Results . --Table 9 gives the conditions and results of each experiment, and figure 10 

 shows graphically the relation between oxygen concentration and percentage of. fish dead, the curve 

 being fitted by the method of least squares, using a log-log transformation of the data. For those 

 experiments employing 14 fish, the times to approximately 50-percent mortality are fairly simi- 

 lar, indicating nnore or less unifornn behavior for the different lots of fish. Comparing the data 

 with the winter lethal values of oxygen obtained from individual fish (table 8b), it is evident that 



the overall range in lethal values is nearly the 

 same, being 0. 55- 1 . 30 and 0. 50- 1. 21 cc. /I. res- 

 pectively. On the other hsind, the mean value 

 from table 8b (0. 96 cc. /I. ) is higher than the 

 mean of the values from the present experiments. 

 Inspection of figure 10 indicates that, in the group 

 experiments, practically all the fish died at oxy- 

 gen concentrations between 0. 55 and 0. 85 cc. /I. 

 whereas figure 11, plotted from the data of table 

 8b, indicates, for the tests on individuals, a 

 more even distribution over the range of lethal 

 values of oxygen. 



Effect of Temperature on Lethal Values 

 of Oxygen for lao 



130 

 OXYGEN CONCENTRATION CC/L 



Procedure . --These tests, in which 

 single fish were used, were performed in a con- 

 stant temperature apparatus consisting of six 

 compartments separated by tin plates with a 

 thernnostatically controlled heating coil in one of 

 the end compartments and a similarly controlled 

 r efrigeration unit in the other. The water in the 

 baths was agitated at intervals by compressed air. 

 After allowing 4-5 hours to establish a constant 



gradient of temperature, the temperature in each of the baths rarely fluctuated more than 0. 3 - 



0.4 C. , with the hotter baths in general showing the nnost variability. 



Figure 10, --Lethal values of oxygen for iao: 

 experiments using several fish. 

 Abcissa represents the oxygen 

 content of the water; ordinate 

 represents the percentage of 

 fish dead at the various oxygen 

 concentrations. 



"- 2 



.7-8 8-9 9-10 10-11 



LETHAL OXYGEN CC/L 



1-12 12-13 



Figure 11. --Frequency distribution of winter 

 lethal values of oxygen for iao, 

 calculated from data of table 8b. 



In the experiments, two 

 Florence flasks, each containing one fish, 

 were immersed in each temperature bath. 

 Handling of fish and water sampling were 

 done in the manner described in the pre- 

 vious sections. Excepting a few experi- 

 ments at 18 C. , all tests were made at 

 tennperatures ranging from 21 to 31.5 C. 

 Preliminary testa indicated that iao die 

 almost innmediately at temperatures over 

 34 C. but, with the exception of a great 

 increase in activity, appear to behave nor- 

 mally at 31°C, For the tests at 18°C. the 

 fish were first gradually introduced to 

 this temperature by placing a gallon jar 

 containing five fish in the water bath used 

 for the temperature versus oxygen con- 

 sumption experiments (fig. 2b) and circu- 

 lating cold water through the copper coils 

 while a slow stream of sea water flowed 

 through the jar. Fined equilibrium at 

 18 -19 C. was reached after 3 to 4 hours, 

 thus avoiding the shock of a sudden 



21 



