ANTIGEN-ANTIBODY REACTIONS 447 



the mixture, but a complex containing more or less of the 

 reagents according to the proportions in which they 

 were mixed. This mechanism affords no explanation of the 

 specificity of such reactions, nor does it account for the 

 fact that on heating the toxin-antitoxin complex soon 

 after its formation dissociation occurs, but heating an 

 hour or more after the reaction has occurred no longer 

 has any effect. It has been shown that the adsorption of 

 toxin by colloidal ferric hydroxide and by antitoxin both 

 follow the adsorption isotherm, but that the iron complex 

 is just as toxic as the original toxin, Avhilst the toxin- 

 antitoxin mixture is non-toxic ; in other words, adsorp- 

 tion alone is not adequate to explain the neutralisation. 

 Bordet claimed that the neutralisation occurred as a 

 secondary reaction after the adsorption was complete. 

 The Danysz phenomenon is readily explained on the 

 adsorption hypothesis as being due to the complete 

 covering of the antitoxin by the toxin first added, toxin 

 subsequently added not being completely adsorbed. 



Precipitin Reactions. — When a soluble antigen, such 

 as a protein, is allowed to react with its corresponding 

 antibody precipitation follows. The precipitate consists 

 of 70 to 90 or 95 per cent, of antibody globulin, most of 

 the remainder being the antigen, although there is often 

 a variable amount of lipoid material also present. The 

 preponderance of antibody in the precipitate is not due 

 to non-specific adsorption by the antigen-antibody .com- 

 plex, since added normal serum is not taken up and does 

 not influence the proportion of antibody and antigen in 

 the precipitate. Dean and Webb showed that the ratio 

 of antigen to antibody which gave the quickest reaction 

 (that is, showed precipitation first) was a constant for 

 any given system. They showed that, in general, in such 

 mixtures neither an excess of antigen nor of antibody 

 could be detected in the supernatant solution by the 

 addition of one or other of the reagents. It was also 



