Ill 



be, roughly speaking, the case. A correction must be 

 introduced here, just as above, (p. 12) for the change of 

 volume through the addition of toxin. In this case, af- 

 ter we have supposed that the reaction proceeds in pro- 

 portion to the quantity of blood also, the correction 

 should be applied as well to the calculation of the con- 

 centration of the blood as to that of the lysin. If the 

 concentration and the velocity of reaction are proportio- 

 nate to one another, in the present case the times, in which 

 the same degree of haemolysis is obtained, should not be 

 in inverse proportion as in 1:2, 4:8 but as (1 : 1.075 2 ) : 

 (2 : 1.05) : (4 : 1.10") : (8 : 1.20*) = 0.865 : 1.814 : 3.345 : 5.554 

 = 6.4: 3.06: 1.66: 1 



Through the correction for the quantity of fixed am- 

 monia, these ciphers will be changed to 7.64 : 3.33 : 1.73 

 : 1.02 --1: 0.44: 0.23 : 0.133. 



Through simple interpolation from the above ciphers 

 it is found that the following degrees of haemolysis re- 

 quire the times mentioned below to become hiemolysed: 



X 



A 

 B 



C 

 D 



In parenthesis ciphers are quoted, which are calcu- 

 lated under the following suppositions: proportionality 

 between haemolysis and the amounts of ammonia -- and 

 that the observations in the first series (A) are right. 

 The concordance between the calculated and observed 

 values is very good, allowing for the great errors of 

 experiment. The greatest deviations are found between 

 the first figures in the D series. It ought not to sur- 

 prise us that rather important errors in the determina- 



29 



