33 
submerged stumps, brush, bushes, or any other material of like nature, 
clustering upon or making their way upward and downward with a 
looping gait, or attached by a minute thread-like spider web, they sway 
with the ripples at or near the surface of the water, often half a dozen 
being attached by a single thread. While these larvae make their way 
up and down these submerged objects with perfect freedom, they do 
not venture above the water, and when about to pupate select a situa- 
tion well down toward the bottom of the stream. In deep water they 
were found 8 to 10 feet below the surface, and also much higher up. 
But in shallow water they may be found in. the pupal stage, clustered, 
one above the other, just above the bottom of the stream, their instinct 
having evidently taught them to provide for a sudden fall in the water. 
Notwithstanding this, with the water falling at the rate of 1 foot per 
day, I found many pupce had been left high and dry. 
These pupae are at first of a light brown color, afterwards changing 
to a pinkish cast, and, just previous to the emerging of the adult, to 
black. During the first of these coloral epochs they are attached to 
these vegetable substances by the thoracic filaments, by threads about 
the body and at the anal extremity, somewhat after the manner of some 
Lepidopterous chrysalids; but during the last two the pupae hang by 
the short anal attachment alone, and in this way swing about freely in 
the current, the adult issuing from beneath the water after the manner 
of others of the genus. 
The time and exact place of oviposition as well as the exact length of 
time required for the insect to pass through either the larval or the 
pupal stage I was unable to determine. But when I left Mill Bayou, 
on March 24, the larvae were nearly all of a uniform size and probably 
nearly full grown, a few only being one-fourth to one-half as large. On 
returning, on April 1, nearly all larvae had passed the pupa stage, and 
the adults had emerged; all of those larvae now remaining being as 
large as the majority were on March 24. This, besides indicating that 
the breeding season was nearly ended, also leaves some grounds for the 
inference that several broods may be thrown off, during early spring, 
in rapid succession ; some strength being added to this theory by the 
fact that, as I now learned from those residing near this bayou, the 
cattle had been driven from the woods in the vicinity of the stream 
about the 20th of February. These are points which the necessarily 
limited period during which I had the adolescent stages under consider- 
ation, and the sudden, and to me rather unexpected, termination of the 
breeding season, prevented my settling. 
The adult gnats are usually observed in the vicinity of places where 
they breed, during the first warm days of spring, and they remain from 
ten days to three or four weeks, seeming to prefer a moderately cool tem- 
perature; and hence, during warm weather, are more numerous in the 
early morning and towards evening, frequently being as troublesome 
during bright moonlight nights as during the day time. They are said 
22340— No. 14 3 
