inhibition were brought about by inactive precipitin (pre- 

 cipitoid) interacting with molecules of antigen, and thus 

 preventing the precipitating action of any active precipitin 

 that may be added, then it should be possible to overcome 

 the inhibition by an excess of antigen, thus providing 

 dK...-tigaged molecules to interact with the active precipitin. 

 But in our experiments, when inhibition was complete, we 

 have never been able to overcome the inhibition by any 

 further addition of antigen. (In these experiments the 

 amount of antigen originally present was sufficient to 

 discharge the greater part of the precipitable content of 

 the antiserum). In similar circumstances we have usunlh 

 been able to overcome the inhibition by a further addition 

 of antiserum— an observation which by itself is not incon- 

 sistent with Ehrlich's hypothesis and the theory of mass 



When regard is paid to the exact amounts of the inter- 

 acting substances in an inhibition experiment, further 

 difficulties arise, since unheated antiserum is never able 

 completely to neutralise the antigen, while the same 

 quantity of heated antiserum may completely inhibit pre- 

 cipitation from like quantities. It would then be necessary 

 to assume that the combining affinities of precipitoid could 

 be effectively distributed over a much greater number of 

 molecules of antigen than could those of precipitin. 



Without going into detail, we may say that our results 

 are consistent with the hypothesis that a heated inhibitory 

 antiserum acts not on the antigen, nor on the antiserum, 

 but on the product of their interaction— the precipitate— 

 and that inhibitory action is of the nature of a solution of 

 precipitate (29, 30). There appears to be a quantitative 

 relation between the amount of heated antiserum and the 

 amount of precipitate that it will inhibit (dissolve). The 

 fact that inhibition can not be overcome by excess of antigen 



