88 BULLETIN OF THE BUREALT OF FISHERIES. 



Experiments 9Jf, 95, and .%'. 



These experiments, dealing- with sea raven (2 cases) and sea bass, will he 

 briefly mentioned, though I attribute far less importance to them, since fresh 

 water is soon fatal to both of these species and we are hence not dealini;- witli 

 normal phenomena. In the case of one sea raven ti-ansferred to fresh water, 

 the loss of chlorine was 0.018 per 100 grams during the first hour and o.ol's 

 during first 2 hours. Death probably occurred not long after this. 



Another sea raven was placed in distilled water and death occurred in about 

 2 hours. Up to this time 0.039 gram chlorine per 100 had passed out. Here 

 and in the preceding case almost the entire epidermis was sloughed ofl'. oven 

 before the death of the- fish. 



A sea bass yielded 0.012 gram per 100 during the -to minutes which elapscii 

 before death. 

 The most instructive cases should be those of fishes which in nature live in 

 either medium, and ma_v be transferred with impunit}- from one to the other. The 

 chinook salmon is of course such a fish, and experiments were attempted with this 

 species, specimens being used which had previously lived 4 days in salt water. It 

 was found impossible, however, to keep up sufficient aeration in the limited volumes 

 of water employed, and the fishes died quickly. The white perch {Morone aiiu'iu- 

 c-ana) is a much more favorable subject for such experiments. It has already been 

 noted (p. 60) that perfectly healthy specimens will, in man}' cases at least, survive 

 ti'ansfer to fresh water. For example, two of the same lot as used below were put 

 into running fresh water and remained in good health for four days, when they 

 were removed for other purposes. This alone would not of course be fully conclu- 

 sive. 



Experiment 97. 



Two white perch from salt water (density 1.015), unfed for 2 days, after 

 preliminary rinsing put into separate jars, each containing fresh water 10 

 times the weight of the fish. Mean results for 2 fishes, 0.0.36 gram chlorine 

 per 100 in 1 day, 0.0.58 gram chlorine per 100 in 2 days. Both of these 

 figures are considerably higher than the ones for E. heteroctitiis. 



The fishes at the end of the first day were in perfect health. By the end of 

 the second day aeration had stopped and l)oth fishes were dead. 

 Experiment 98. 



The control experiment of extracting the salts from the viscera was per- 

 formed once more with fishes having the same history as the preceding. Not 

 only were all of the viscera removed and the body cavities rinsed, but the 

 former were boiled for 15 to 20 minutes. It was found that the chlorine 

 derived from all of these sources was (in proportion to the weight of the fishes 

 used) considerably less tlian two-thirds as much as passed from the liody in a 

 single day and only al>out one-third as much as passed from the l)ody in l' 

 days. 

 This does not of course entirely dispose of the objection that tlie salts may leav(^ 

 the body in the excretions of the kidneys. 



It remains to be considered whether chloj-inc in appreciable (juantities is given 

 oft l)V fresh-water fishes while in their normal medium. It will be remembered that 



