipositor in between the two con- 
tiguous faces of the leaflet as far as possible ; at intervals a slight wave 
of distention passes back along the ovipositor, indicating probably the 
passage of an egg. Usually several eggs are laid" on the same leaflet— 
sometimes a dozen or more. After many eggs are laid, the abdomen of 
the female is noticeably smaller. 
■ In some way the insect prevents the leaflet from opening out. The 
larva is quite unable to fold a leaflet that has already spread out. If 
placed on such a leaf the larvae cannot even hold on to it, and rolls off. 
Like other midge larvae, they require some tight crevice in which to 
develop. 
Like them also, the maggots of the clover leaf-midge are very 
sensitive to moisture, contracting and becoming motionless when it is 
dry, and resuming activity when it is moist. Dryness prolongs the 
period of development and retards the emergence of the fly. The 
long record just given of twenty-two days from cocoon-making to 
emergence was due in some measure to dryness. The larva found 
making a cocoon August 5, was put into a small glass-covered box and 
pressed up against the glass by means of dry cotton-wool, so that I 
could watch the process of cocoon formation under the microscope. 
The larva spun for awhile, and then left its cocoon unfinished and 
wandered about, but it was contracted and motionless August 6 to IL 
On the latter date the cotton was moistened, and soon the larva began to 
wriggle about in all directions, away from the light, and kept this up all 
day. Coming to rest again, the larva contracted without making a 
new cocoon, and gave an imago August 27. The larva often pupates 
without making a cocoon, if taken out of its gall. 
The cocoon is evidently composed of silken threads, as Comstock 
said. Under the microscope the thread can be seen to issue from the 
mouth of the larva, which swings the fore part of the body to and fro 
as it spins. This fact is mentioned to counteract the old statement 
that the cocoon is an exudation around the body of the larva. 
The leaf-midge is not without natural enemies. I have found the 
larvae and pupae of a chalcid in the galls among the midge larvae. The 
chalcid is not common ; I found it in only 1 percent of the galls, out 
of hundreds examined. The chalcid larva might hastily be mistaken 
I 
