Muttkowski—The Fauna of Lake Mendota. 
399 
estimates have been made, as it forms a negligible item. Vari¬ 
ous species of Hemerobins occur in the sponges of other wa¬ 
ters, but I have not found the larvae. 
F. Order Megaloptera. 
Corydalis cornuta Linne. 
I have on several occasions taken larvae behind the bar of 
University Bay, but never in the lake proper. It is not an 
abundant species. 
Sialis infumata Newman. (Table 5, p. 475.) 
The larvae of this species constitute one of the major items 
of insect fish food. Late in April, or in early May, during a 
very brief period, the larvae migrate to the shore, emerging 
there in immense numbers to pupate in the earth. After two 
weeks the adults emerge; they may be so abundant in places 
along the shore as to be an annoyance to the pedestrian or 
boatman frequenting the spot. In oviposition the females lay 
their eggs in thick, velvety masses on the branches of trees, on 
sticks, and on piers along the water’s edge. 
In emerging the larvae drop into the water, where fish are 
attracted in numbers, apparently relishing the feast. ^ The 
larvae immediately migrate through the plant zone to the sub-f 
littoral area, where they are very abundant. On the bottom 
they will wander downward as far as the oxygen content of 
the water will permit them to go. As the oxygen goes out 
with the warming up of the lake, they migrate to the shallower 
regions; vice versa, they go downward with the fall overturn. 
Perch brought up from the deepest portions of the lake in 
winter were found to contain many of these larvae. 
A two-year cycle is indicated by the fact that shortly after 
the annual shore migration half-grown larvae are quite abund¬ 
ant in the lake. I secured them freely with the dredge through¬ 
out the summer. 
The larvae live chiefly upon organic detritus of various 
origin. They do not refuse live Chironomids and worms, how¬ 
ever. They are water-breathers and may descend to any 
depth, controlled by the oxygen supply. 
The numbers quoted are incomplete, since they do not ac¬ 
count for the distribution below 6 meters, altho the maximum 
is found below the depth indicated. 
