MICHAEL HEIDELBERGER 



lying the estimation of antibody in absolute terms ; namely, that anti- 

 body, and antibody only, is quantitatively measured. Although the 

 serum of the type Ill-immunized rabbit contained up to 1 atom per 

 cent N^^ excess, the type I antibody isolated from this medium contain- 

 ing other antibodies, globulins, and albumins rich in heavy nitrogen re- 

 tained none of the N^^ within the error of the mass-spectrographic 

 method, about t. 0.03 atom per cent N^^ excess. 



The immunochemist has learned quantitative relationships 

 with even so complicated a test tube as the guinea pig, for Kabat and 

 Landow (26,27) have shown that 0.2 mg. of antibody suffices to sensi- 

 tize the animal so that fatal anaphylactic shock ensues with 1 mg. of 

 antigen such as egg albumin or 0.1 mg. of a type-specific polysaccharide 

 of pneumococcus. Either of these shock doses is far in the region of 

 antigen excess in the in vitro precipitin reaction, suggesting at once the 

 first simple explanation of how much smaller quantities of antigen 

 may desensitize the animal. Because small quantities ot antigen com- 

 bine with relatively much larger amounts of antibody, after the in- 

 jection of "desensitizing" small doses of antigen relatively little anti- 

 body remains for the final destructive effect of an injection of excess 

 antigen. 



All of which must not be taken to indicate that quantitative 

 immunochemistry, like a parachute flare, has had its episodic illu- 

 minating burst and is fading out. It has shown the way to objectives 

 of theoretical and practical interest, and many more applications are 

 apparent. One quantitative immunochemist has at last irritated our 

 native virologists into an appreciation of its potentialities in their vast 

 field of endeavor (25). Quantitative immunochemical methods are 

 also indispensable guides in the study of bacterial antigens and the 

 fractionation of many other natural carbohydrate and protein mixtures. 

 A recent application in the little studied field of lipoproteins was the 

 report of Cohen and Chargaff (4) on thromboplastin. Suspensions 

 of bovine thromboplastin removed measurable amounts of antibody 

 from specific antisera in rabbits, and, remarkably enough, the specific 

 precipitates showed greater thromboplastic activity than the antigen 

 contained in them. Displacement of the phosphatide in the antigen 

 by means of heparin did not destroy the power to precipitate with 

 antiserum. 



Moreover, if proteins or proteinlike structures are ever synthe- 



