148 



COLLECTED STUDIES IN IMMUNITY. 



previous experience we took two agar cultures to be the normal 

 measure for one rabbit. Since small amounts of bacilli were lost 

 through the centrifuging, we often employed somewhat larger amounts 

 for the injection of the agglutinated bacilli; while, on the other 

 hand, the control animals frequently purposely received less than 

 two agar cultures. This was done to meet the objection that the 

 animals injected with agglutinated bacilli had received fewer bacilli 

 than the control animals. But just in these control animals which 

 therefore received different amounts it was seen that a strict paral- 

 lelism between the amount of bacilli injected and the agglutinating 

 value produced thereby does not exist. Many animals with smaller 

 doses exhibited higher agglutinin values than other animals with 

 larger doses, as is seen by the following table I. 



TABLE I. 



The injection was usually subcutaneous, a few times intraperi- 

 toneal. The blood was abstracted from the ear vein. 



Testing the agglutinating value of the serum was accomplished 

 according to a method long in use in the bacteriological division, as 

 follows : 



The serum dilutions (in 0.85% salt solution) were usually Vso, 

 1 Ao, Vso, Vieo, etc.; finer gradations were not employed, as they 

 are of no value in measuring the agglutination. The culture used 

 was a living 20-hour agar culture which was suspended in 10 cc. 

 of bouillon. To each serum dilution, whose volume was 1 cc., the 

 same amount of bacilli was added (1 cc. bouillon culture), so that 

 the total volume of each specimen was 2 cc. Each specimen was 

 then poured into a little Petri dish and placed into the thermostat 

 for two hours. Thereupon the specimens were examined with the 

 low power of the dry objectives. In this way the occurrence of 

 larger or smaller clumps is very distinctly seen. In the protocols 



