THE HOMOGENEITY OF MONOSPECIFIC ANTIBODIES 249 



fixation. We have noted that precipitation is a general form of reaction between 

 all antigens and all antibodies, including toxin and antitoxin. And we have 

 seen that the union of antigen and antibody at the surface of a bacterial cell sensi- 

 tizes it alike to the flocculating action of electrolytes, to the lytic action of comple- 

 ment, and to the phagocytic action of leucocytes. 



There are too, in records of the assay of the protective power of antisera, many 

 instances of complete parallelism between the protective power and some in vitro 

 agglutination, antihsemolytic or antitoxic reaction. For instance, the type- 

 specific precipitin content of antimeningococcal sera was found to parallel their 

 protective power in mice (Pittman 1943). 



The detailed evidence has been reviewed by Marrack (1938). It remains to 

 note that Delves (1937) found that precipitating antibodies to pure human albumin 

 and pseudoglobulin would agglutinate and opsonize collodion particles coated 

 with homologous antigen ; that Gerlough, Palmer and Blumenthal (1941) were 

 able to establish the identity of precipitins, agglutinins and protective antibodies 

 in antisera to six different types of pneumococcus ; and finally, by their method 

 of dissociation of agglutinated bacteria, Heidelberger and Kabat (1936) and Alex- 

 ander and Heidelberger (1940) were able to demonstrate directly the identity of 

 the agglutinin and precipitin to Type I pneumococcus and to Hwnwphihis inflxienzcr. 

 Type b. 



This conception of the serum reactions does not, of course, in any way modify 

 our belief in a multiplicity of antibodies corresponding to a multiplicity of antigens. 

 A red cell, a bacillus, or a crude protein solution such as horse serum, contain many 

 antigens and give rise to many antibodies. The unitarian hypothesis, as Zinsser 

 (1921) has emphasized, implies simply that the injection into the tissues of a chemi- 

 cally pure antigen will lead to the formation of one antibody capable of producing 

 all the various manifestations of antigen-antibody union. 



When we discuss, in the next chapter, the antigenic structure of bacteria, we 

 shall see that there is a sense in which it would be correct to differentiate one 

 antibody from another in terms of function — to say, for instance, that a particular 

 antibody is an agglutinin but not a lysin. This difference in functional activity, 

 however, is determined not by a difference in the nature of the antibody but by 

 a difference in the structural position of the antigen to which it is attached. 



The Homogeneity of Monospecific Antibodies. — It is often assumed that the 

 behaviour of antibody in solution, provided that it has been formed in response 

 to a single antigen, is the sum of the effects of a homogeneous collection of anti- 

 body globulin molecules. We have already discussed a number of phenomena that 

 indicate a heterogeneity of the antibody in a given serum. Thus, in monospecific 

 sera there is an antibody that precipitates with antigen only in the presence of 

 fully reacting antibody (Marrack and Smith 19316, Heidelberger and Kendall 

 1935a, h, c, Heidelberger, Treffers and Mayer 1940). Heidelberger postulated 

 univalence for this low-grade antibody, and multivalence for the more "avidly" 

 reacting antibody, Landsteiner and van der Scheer (1940) found in an anti-hen 

 ovalbumin serum two kinds of antibody both precipitating with hen ovalbumin ; 

 turkey albumin precipitated readily with one, and failed to precipitate with the 

 second, though there was evidence of combination. Other evidence of hetero- 

 geneity of antibody is provided by the change in quality of antibody during the 

 course of immunization. The broadening of reactivity and the increase in avidity 

 of antiserum with prolonged immunization has been observed many times. Hooker 



