324 U. S. BUREAU OF FISHEBIES 



On the other liand, from a lot of trout n few of the most vifjoroiis 

 will outlive the m!ljorit3^ 



Little characteristic change in the activities and appearance of 

 the fish will be noted during the progress of chlorine poisoning. 

 "Where the quantity of chlorine is insufficient to kill quickly a certain 

 sluggishness is evident among fingerlings and older specimens. In 

 the later stages of the poisoning there is a tendency to collect at 

 the surface of the water. No other superficial reactions are evident, 

 and in the sac fry no special symptoms can be noted. 



As the deaths from heavier concentrations occur generally at night, 

 no records from the aquarium under these conditions have been made. 

 In experimental work rainbow-trout fingerlings subjected to heavier 

 concentrations (approximately 0.3 part per million) show a moderate 

 restlessness, with slight dj^spnoea (rapid breathing) and loss of 

 equilibrium, foUowecl by death in two hours or less. A greater 

 quantity of chlorine tends to increase the rate of respiration rapidly 

 and to accentuate the other symptoms. It will be seen that these 

 responses might follow the action of a number of chemical agents 

 other than chlorine. An important consideration lies in the fact 

 that fish whose " gassing " has jDrogressed to a serious stage generally 

 do not recover when transferred to water from which the toxic 

 agent has been eliminated. 



In regard to the concentrations of chlorine involved in the process 

 of purification of civic water supplies, more is said in another section 

 in the discussion of the chemistry of the subject. It should be stated 

 here that the earlier conjecture (" chlorine in this proportion — 0.3 

 part per million — might not have been injurious to fish ") does not 

 represent the facts. Experiments can be cited to show that 0.3 part 

 per million chlorine will show its effect on yearling rainbow trout 

 within 15 minutes and kill within 2 hours. A smaller amount (0.25 

 part per million) is fatal to fingerlings of the same species in 4 to 5 

 hours, but a solution of equal strength leaves goldfish apparently 

 unaffected at the end of 42 hours. It is apparent that the foregoing 

 concentrations are not maintained regularly in the water as received 

 from the mains, but too much emphasis can not be placed on the fact 

 that periodical " surges " or " pulses " at times may elevate the chlo- 

 rine content above these levels. 



Compilations of the toxicities of fish life to various chemicals make 

 no mention of chlorine. The only available refeience to this phase 

 is that to the work of Rushton,^ who experimented with chloros, a 

 commercial sterilizing agent supposed to exert its effects by the 

 action of free chlorine. He cited a 1 to 400,000 parts free chlorine 

 solution as being virtually harmless to trout ; but his chlorine deter- 

 minations were made by the silver-nitrate method, which shows only 

 chloride-ion and not CL. Hence, no admissible estimate of the 

 gaseous content of the water is given. Utilizing pure chlorine water, 

 he made further experiments on trout, but here no statement of actual 

 analyzed chlorine content is made ; computation from his data indi- 

 cates that his specimen was uninjured at a possible maximum of 0.4 

 part per million. How far this was the actual content can only be 

 guessed. 



5 Rushton, W., Salmon and Trout Magazine, October, 1021, p. 296, et seq., and .January, 

 1922, p. 40. London. 



