92 BULLETIN OF THE BUREAU OF FISHERIES. 



In 1 (lav the fishes lost over 16 per cent of their chlorides; in 11 ilays nearly '■',' 

 per cent. As I have pi'eviously pointed out in discussing the analysrs of clilniinc in 

 the water, the loss of salts from tlie Ixxly takes phi.-c at a diininisiiing rate. In the 

 present case nearly as niurh passed out dniMni;- the first day as during the ncxi in 

 days. 



Erp.rnu.ut inc. 



Fishes kept ill salt water. 



(«) 1 (lay (4 fishes 0. 181 \ ^^^^.^^^ ^ ^^^ 



(((') 1 day (4 fishes 0. 196* ""'• " ^ 



ih) .T clays (2 fishes) ■. (i. U:; 



(/■) 10 (lays (3 fishes) , e. i:.! 



For the rather anomalous fact that in two diiierent tests the 1-day tishes oa\ ( 

 niucli higher percentages of chlorine than the 5 or the 10 day ones, I will oti'ei' tin 

 following explanation: (1) As shown above (experiments G.5-(W). the immediate result 

 of the immersion of these fishes in salt water is a loss of water amounting in 1 day 

 to about lit per cent of their weight. This would of course result in a higher con- 

 i-entration of the soluticnis in the body. It was found, however, that part at least of 

 this loss of water is made up later. (2) A certain (often a large) proportion of these 

 tishes succumb within a few days after transfer to .salt water. It is quite jjossible. 

 therefore, that the lots analj-zed included individuals which would soon have died 

 and thus had undergone greater changes than normal tishes would have done. 

 Those, on the contrary, which had survived :> or more days in .salt water were proba- 

 lily individuals that had not been harmed by the change and would probably have 

 lived indefinitely." 



Omitting the figures for the fishes which had passed 1 da\- in salt water, it will 

 be seen that those for the other tests fall into a very suggestive series. 



I 11 days 0.085 



Fresli water ] ?, days 0. 108 



I I day 0. 1 1 2 



Brackish water (1.002) 0. 1.34 [or 0. 142] 



Sait water! 5<l7« - «• '-^^ 



UO days 0.151 



It will be seen that the last of these figures is aliout 7s per cent greater than tlie 

 first. It will be seen likewise that whichever figure be regarded as the more cor- 

 rect one for the brackish-water fishes, the latter agree much more closelv with the 

 salt-water than with the f re.sh-water individuals (the comi)arison being of course with 

 the extreme members of the series). 

 Experiment 107. 



Eight fishes kept in water of density 1.018 for 7 days. The percentage of 



chlorine for this lot was 0.134, i. e., the same as for tlios(> kept in lirackish 



water. 



It was thought desirable to test .some species which in nature inhabits fresh. 



bi'ackish, and salt water equally well, using individuals from each of the.sc sources. The 



white perch (2£orone americana) seemed likely to be. a favorable object for such 



<t .\nother explanation of this apparent anomaly was suggested to me by Prof. W. C. Sabine. It is quite possible that 

 after the first shock of change, resulting in considerable osmotic disturbances in the body of the fish, an effort would be 

 made by the latter to reduce its salt content to the original (normal) level, the excess or. part of it being eliminated by 

 way of the organs 



