through a "Y" tube. Circulation of the water 

 caused an increase in temperature which was 

 maintained at approximately 13.6°C by means of a 

 refrigeration system (7) and recorded on a ther- 

 mograph (8). Water level in the tank was main- 

 tained by float valves (9). Each trough was 

 supplied with 1 liter per minute of water regulated 

 with flow meters (10). The water used was from 

 the municipal supply, was soft, and was passed 

 through activated charcoal to remove the chlorine. 

 A greater depth of water was needed for absorp- 

 tion of the gases than was afforded by the tank (1), 

 so two towers (11) were added to the system. The 

 spout (12) at the top of the towers was to direct 

 possible overflow water back into the system. 



Inside dimensions of troughs in the fish holding 

 area (13) were 104.5 x 23.5 x 20 cm high. Water 

 depth was maintained at 14 cm. Each trough could 



Figure 1. -Apparatus for subjecting 

 fish to constant-temperature, flowing 

 water with a definite oxygen and ni- 

 trogen content. 



be separated into three compartments with 

 screens— "A" was at the inflow end of the trough, 

 "B" middle, and "C" outflow end. In a few cases a 

 compartment was divided longitudinally so that 

 two groups of fish could be subjected to almost 

 identical conditions. 



Results 



Effect of Variation in Oj /N^ Ratios on Mortality 



Times to death (LE2.-, and LE50 ) of juvenile coho 

 salmon at various concentrations of 0^ and N2 

 during constant total gas saturation of 119^^ ap- 

 pear in Table 1 are shown graphically in Figure 2. 

 All tests were run in duplicate with 50 fish per test, 

 except one at 229% 02and 90% Ng which involve 50 

 fish but one test. With one exception (192% 02and 



916 



