101 
segments, as against 16 or 17 in the seed-midge. The male, distin- 
guishable from the female by the large clasping organs at the end of 
the abdomen, is also smaller than the male of the seed-midge. 
Life History. — The podlike galls ( PI. II., Fig. 4) of the leaf-midge 
are common on white clover thruout the growing season, during which 
the insect in one or another of its stages can always be found in some 
of the galls. Usually several stages in the development of the insect 
are represented at the same time in a single gall. Frequently one to 
six individuals inhabit the same gall, and there may be more; I know 
of twenty eggs being deposited by a single female on one leaflet. 
Larvse from eggs laid at the same time do not all develop equally ; 
some of them outdistance the rest, and when the gall is crowded with 
the maggots, some of them die oft' in the competition for nourish- 
ment. 
The earliest date that I have recorded for the galls is June 21, at 
which time larvae and cocoons were present ; thereafter I found the 
species in various stages every few days up to October 5, on which 
date larvae were common. The latest eggs and larvte are killed by the 
frost, along with the leaves which they inhabit. This fact in connec- 
tion with what we know about the habits of the species warrants the 
inference that it probably winters in the cocoon, tho possibly as a fly. 
My observations indicate four full broods a year in this place, 
with scattering and ill-fated members of a fifth brood. Each genera- 
tion requires about one month, on an average, for its development. In 
the field the broods overlap enough to make their separation by field 
observations a little difficult. Yet there are times when almost all the 
galls are empty, indicating an interval between two broods, when the 
species is represented almost entirely by winged adults. Thus, Au- 
gust 14, 1903, when I examined a great number of recently made galls, 
1 found 85 without living contents, tho remnants of cocoons were 
often present; 9 with cocoons (one to four in each instance) ; 2 with 
full-grown or large larvae; and 2 with larvae that had just hatched. 
Again, August 15, 1907, I opened several hundred of the galls and 
found nearly all of them to contain cocoons or full-grown larvae ; only 
one or two had white larvae, and very many of the galls were empty, 
tho not more than a month old. Thus a break between two genera- 
tions occurred about August 15 in two years. 
Eggs laid August 1 gave six larvae, the flies from which issued 
August 22 ; this is the shortest life history that I have found for this 
species. The egg period is six days as a rule; thus, eggs laid August 
13 (2:45 to 3:15 p. m.) hatched August 19 (at noon). The period 
in the cocoon varies. From two cocoons made August 3, the imagines 
emerged August 13 ; in several other instances, nine days was the 
period. On the other hand, a larva that was making a cocoon August 
5 did not give the adult until August 27. 
Habits. — The midges can be obtained simply by picking the galls 
and keeping them on damp sand or cotton- wool. A better way, how- 
ever, is to transplant such plants as bear the galls, because the leaves 
do not keep fresh very long after being picked. 
