ANTIGEN-ANTIBODY REACTIONS 457 



contact was close over a considerable area then firm and 

 specific combination would occur. The forces which 

 hold the molecules together are not themselves specific ; 

 the specificity depends entirely on the appropriate 

 distribution of the atoms involved. The method by which 

 such complementary structures might be built up in 

 antibodies has been outlined on page 421. 



The evidence points to the fact that the second 

 stage of antigen-antibody reactions, especially in agglu- 

 tination and precipitin reactions, is also specific. Thus 

 agglutination carried out with mixtures of two bacteria 

 and the corresponding antisera normally gives clumps 

 which contain one type of organism only, and not the 

 mixture of organisms which would be expected if the 

 aggregation were non-specific. Similarly, precipitin 

 reactions carried out in the presence of a heterologous 

 antigen do not give precipitates containing the second 

 antigen. It is reasonable to suppose, therefore, that the 

 second stage, the actual formation of a precipitate, is a 

 continuation of the first stage until aggregates of sufficient 

 size to be insoluble are formed. This suggestion forms 

 the basis of the " framework " or " lattice " theory put 

 forward by Marrack and elaborated by Heidelberger and 

 by Pauling and their collaborators. 



The framework hj^jothesis requires that antibodies 

 be at least divalent, otherwise antibody molecules could 

 not form links between two or more antigen molecules 

 to give the framework structure illustrated in Fig. 11 

 which represents the state at the equivalence point. 



re- Anfigen 

 Antibody 



Fig. 11. 



