EIE PRECIPITIN REACTION. 331 



(the original antigen being sufficient to neutralise to 

 greater part of the antiserum) is explained by our finding 

 that, in these circumstances, no marked increase of pre- 

 cipitate occurs. The further fact that, in similar conditions, 

 an increase of unheated antiserum can overcome the 

 inhibition, is consistent with our finding that there is 

 always sufficient antigen left to produce fresh precipitates 

 with fresh addiments of antiserum. 



We may add that our precipitation (11, 31) and inhibition 

 (29, 30) experiments with closely related heterologous 

 proteins support our interpretation of the interaction of 

 antisera with specific antigens, and so strengthen our 

 general position. 



Conclusions.— 

 1. The main mass of the precipitate is formed by con- 



2. There is a close relationship between the weight of pre- 



cipitate and the amount of antiserum (cf. Table I). 



3. In total interactions the weight of precipitate is inde- 



pendent of the weight of antigen (cf. Table III). 

 *. In partial interactions the weight of precipitate is con- 

 ditioned by the quantity of antigen (cf. Table II). 



5. It is therefore erroneous to speak of the precipitin of the 



antiserum coagulating the antigen, or to regard the 

 antigen as the precipitable substance. 



6. Inhibition phenomena are inconsistent with the hypo- 



thesis that precipitoid is developed, but are consistent 

 with the finding that heated antisera act directly on 

 precipitate by specific solvent action. 



7. Our interpretation of the precipitin reaction appears to 



have important practical bearings. 



(a) On the identification of the specific origin of 

 proteins, and on the separation of closely 

 related species (11, 31). 



