J 32 
Bulletin Wisconsin Natural History Society. [Vol. 10, Nos. 3-4. 
swarms and toss up and down in the air over meadows, pastures, 
and stream sides.” Evidently, then, the midges did their flying 
normally at dusk, instead of mid afternoon, except where they 
could secure the advantage of shade or a cloudy morning as we 
observed to occur next morning. It is concluded that they begin 
to form up for the nightly dance above the taller vegetation of 
the region; whether trees, crops, or high grass tufts, seems to 
make no perceptible difference. In some cases the cloud of midges 
near the woods would stretch out in a wavy layer nine to two 
yards thick, about ioo yards in length, and with a width I could 
not estimate for, seeing it in side view. This formation was of 
short duration as the midges would ball up in the thicker clouds 
noted above. 
On going down town after mess, observations of the midges 
were continued till almost midnight. There was bright moon¬ 
light so that the midges could be seen flying over the woods as 
late as that hour. Schuster’s (1904, 344-5) note on the swarms 
of the same species, July 16, 1904, by a pond near the town of 
Lich, northwest Germany, states that the swarm discovered at 
8 45 p. m., varied greatly in shape; and he, likewise, likens it 
to smoke—“a heavy black cloud of smoke,” easily seen at 300 
meters distance. He says “the whole air was filled with thou¬ 
sands upon thousands of midges flying around in every direction, 
so that one had to close his eyes amidst the mass of flying insects. 
The sight of numberless interweaving hordes wavering up and 
down is indescribable.” Then there was a swarm amidst these 
“which had the form of a column and a height of 10-15 meters, 
was stationed above the highest bush of a hedge extending along 
a hillside. A second but much smaller cloud danced about 300 
meters distant from the main swarm over an oak, and a third 
above the hilltop itself, .these three as thick as black clouds. .The 
form of the cloud changed every second. Now the nucleus was 
at the upper now at the lower end; now the column parted in two, 
then the parts united again. The midges danced in clouds rising 
and sinking and all the time with a peculiar jerking flight.” Once 
or twice only did the irregular cabin-sized masses I witnessed rise 
at all like towering smoke, or a column as Schuster calls it. and 
then not long enough to be recorded as a normal formation. Knab 
(1906, 127-9) cites many such columns from early literature. 
