EFFECT OF ETHER AND CHLOROFORM ON FISHES. 243 



c.c. per liter (3,240 mg.) the time was 1-2 minutes but below 

 this concentration there occurred either a sudden decrease in 

 initial toxicity or an equally strong increase in the resistance of 

 the fish. This culminated at 4.3 c.c. (3,096 mg.) per liter in an 

 "abnormal" time of n minutes, and from there, with further 

 decrease in strength of solution the initial toxicity increased to 

 the norm (1-2 min.) at a concentration of 3.7 c.c. per liter 

 (2,664 mg.). 



If a relation such as this be found for fish in standardizing 

 some drug, it will be both quicker and cheaper to use it instead 

 of death. Decreased initial toxicity with decrease in strength 

 and close approach to hour killing concentration would be in 

 no wise remarkable if it did not increase again to the norm as 

 the concentration fell to the hour killing concentration as deter- 

 mined. But enough fish were used to render this inexplicable 

 phenomenon reasonably certain. 



4. Chloroform: the Sunfish. The series run here were also 

 plotted as explained for ether. See chart III. The hour death 

 concentration lies between about .07 and .1 c.c. per liter (106.89 

 and 152.7 mg.) and much variation is exhibited between these 

 points. The rest of the curve shows no such variation but it 

 may be that this was due to the small number of experiments 

 used to determine them. It is equally likely that only certain 

 limited regions exhibit variability. Should this be the case in a 

 drug needing standardization the variable areas should be 

 located and no standard set anywhere within one of them. 

 Continuous variation throughout a curve would, if wide enough, 

 forbid the use, for accurate standardization work, of the experi- 

 mental animal used in making the curve. 



IV. COMPARISON OF ETHER AND CHLOROFORM 



KILLING CURVES. 



It has been stated that these curves are hyperbolae or very 

 close to them. It should be further noted that the maximum 

 death variability seems to coincide with that part of the curve 

 showing greatest curvature. The laws governing hyperbolae 

 will probably be found to be of great use in fish standardization 

 since they seem to die according to one. Twenty-four experi- 



