Ill 



Tab. 11. 



Corresponding experiment with double concentration at 20 C. 



Experim. 3. 



Table 11 a. 



Same experiment at 20. 

 Experim. 4. 



The ciphers betwen brackets are extrapolated by 

 comparison with the next ciphers in the table. 



It appears from comparison of the different experi- 

 ments, that the results of the experiments at 37 C show 

 a very great mutual agreement, when we except the 

 ciphers for 0,1 antitoxin. Thus the errors of experiment 

 at the low temperature are much greater. By forma- 

 tion of new means the following values of x are ob- 



