242 HAROLD D. CLAYBERG. 



died in the same time. Above that concentration the toxicity 

 of the poison increased for the 10-20 gram fish and decreased 

 for the 2-9 gram fish up to about 5.3 c.c. per liter (3,816 mg.), 

 beyond which it remained constant. Below about 4.8 c.c. per 

 liter, on the contrary, the toxicity of ether increased in the case 

 of class I, and decreased in class 2 to a maximal difference at 

 about 4.1 c.c. per liter (2,952 mg.) below which the two converged 

 until they met at 3.7 (2,664). Their crossing point is at the 

 hour death concentration. If, at certain points, the death 

 curves of different weights of fish are found to cross, in standard- 

 izing aconite, for example, it would be well to take these points 

 as the standard and disregard the hour standard or any other 

 previously set. 



The third class furnished complete data only between 4 and 

 4.35 c.c. per liter (2,880 to 3,132 mg.). At the upper concen- 

 trations the toxicity is intermediate between classes I and 2. 

 But below 4.1 c.c. per liter (2,952 mg.) the toxicity falls away 

 far more rapidly than for the others. It is very likely that 

 concentrations below 3.9 (2,808 mg.) will cease to have toxic 

 effect of any sort on class 3. Since the critical concentration 

 is well below this point the significance of weight of the fish used 

 is seen to be of the utmost importance. It is regrettable that 

 no large fish were available to test out the end of the class 3 

 curve. Very small fish below 2 grams are, as might be expected, 

 killed in much less time than classes I or 2. Thus the resistance 

 of the fish is seen to be consistently different for different weights 

 and also of considerable complexity. 



The crossing of the curves of classes I and 2 renders both 

 classes of data available in the determination of the hour death 

 point. The above results point out that, in standardization of 

 drugs by fish, the different sizes may well prove to differ con- 

 sistently and considerably. For this reason it would seem that 

 the weight problem needs careful analysis and extensive experi- 

 mentation to determine the proper weight limits to be allowed 

 for results of the accuracy demanded. 



Examination of the fish led to the graphing on the same sheet 

 (see chart II.) of the time it took the fish in any concentration to 

 become definitely abnormal (loss of balance, etc.). Above 4.5 



