MICHAEL HEIDELBERGER 



is a static affair and aggregation a nonspecific aftereffect. Insisting 

 that the evidence for the multivalence of antibody, at least, is inade- 

 quate, he has produced no satisfying alternatives and his publications 

 show evidence that underground movements are weakening his hold. 



Study with Kabat (9) of a system of bacterial agglutination 

 showed this reaction to be subject to laws similar to those underlying 

 the precipitin reaction, and for the first time placed the important 

 function of salts in a perspective consistent with modern protein 

 chemistry. But a theory that does not lead to useful and exciting 

 predictions is like a sterile "Man O' War." In its explanation of salt 

 effects observed in a study with Teorell (14), the quantitative theory 

 developed with the new analytical methods led to a prediction as to 

 how analytically pure antibody globulin might be obtained (13,14); 

 and this was realized in two laboratories, (5b, 10). Since, with Pedersen 

 and Kabat (17,24) it was shown, with the help of Svedberg and his 

 ultracentrifuge and Tiselius (35) and his electrophoresis apparatus, that 

 purified antibodies have the properties of serum globulins, the protein 

 nature of antibodies can no longer be doubted. 



The technically important toxin-antitoxin reaction has been 

 worked out analytically by Pappenheimer and Robinson (32), so 

 that it is now possible, aided by a single pair of nitrogen estimations, 

 to measure the unitage both of an unknown diphtheria toxin and an 

 unknown antitoxin. To Ehrlich this would have seemed a pro- 

 digious feat, but the "brass" of immunology has failed to legitimize it. 



Quantitative immunochemical methods are, however, being 

 applied in many other directions. They were used, with Mayer, for 

 a study of reversibility and velocity in the precipitin reaction (20,31). 

 Through further modifications, with MacPherson, they have been 

 extended to the measurement of micrograms of antibody in human 

 sera (7,16). The new methods have also supplied an absolute measure 

 of complement, furnishing a weight unit independent of the lytic 

 function of this strange and little understood complex (6,15,18,22). 

 With this measure, the reacting proportions of complement, hemolysin, 

 and red cells could be calculated, as also those of antigen, antibody, 

 and complement involved in the process of complement fixation. 

 Oddly enough, complement could be fitted handily into the quanti- 

 tative theory of the precipitin reaction, which had been elaborated in 

 blissful neglect of what might have been an embarrassing apparition. 



