66 



BULLETIN OF THK BOKEAU OF FISHERIES. 



's with tlic preceding- except that tl 



Experiment SO. 



Woods Hole, Septenilier. I'.tOt. Ag-iv 

 water density was \SW.\. The results were tlie same. 



Water of dciisitj' 1. 002 was employed, and only 1 tish died, on the loth day. 

 after which the experiment was discontinued. 



During the season of 190.5 some individuals of this species were found to survixc 

 1.5 daj's (i. e., until the experiment was discontinued) in water having only al)i)ui 

 3 per cent of the salinity of the local sea water (experiment 31 bis). In this ca^i . 

 however, a large majority of the fishes had died in the meantime, and it would thcK 

 fore seem that the physiological limit of dilution had been passed. In a few ca^'^ 

 the same symptoms were observed as in the case of F. heteroditus in pure fro^h 

 water. (See p. 61.) 



The same result is less strikingly shown in the case of sculi)ins (experiment A*'<]. 

 Two of these tishes put into fresh water died within 24 hours (perhaps much les>). 

 while of 3 from the same lot which were put into water having a specific gravity nt 

 about 1.001" 1 survived 6 days, when death occurred, the remaining 2 for 3 day>. 

 when they were put back (living) into salt water. 



The water used in the last two experiments had only 3 to 4 per cent of the saliii 

 ity of pure sea water. If the mere change in the osmotic pressure of the surround 

 ing medium were responsible for the harmful effects, it would not be expected tiiat 

 transfer to water of a very low degree of salinity would result in little or no harm tn 

 the fish, when transfer to pure fresh water was so quickly fatal. This salutary influ- 

 ence of a very small proportion of salt in the water is later shown to be related to an 

 important difi'erence in its effect upon the salt content of the body (pp. 90 et sc(|). 

 The latter is shown to undergo a serious diminution in fresh water, while little i>r 

 no diminution occurs if the water be even slightly saline. 



EFFECT OF TRAXSFElt OF BRACKISH AND FUESH WATER FISHES TO SALT WATER. 



Experbiien ts 32-35. 



Woods Hole, July, August, September, 1904:. Eundulm (Uapha-nns from a 

 brackish pond (density 1.005 to 1.006) transferred directly to sea water. Tiie 

 record of deaths in the four cases reads: 



(I Discontinued. 



These differences in the death rate are doubtless due to the condition of the fishes 

 at the time of experiment. Those used in experiment 33, for example, had remained 

 unfed for about 3 weeks. During the summer of 1905 the experiments were repeated 

 with similar results, confirming (1) the harmful effects of this abrupt change of water 

 density, and (2) the variability of different lots of fishes in their capacit}' to endure 



the salinity fell to about one-half of this. 



