ANTIGEN-ANTIBOr>Y REACTIONS 449 



about 2 per rent, of the jjiecipitate at the optimum 

 proportion and about 0-6 per cent, at the end of the zone 

 of precipitation in the antibody excess region. The 

 pseudoglobuhn of horse serum, as antigen, forms about 

 20 per cent, of the precipitate at the optimum proportion 

 ratio. When antigens which can be separately estimated 

 in the precipitate and in the supernatant sohition are 

 employed it is possible to investigate the composition of 

 the precipitate also in the region of antigen excess. Thus 

 Heidelberger and Kendall, using R-salt-azo-diphenyl-azo- 

 egg albumin, a red dye, as antigen and estim^ating its 

 concentration colorimetrically, found that the precipitate 

 contained 13 per cent, of antigen at the optimum pro- 

 portion, 6-7 per cent, at the limit of antibody excess 

 and 33 per cent, at the limit of antigen excess. 



It will be seen that the precipitates do not have a 

 constant composition, the amount of antigen present 

 increasing with the proportion of antigen to antibody 

 in the mixture. The range of composition varies with 

 the particular antigen-antibody pair concerned. Marrack 

 and Smith showed that, using diazotised atoxyl coupled 

 to crystalline egg albumin or to horse pseudoglobulin, 

 or using iodo-egg albumin or horse pseudoglobulin as 

 antigens, the proportion of antigen to total protein in 

 the precipitate increased with the amount of antigen 

 added to a given amount of antibody. At the optimum 

 proportion, addition of either antigen or antibody to the 

 supernatant solution caused precipitation, showing the 

 presence of small amounts of antigen and antibody in 

 the solution, due to dissociation of the precipitate ; at 

 other ratios only antigen or antibody was to be found in 

 solution. They showed that the change in composition 

 of the precipitate was not due to non-specific adsorption 

 of the antigen or the antibody since the introduction of 

 normal serum or another antigen caused no change in 

 the composition of the precipitate ; for example, no azo- 

 protein was carried down in a toxin-antitoxin flocculation. 



