io8 



A. P. Mathews 



average value of K in these results is almost exactly the same as that 

 of McGuigan; i. e., about 2.2. The exceptions in these results are, 

 with the expection of cadmium, the same as those recorded by Mc- 

 Guigan, cobalt being too little toxic and zinc too toxic. I have deter- 

 mined K for zinc chloride in comparison w^ith all other metal chlorides, 

 expecting that, while the ratio would be too low for all other metals 

 above it in the scale of ionic potentials, it would be too high for all 

 metals below it, and these two errors should neutralize each other, 

 provided the metals were about equally distributed above and below 

 zinc. The result (Table 7) gave for the mean K 2.03, which is a 

 litttle low, but fairly close to the mean 2.2 already obtained. 



I have also determined (Table 8) the fatal dose just sufficient to stop 

 swimming in two minutes in rapidly swimming cultures of Volvox 

 globator. Four metals were investigated — i. e., silver, cadmium, 

 manganese, and magnesium. K was assumed to be 2.2, the same as 

 that for Fundulus and diastase, and the constant C was calculated. 



The result was as follows : 



Mean — 3.33 



The constant C is thus found as constant, as could be expected 

 from the methods used and the variability of the cultures. 



Undoubtedly the most consistent and accurate results thus far ob- 

 tained are those of McGuigan upon the minimum fatal dose of salts 

 for the diastatic ferment. These results are plotted in Fig. 3. I have 

 used only those results which were obtained with the metals Mg, Mn, 

 Zn, Cd, Co, Ni, H, Cu, Hg, and Ag, for the reason that the solution ten- 

 sion, and hence the ionic potential, of all metals above magnesium — 

 i. e., Ca, Sr, Ba, Na, K, Li, and Cs — are still so very uncertain as to 

 make the comparison of ionic potential and fatal doses of little value. 

 In Fig. 3 the hne A B represents the formula. 



JOg K =C- + A(^£,ca{JQji salt -C-cation colloid j ~r(.-£^anion salt -C-anion coUoidj • 



