86 BULLETIN OF THE BUKEAU OE EISHERIES. 



amount (Imiiiy the tirst liouf. iiiid siiuilurly only 37 per cunt more ]1a^^sed out (iuriiii;- 

 two diiys than during tln' lirst day alone. Thus the loss of chlorides from the h()(l\ 

 occurs at a steadily diminishinu- rate. This conclusion is conlirmed l)y all the -uc 

 ccedini'- experiments. 



Fishes with same history as last. Al)Out twice as much chlorine was recorded 

 for the tirst hour. An accid(Mit jirevented furtlier determinations. 

 Erper!iHri,f S.J. 



F. }i,'frr<H-i;f„s. which had been in water of density l.no.", ( 1.004 to l.uo.;) 

 for 5 days, and were taken originally from a salt-water taidc ( l.nl:-i): ().(«(; 

 gram per Ino in I day. ().(i.")l gram per ion in 2 days. 

 E.vperhn,i,t Sl^. 



Same species, with same histoiy: n.Oii-J gram per lno in 1 day. ().(»:i.") g-ram 

 per 100 in 2 days. 

 Experimt'nt S-j. 



Hame species from same tank originally, hut in 1.O05 water 10 days: 0.O31: 

 gram per 100 in 1 day, O.olr-t gram per 100 in 2 days. 



In each case the tishes were well at the end of the second day of the 

 experiment. 

 Experiments S3 and 85 agree fairly closely, but experiment 84, for some 

 unknown reason, shows a much smaller loss of chlorine. The mean results of these 

 three experiments are 0.031 gram per 100 in 1 day; 0.040 gram per 100 in '2 day-. 

 In experiment 81 these figures were 0.030 and 0.041, i-espectively — i. e., there is a 

 practical identity between the two sets of figures. Where such great variation- 

 occur as in the above results the mean of a large number of determinations should of 

 course be taken. Provisionally it ma_v be concluded, however, that fishes from 

 water of density 1.005 will yield aijoutas much chlorine in fresh water as fishes from 

 water of density 1.013. 

 Keperlir.ent 86. 



In this case the fishes had been kejat 3 days in water of density 1.023 before 

 the commencement of the experiment (originally from l.(il3). Results: 0.01.' 

 gram per 100 in 1 hour, 0.042 gram per 100 in 1 day. 



The figures are larger than any of those previously given. 

 In order to test the possibility that the chlorides thus found had left the body by 

 way of the alimentary canal or reproductive organs, the following control experi- 

 ment was performed. 

 E.,'perl,i,ent 87. 



Ten fishes having the same history as those used in the preceding experi- 

 ment were killed, and from them were taken the entire alimentary canals and 

 the gall bladders, testes, and ovaries. These were cut into fine bits; then 

 macerated 15 minutes or more in fresh water. This latter was found to vield 

 0.005 gram chloi'ine. Thus this lot of fishes, weighing 106 grams, could have 

 furnished from the contents of their alimentary canals and reproductive organ- 

 not more than five thousandths of a gram of chlorine, and probably much less 

 than this. 



