160 COLLECTED STUDIES IN IMMUNITY 



The blood is once more washed and centrifuged and then filled up 

 to 100 cc. The blood-cells thus saturated with immune body are in- 

 jected in rabbits intraperitoneally, each animal receiving 25 cc. of 

 the mixture. 



At the same time control animals are injected with the same 

 amounts of normal ox blood. 



Usually on the tenth day after the injection, as this had shown 

 itself the most favorable time, serum was withdrawn, inactivated and 

 tested for its content of immune body by adding it in decreasing 

 quantities to 1 cc. 5% ox blood plus sufficient complement. Either 

 rabbit serum, 0.4-0.5 cc., or guinea-pig serum, 0.1-0.15 cc., served 

 as complement, for these are equally well adapted for this purpose. 

 The results of the experiments are as follows: 



Out of eight rabbits injected intraperitoneally with ox blood 

 saturated with immune body, only three corresponded to the requirements 

 which follow from von Dungern's results. Their serum, tested exactly 

 like the immune body, failed even in amounts of 1.0 cc. to produce 

 a trace of haemolysis, whereas when the serum of the corresponding 

 control animals was tested, 0.025 and 0.05 cc. respectively sufficed 

 to effect complete haemolysis. 



These results are approached very closely by the serum of a 

 fourth animal. The haemolytic action of this serum compared to 

 that of the serum of the corresponding control animal was 1: <135, 

 i.e., was exceedingly slight. The remaining four rabbits had pro- 

 duced an immune body in greater or less amounts, though this amount 

 was always far less than that produced by the corresponding control 

 animals. When the absence of a zone of marked complete solution 

 rendered it impossible to make an exact determination, the compari- 

 son of the immune body values of the sera in parallel tests was accom- 

 plished by comparison of tubes whose colors corresponded. The 

 amount of immune body possessed by these animals compared to that 

 of the corresponding control animals was as follows: 



(1) 1:5; (2) 1:7; (3) 1:10; (4) 1:10. 



I have supplemented these experiments w r ith a smaller series of 

 experiments made with intravenous injections. In these, of course, 

 very much smaller amounts of blood were used for injection because 

 when the blood-cells loaded with immune body are injected directly into 

 the circulation, they suffer haemolysis through the action of the com- 



