60 ILLINOIS BIOLOGICAL MONOGRAPHS [180 



Results which will allow the above interpretation have been derived in- 

 dependently by dijBferent workers (but not thus interpreted by them) in 

 different groups of the animal kingdom as well as the plant kingdom. All 

 will stand the above tests. Ostwald (1905) in his work on toxicity of sea- 

 water and the constituents of sea-water on a fresh water isopod {Gammarus 

 ptdex De Geer) plotted survival time of the isopod as ordinate and proportion 

 of sea-water or its constituents as abscissa. The curve thus formed when 

 examined casually resembled an hyperbola, however he did not designate the 

 curve as such, but stated, "I have not been able so far to figure out a formula 

 that holds for the whole curve." Ostwald did not construct the reciprocal 

 or velocity of fatahty curve. When this curve is constructed from his data, 

 (Table, page 72, showing survival times of male and female Gammarus in 

 different proportions of sea-water, 1905) it resembles very closely the velocity 

 of fatahty curves shown in Fig.-.. 1 to 20. Doubtless the resemblance would 

 have been more striking had Ostwald tabulated his data which were omitted 

 and had he run a few experiments with higher concentrations, as he states, 

 "it was only from concentrations of 20 parts of fresh water with 80 parts sea- 

 water and higher, that visible toxic effects began to appear. But these figures 

 thus obtained were still so large and varied so much that I excluded them 

 from my experiments. " And later he says that the toxicity of all salts inves- 

 tigated increased very slowly at very low concentrations with increased con- 

 centration of solution depending on the nature of the salt and at stronger 

 concentrations there was a sudden rise, while at still higher concentrations 

 there was again a slow rise in toxicity. This is in complete accord with data 

 obtained in this investigation. He then suggests that this din"erence 

 in the rate of increase of the toxicity of a substance with increase 

 in concentration at very low and at very high concentrations was due 

 to the inability to measure very exactly the low concentrations and to 

 inexactness in determining the death point in a very short survival period. 

 Ostwald (1905, 1907) then disregards these extremes and applies to his 

 mean data the absorption formula (Ostwald, 1906) c = KC" where a = 

 amount of salt absorbed, C = concentration of the solution, and K and m 

 are constants depending on the nature of the salt investigated. He then 

 assumes that the survival time of the Gammarus is inversely proportional to 

 the amount of salt absorbed, i.e., l/t = a, thus l/t = KC'" or tC = Ki. Later 

 Ostwald (Ostwald and Dernoscheck, 1910) recognized the inabiUty of the for- 

 mula to fit his experimental data, and revised his formula by substituting 

 (C-n) for C where « = amount of the salt tested normally found in the blood 

 of the experimental animal. But it is difficult to see why such a substitution 

 should be made wih substances not normally found in the blood of the animal. 

 This as has been pointed out would be necessary with all substances tested, 

 since all undergo the same variation in toxic activity with variation in con- 

 centration as that pointed out by Ostwald. The only exceptions are CuCl2, 



