310 PHYSIOLOGY OF RESPIRATION OF FISHES 



When the figures are examined more closely it is seen that the opti- 

 mum pH for the absorption of oxygen at low tension varies with the 

 different species. The herring has the lowest pH optimum which 

 seems to be correlated with the habits of this fish. 



Results of Experiments on the Pacific Herring {Clupea pallasii 



Cuv. and Val.). 



The foregoing experiments were performed at the Marine Biological 

 Laboratory, Woods Hole, during the summer of 1920. It was 

 thought desirable to test the Pacific herring, Clupea pallasii, in 

 the same manner since its behavior was. better known (Shelf ord and 

 Powers, 1915 and Powers, 1921) than the Atlantic herring Clupea 

 harengus. These experiments were performed at the Puget Sound 

 Biological Station, Friday Harbor, during the summer of 1921. The 

 methods employed in the experiments were similar to those at Woods 

 Hole. In order to eliminate individual variation and make the time 

 until death shorter, five 1.5 to 2.5 gm. fish were put in a 2 quart Mason 

 jar of sea water instead of only one. The oxygen was determined 

 immediately after the last fish had died. A total of three hundred 

 and seventy-five fish were tested. Instead of taking the means of 

 the extremes of the oxygen content at the end of an experiment, as 

 was done in the preceding experiments, a curve, which is the 

 mathematical mean of all experiments, was drawn (Fig. 3). The 

 greatest deviation from the mean was 0.18 cc. per liter with the 

 exception of three experiments. In two, one at 6.95 pH where 

 the fish died when the oxygen content was 1.08 cc. per liter and 

 another where the pH was 7.88 and the oxygen content was 0.98 cc. 

 per liter at the end of the experiment, the variation from the 

 mean was plus 0.25 cc. per liter. These two experiments seemed 

 somewhat erratic but they have been included in all averages and 

 calculations. In another, where the pH was 7.43, the variation from 

 the mean was 0.20 cc. per liter. The average variations were plus 

 0.089 and minus 0.085 cc. per liter. This is less than 2 per cent of the 

 total oxygen content of the sea water at the beginning of the experi- 

 ments and less than 12 per cent of the difference of the extremes of the 

 oxygen content of the sea water at the time of death of the fish. The 



