American Fisheries Society 285 



in spite of its fineness and the care with which it was fitted 

 and caulked. Even in the second trough with its small 

 flow they constantly drifted toward the foot-screen, so that 

 it seemed best to siphon them and return them to the head 

 of the trough occasionally. The fish seemed to hatch faster 

 after the electric light was turned on in the evening, but 

 this appearance may have been due in part to the greater 

 distinctness with which they could be seen and the presence 

 of their sharp shadows. About midnight of the third day, 

 April 25th-26th, a second shipment of 400,000 was sent out, 

 and 24 hours later an additional 200,000. 



The remaining eggs were then cleaned and put into three 

 jars, thus reducing the flow into the retaining trough. 

 During the fifth day the eggs were again handled and the 

 number of jars reduced to one. On the sixth day it was 

 found necessary to clean the screen but twice and that night 

 no watchman was needed. The number of fish hatching- 

 had now become so small that a siphoning at 9 P.M. and 

 another at 3 next morning sufficed for the first trough. 

 The other, which contained more fish, was protected in 

 another way. By experiment it was found that the fish were 

 strongly attracted by light and, by making the lower end 

 of the trough very dark by means of a close cover and 

 leaving the upper part open and an electric light over it, the 

 fish were induced to remain in the upper part of the trough 

 and away from the screen. 



The seventh day, April 29th, careful estimates were made 

 of the dead eggs and fry, and a final shipment calculated at 

 400,000 was sent out. A few strays found in the trough 

 were held for observation. These grew rapidly for a few 

 days and, their exact age not being known, attempts to feed 

 them were begun May 2d, moss supposed to contain micro- 

 scopic food, water from among such moss, fine cut liver 

 strained through cheesecloth, and finely divided pulp from 

 ground crawfish being offered. 



So far as observed they did not feed at all until May 10th, 

 when they began to eat one another. Two days later 



