504 STUDIES IX [MMUNITY. 



li was quite reasonable to presume thai freshly obtained, unheated 

 cholera scrum should act both as an immune serum and as fresh 

 normal Berum; and this, indeed, was what happened. Under these 

 conditions il alone produced metamorphosis of the cholera vibrio. 

 When healed to 55degrees il losl its bacteriolytic power, but recov- 

 ered il on the addition of fresh normal serum. Such was the dem- 

 onstration of the two substances, the collaboration of which is 

 necessary to produce bacteriolysis, — the sensitizer or preventive sub- 

 stance, thermostable, specific, and characteristic of immune serum, 

 and the alexin destroyed at 55 degrees, — non-specific, and present in 

 1 »ractically the same amount in the sera of normal and of vaccinated 

 animals, as may be shown by proper mensuration. While carrying 

 out these experiments (1895), it was noted that cholera serum im- 

 mobilizes and agglutinates cholera vibrios in clumps ; this was the 

 first instance of agglutination of a bacterial culture on the addition 

 of a small dose of immune serum. It was further found that this 

 agglutination would still occur, even with cholera scrum that had 

 been heated to (10 or 65 degrees, from which fact it was evident that 

 the agglutinating property is separate from the bactericidal property. 

 It became obvious at once that one might use this test which is so 

 easy to perform in vitro, in the diagnosis of the cholera vibrio, 

 instead of the in vivo method which Pfeiffer discovered. This 

 marked the introduction in bacteriology of serum diagnosis in vitro. 



As may be seen, the idea of a participation of two substances in 

 bacteriolysis is not, properly speaking, a theory, for there is nothing 

 hypothetical about it; it is simply a literal translation of observed 

 facts, particularly the experiments of "reactivation" which have 

 just been mentioned. Following is an account of the discovery of 

 hemolytic sera: 



In my experiments in bacteriolysis I frequently used as an im- 

 mune serum, the serum of a goat that had been immunized against 

 the cholera vibrio, and as alexin (complement), fresh normal guinea- 

 pig serum. It frequently happened that the latter contained a cer- 

 tain number of red blood cells and I found that these were agglu- 

 tinated and would be agglutinated even when mixed with normal 

 goat serum. I had, moreover, already noted that neither motility 

 nor vitality was a necessary condition for the agglutination of bac- 

 teria; cholera serum agglutinates vibrios that have been killed by 



