94 American Fisheries Society 



haust the oxygen supply. Related to this would be an un- 

 usual sensitiveness to carbon dioxide or other products 

 of respiration. Another cause suggested is the purely 

 nervous reaction of fright with its consequences which 

 seems to cause almost instant death with some species of 

 fish. 



In a discussion of the problem of artificial infection 

 (Howard, 1914) the writer suggested a method for car- 

 rying out the experiment. This was to hold the fish in 

 a pound net made of a seine in waters favorable to the de- 

 velopment of the mussel, infecting without removal from 

 the water if necessary. Last year the opportunity 

 came for trying out this plan. On the proposal of 

 Mr. Shira, Director of the Biological Station, Fair- 

 port, Iowa, the writer, in company with Mr. H. L. 

 Canfield, Superintendent of Fish Culture, went to Lake 

 Pepin where a few of the herring were being taken in 

 connection with the propagation of mussels carried on in 

 this lake by the U. S. Bureau of Fisheries. The following 

 plan was carried out: A fish-pound was constructed at a 

 protected place in the lake, as near as practicable to the 

 point where the fish were to be taken in order to reduce 

 the carry to a minimum. The lake is a widening out of 

 the Mississippi River but has the characteristics of a true 

 lake. The length is 25 miles, the maximum width, 2^ 

 miles. No river-current is perceptible and the water 

 reaches a maximum depth of 56 feet. As the fishing oper- 

 ations are carried on at one of the wider places, it was 

 necessary to select a protected site for the pound where 

 the fish and any mussels shed from them would escape 

 wave action. A fortunate location was found convenient 

 to the seining beach and near the head of a bay where 

 the only wave action possible was from directly across 

 the lake, a quarter from which rough water can scarcely 

 develop to an injurious extent. The pound was construct- 

 ed by fastening a 300-foot seine to stakes driven into 

 the bottom. A depth of 4 to 6 feet met the requirements 

 in most respects and consequently the difficulties of con- 

 struction for greater depths were avoided. 



