THEORIES RELATING TO ANAPHYLAXIS 105 



as we have previously shewn. Now, the toxin 

 formed in this case is evidently very different from 

 the one which might possibly be formed in the 

 organism. 



But what should especially make one reflect is the 

 fact that an active anaphylotoxin is obtained when, 

 of the three substances, one, the antibody, is com- 

 pletely suppressed. Thus, simply by making the 

 complement act on the antigen, as in the case of the 

 tubercle bacillus or a bacillus of such slight virulence 

 as the B. prodigiosus, a highly toxic anaphylotoxin 

 is obtained. 



That is not all. Keysser and M. Wassermann^ 

 have shewn that the anaphylotoxin can also be 

 prepared by suppressing the antigen and making the 

 complement act on a substance as inert as barium 

 sulphate or kaolin. 



Finally, Doerr and Russ,^ as well as Seitz,^ have 

 gone still further; they have discovered that com- 

 plement is not indispensable to the constitution of 

 an anaphylotoxin. 



Thus, Doerr and Russ mix horse serum with the 

 serum of a rabbit immunised against horse serum. 

 After twenty-four hours the precipitate is separated 

 by centrifugalisation from the fluid part. Now, 

 whether one injects into the guinea-pig's veins the 

 precipitate or the fluid, the animal manifests the 

 symptoms which are produced by the anaphylotoxin 

 prepared with complement. 



Seitz has arrived at the same result in a more 

 convincing manner; he prepares dysentery anaphylo- 

 toxin by treating the dysentery bacillus with guinea- 

 pig serum, which he heats beforehand for an hour 

 at 65° C. Now, in spite of the destruction of the 



^ Folia Serologica, vii., pp. 243, 593, 191 1; Zeitschr. f. Hygiem 

 Ixviii., p. 535, 191 1. 

 2 Centralbl. f. Bakt., I. Orig., Ixiii., p. 243, 1912. 

 ^ Zeitschr. f. Immuniidisf., I. Orig., xiv., p. 91, 1912. 



