:^t; 



that lends support to the common view of the precipitin 

 reaction, since in the earlier partial interactions (Nos. 1 to 

 5) the weight of precipitate is increased with increasing 

 weights of antigen. 



Table III. 



In the second experiment (Table III) 14'4 mg. antigen 

 served completely to precipitate the content from 2 cc. 

 antiserum, and further increments of antigen, even to 

 thirty-fold, did not increase, nor diminish, the weight of 

 precipitate. This antiserum is characterised by a moder- 

 ate precipitablc content (about 1*7 mg. per cc. of antiserum) 

 and a relatively low precipitability. In these total inter- 

 actions the weight of precipitate is independent of the 

 weight of antigen. 



These observations show that it is only in partial inter- 

 actions that the weight of the precipitate alters with the 

 weight of antigen (c/\ Table II). In a total interaction 

 the precipitate has reached a maximal weight, and does 

 not vary with further additions of antigen (cf . Table III). 



(<) Relation of antigen to precipitate.-When a fixed 

 quantity of the antigen, however minute, is allowed to 

 interact with successive amounts of the antiserum, the 

 antigen is not completely removed from the superfluid. In 

 many of our experiments with quite small amounts of 

 antigen and relatively large amounts of antiserum, we had 

 no evidence that the antigen was sensibly diminished. We 

 found that the superfluids were capable of interacting again 

 and again to yield similar precipitates with fresh addiments 



