Ill 



As usual a 2,5 % blood dilution in salt water was 

 used. The figures before S indicate how many 100 000 

 parts of serum were contained in the fluid added to 

 the blood. Already, when the serum reaches an amount 

 of 1,3 million parts of the bloodmixture , (addition of 

 0,1 cc. 0,013 % toxinfluid to 10 cc. blood emulsion) its 

 action is perceptible. The action increases nearly in 

 proportion to the amount of serum, until this has 

 reached 16 million parts (= 0,2 cc. of 0,081 %). In this 

 case the decrease in the action of lysin is about 50 / . 

 During the increase of serum from 16 to 250 million 

 parts, the action remains just the same. But from 

 250 million parts the action of the serum becomes again 

 effective. 



The figures in the last column are very unreliable, 

 because the blood dilution becomes strongly coloured 

 (yellow), when greater amounts of serum (5 10 %) 

 are added, and by this means the hasmolysis appears 

 much greater than it really is. And the determinations 

 are generally rather unreliable, when the haemolysis 

 is as slight as here (under 2). It is also observed in 

 this case that the increase of haemolysis with increasing 

 amounts of toxin becomes slower as more and more 

 serum is added. 



It appears from this circumstance how necessary it 

 is in these experiments to wash away the normal se- 

 rum from the blood corpuscles, before they are used. 

 The following experiment was carried out, partly with 

 blood corpuscles, twice washed with an isotonic NaCl 

 solution and centrifugalised, partly with corpuscles from 

 the same blood, only once centrifugalised and not 

 washed. The difference is very evident (amounting to 

 about 30 / ). The lysin was a 0,1 % solution, the 

 blood dilution 2,5 %. 



49 



