568 



COLLECTED STUDIES IN IMMUNITY, 



suspended in salt solution. The result of the experiment is shown 

 in the following table: 



TABLE I. 



Here we see the curious result that with a certain excess of the 

 antiserum there is no inhibition of haemolysis. This paradoxical 

 phenomenon we observed only with the antiserum produced by 

 immune serum injections, and then only in the "curative" experi- 

 ment. If the antiserum was used for "protective" experiments, 

 i.e., mixed with amboceptor previous to adding the blood-cells, or 

 if the antiserum produced by injections of normal serum was employed, 

 the course of the experiment was entirely uniform, an increase in the 

 amount of antiserum causing an increase in the antilytic action. 

 For the present we are unable to say whether we are here dealing 

 with an essential difference between the antiserum produced by 

 normal serum and that produced by immune serum, or whether we 

 have to do with an individual fluctuation. So far as the mechanism 

 of the phenomenon is concerned we were able to clear up at least 

 one point, namely, that the essential factor in the experiment is the 

 presence or absence of the very small quantities of normal rabbit 

 serum which contains the amboceptor. Thus if the blood-cells are 

 sensitized with amboceptor without subsequently removing the serum 

 by centrifuging, it will be found that the course of the "curative" 

 experiment is perfectly regular. There is no inhibition of the antilytic 

 action with an excess of antiserum. The same holds true if we sepa- 



