aunount of water circulated through the tank. Information on lethal values of oxygen is widely 

 scattered through the literature and there are few extensive experimental investigations of the 

 subject. In view of these considerations, it was thought worthwhile to deternriine the lethal value 

 of oxygen for the local species of baitfish. Both individual fish and groups of 10 to 15 fish were 

 asphyxiated in closed containers. In addition, the effect of temperature on the lethal value of oxy- 

 gen was determined, single fish being employed for these tests. Complete cessation of all move- 

 ments, breathing or otherwise, was taken as indicating the death of the fish, the validity of this 

 criterion being evidenced by the fact that no recoveries were obtained upon return to normal sea 

 water. 



With the exception of the first summer's experiments, all oxygen samples were ana- 

 lyzed by the Rideal-Stewart modification of Winkler's procedure. A comparison between the modi- 

 fied and unmodified methods is reported in table 6. The oxygen samples were siphoned off si- 

 multaneously from large, sealed Erlenmeyer flasks in which fish had been using up the dissolved 

 oxygen for some time. The mean difference in the values obtained by the two methods is not sig- 

 nificant (P = 0.08), but the discrepancy in some of the trials is pronounced; to be Seife, therefore, 

 it was decided to use the Rideal-Stewart procedure. 



Table 6. --A comparison of the Rideal-Stewart and unmodified 

 Winkler procedures for determining the dissolved 

 oxygen content of water in which fish have extensively 

 respired 



