EFFKCTS UPON FISHES OF CHANGES TN SALINITY OF WATER. 89 



ill experiments 88 and 89 some F. hetei^ocUtm which had been living in fresh water 

 foi- some daj's were found to pass out about half as much chlorine in a day as did the 

 fishes recently taken from salt water. In qualification of these figures it will be 

 recalled (1) that small amounts of chlorine were apt to be overestimated, and (2) that 

 in each of these two experiments one or more of the fishes had died. 



Experiment 99. 

 A cattish {Ameluruft nehuloifus) was found to pass out during li» hours 



O.OOJ: gram chlorine per 100." This is thus between one-eighth and one 



seventh the first day's output of F. hi-tn-dclifux from salt water. 



E.cpH''ni,<niU 100 and 101. 

 These experiments, in which carp were used, are decidedly puzzling. The 



output of chlorine per 100 grams of body weight was in one case 0.009 gram 



in 5 hours, in another it was 0.015 gram in 16 hours. 

 Since these fishes had not been fed for 10 and 13 days, res,pectivcly, the salts 

 here indicated can not have been pas.sed out in feces. I can only mention, by 

 way of possible explanation, that the fishes here used had been living in fresh water 

 which received a slight, though undetermined, admixture of salt water. It will })e 

 shown below that even very small proportions of sea water have a pronoiinc(>d effect 

 upon the salt content of the body. 



It will be better to defer any general discussion of the foregoing results until 

 the analyses of tiie bodies of the fishes themselves have been considered. Certain 

 doubts which the reader may have formed regarding the significance of the preceding 

 (igures will then perhaps be dispelled. 



Experiments in which the fishes died, or in which they were killed and returned 

 to fresh water, seem to show (1) that a dead fish will yield a larger proportion of 

 its salts than a living one, but that (2) there is no sudden increase in the difl:'usion 

 of these after death. These conclusions were not tested very thoroughly by 

 experiment, iiowever. and they are hence only offered provisionally. 



DETEKMINATIONS OF THE CHLORINE IX THE TLSSUES OF FISHES. 



Each lot of fishes here used was first rinsed in fresh water for 5 to 10 minutes 

 ill order to remove any salt water which might adhere. They were then weighed, 

 after which the entire fishes were cut up into fine pieces and placed in a porcelain 

 evaporating dish or crucible. Considerable (juantities of a mixture of equal parts of 

 sodium (or potassium) nitrate and sodium carbonate were now added in order to 

 facilitate combustion, and the whole mass was heated to the point required for igni- 

 tion. It was, of course, necessary that all chemicals used should be chlorine-free, 

 and it was likewise neccssarj' to avoid any contamination of materials with salt water. 



The soluble salts were extracted from the residue with ])oiling water. The 

 resulting solution was first treated with nitric acid to transform any phosphates 

 present, and then rendered neutral with calcium carbonate. After filtration the 

 solution was titrated with silver nitrate according to the method already described. 

 If care had been taken to insure the complete combustion of all organic matters, the 

 determination was simple and fairly exact. 



<i This figure is probably ton high (see p. S51. 



