165] 



THE GOLDFISH AS A TEST ANIMAL— POWERS 



45 



T (toxicity) = a' 



tan, e 

 a 



This equation cannot represent the absolute or the 



exact relative toxic value of a substance since it is based only on a portion of 

 the velocity of fatality curve, the portion A to B, Figure 1, and the assumption 

 that this portion and the two extremes, i.e., the two extremes C to A and B to 

 G, follow the reciprocal of an equilateral h}^erbola which is not in keeping 

 with experimental data or a curve dra\vn from the hypothetical equation repre- 

 senting the true velocity of fataHty curve. See pages 48, 52 for a discussion 

 of this h}^othetical equation. However this expression has been chosen to 

 represent the relative toxic value of a substance rather than the curve itself 

 as suggested by Osterhout (1915), as the latter leaves the formation of a 

 criterion to the reader. It has the advantage in that the relative toxicity 

 can be given a numerical value and is a natural criterion and not an arbitrary 

 one. This equation, as has already been pointed out, does not express the 

 specificity of the toxic activity of the salt, which can be shown only by the 

 curve itself, as Osterhout has suggested, or by taking into consideration the 

 factors which go to make up an equation which represents the curve itself 

 (See h}^othetical equation page 49). Taking the above equation as repre- 

 senting the relative toxicities of the alkali and alkahne earth metals, as chlorides 

 and nitrates, the toxic values are found as given in Table XXXVI. These 

 values are comparable in general, not so much in comparative numerical values 

 but in position, to values obtained by other workers after the measurements 

 of the elements are reduced to a common standard. See bibliography for 

 citations. However, among all workers there is more or less difference of 

 opinion. 



TABLE XXXVI 



Relative Toxicities of the Alkali and Alkaline Earth Metals \\tien in 



Combination wtth CI and NO3 



