58 HULLKTrN OK THK BCREAH (»F KISHEKIES. 



Throughout all of these (ixpoiiincnts. licaltliy Hs!u>s were selected unless otlu i 

 wise specified. Fishes having the siunc iii-^torv as tiiose used for experiment w i. 

 counnonly kept in the res(>rvc tanks an, I served for purposes of control. Spe( ii 

 control experiments were likewise carried on from time to time, as will appear. 



In a number of other expei'iments tiie harmfid ettects of transfer to fresh wair 

 upon F. h,i,r,><lU>i.< were manifest. 



E.rp.nnn.tS. 



August, ItXH. Of iftj fishes, none died until the lapse of (> days, whn 

 they began to succumb. Fifty -eight per cent had died in l(i days, when tin 

 experiment was discontinued. 

 E.,-p,:rhiient 9. 



August and September. 11)04. The fishes began to die in 3 days, and .'.' 

 per cent died during the first Kt days. From tlie thirteenth to the nineteenil 

 day, however, no deaths occuned and tlie experiment was discontinued. 

 Krj„r!inent 10. 



September, l!t04. Of l'02 fishes used, :5 died during the first day, wiiile ;•: 

 per cent died within In days. 

 Garrey (li)( );")), speaking of some experiments of his own upon this species of tish 

 states: "It was found that if care was taken to select individuals which were ih' 

 injured in catching, about SO percent lived in fresh water for 6 weeks, when tin 

 expei'iment was discontinued. This is as liigh a percentage as I'an be kept alive ii 

 the sea-water aquaria of the laboratory." ( iari'e3-'s statement is suriu-ising in \ iev 

 of my own results alreadv cited, and of the confirmatory experiments recorded below 

 It is to be remarked that many of my experiments were carried on at about the >ainc 

 time as those of (larrey, and that the fresh water employed came from the same w atei 

 system as that used at the Marine Biological Laljoratory, where Garrey's work wa 

 done. The fishes that I used were generally treated with the utmost care and in man\ 

 specified cases were fed throughout the experiment. Fishes from various localitie 

 were used. The conditions were likewise varied from time to time so that irrelevan 

 factors might be eliminated. 

 E.rp.rhiKiil 11. 



.Iidy. li»(>,'i. Thii'ty specimens of F. IntircdituK put into a hatching bli^ 

 filled with running fresh water. Another oO specimens taken from the sann 

 lot of fish were placed in an adjacent hatching box filled with running sal 

 water. Both sets of fishes were fed. although those in the fresh water show et 

 little desire for food. The temperature of the salt water was about 3 (' 

 higher than that of the f resli. 



In the fresii-water tank 1 fish died within •> days, and the last died withii 

 14 (lavs. In the salt-wat(>r tank not a single fish died during this period. 

 F.rp.rh..„i 1.1. 



Summer of 190.j. The first fish died after 7 days in fresh water, and a single 



fish remained living at the end of %) days. 



It is to be added that I have given the results for all of the experiments of tlii^ 



nature performed by me, no contradictory figures being suppressed, here or elsewhere 



That the fresh water hci-e used was not contaminated with any actively poison 



ous substance is shown by tiie fact that the addition of even a very small proportioi 



