KFFECTS UPON FISHES OF CHANGES IN SALINITY OF WATER. ( i 



water (density duriug; experiment 1.007 to 1.009). Result, a decrea.se of L'.tl 

 percent diirino- first day: 3.8 per cent during 2 days; 4.7 per cent durini;- 4 

 days; 6.1 per cent during- H days. 

 Experiment- -5^. 



New Yorlv. jNIarch and April, 190.3. Performed as control of above. >ix 

 fishes of same lot left in fresh water and weighed to determine loss due to 

 waste. It was found that there was a decrease of l.H per cent during tii'st 

 day; 2.(! per cent during -1 days; -i:.7 percent during 4- days: <> i)erceTit during 

 6 days. 

 The initial loss is thus seen to be considerably greater in the fishes transferred 

 to even this dilute sea water. But the total loss during the first 6 days, and even 

 during the first -4 days, is practically the same in each case. This is precisely 

 what we should expect if the salt water were hypertonic and the membranes perme- 

 able both to water and (in lesser degree) to salts. There would result, first, a loss 

 of weight, followed by a relative gain in weight as equilibrium was reestablished. 

 This experiment I regard as an extremely important one from the present point of 

 view. Accordingly I repeated it, though the density of the salt water was considei- 

 ably greater at the time of the second experiment. 

 E.cperiment oo. 



New York, May, 1905. (</) Six (Jiu-urliyiirlim put uito sab, water (density 

 1.013). Loss 3.1 per cent duriug first day, -±.5 per cent during 4 days. 



(1>) Six fishes left in fresh water. Loss of 1.9 per cent during first day. .5 per 

 cent during 4 days. 



(r) Six fishes put into salt water (density i.UL'3) in a hirge can. without aera- 

 tion. The fishes, as might have l)een expected, were all dead (prol)al)ly asphyx- 

 iated). The loss in 1 day was nearly 12 per cent. 

 That water of the density 1.013 should prove hypertonic to these fishes is not 

 perhaps surprising. But that water of densit\- 1.007 to 1.009 should have done so 

 (experiment 53) was. I confess, a source of astonishment, since such water has 

 between one-fourth and one-tliird the salinity of ocean water. 

 Erperhnent 56. 



In this experiment the water was of a density 1.007 to L.008, and 4 fishes 

 showed an average decrease of 3.ti per cent in 24 hours. These fishes, it is 

 true, had had a patch of scales removed on one side, but this does not affect 

 the result as determining whether or not the water was hypertonic. All of 

 the fishes in this and in the three preceding experiments survived this treat- 

 ment and continued to live in health as long as they remained under observation. 

 After making due allowance for gain or loss of weight through various acci- 

 dental cii'cumstances, in no wa}' related to osmosis, and for the imperfections of the 

 method emplo3'ed, it seems to me to l)e proved by the foregoing figures that water 

 may enter or leave the body of a fish after immersion in solutions having a different 

 osmotic pressure from the latter. It may be questioned, however, to what extent 

 this process is a normal one, and to what extent it is pathological. In certain cases, 

 without doubt, it is pathological, the death of the fish resulting from, or at least 

 being accompanied by, a considerable gain or loss of water. In earlier experiments 

 it was found that FnuduJns heferocJIfus sooner or later died from the effects of fresh 



