ANTIBODIES II 29 



In any case, the cavity is not necessarily the deep invagination sug- 

 gested by Fig. 2-11 and example 1 in Fig. 2-12. The antigenic deter- 

 minant might alternatively be accomodated lying on its side in a 

 shallow trough (example 2 in Fig. 2-12), or sidewise in a sort 

 of slit trench (example 3 in Fig. 2-12). According to Pressman 

 (1957), there is evidence that all three types of antibody cavity 

 exist. 



If the combining group of an antibody molecule is relatively small 

 (one such group, according to the above estimate, would amount to 

 about 2 per cent of the surface of an antibody molecule), we naturally 

 ask how many such groups an antibody has. It seems conceivable that 

 one group would be enough to account for the reactions of antibody. 

 For some time Dr. S. B. Hooker and I and some other workers in 

 this field maintained that on the basis of economy of hypotheses 

 (Occam's razor) it should be assumed that antibody was univalent. 

 Others assumed that antibody was multivalent. There is now con- 

 siderable experimental evidence indicating that neither party to this 

 controversy was wholly right, for the valence of antibody seems to 

 be two. There are certain antibodies, especially in connection with 

 the Rh blood groups, behaving in peculiar ways which have led to 

 their being described as "incomplete" or "univalent." The presently 

 available evidence, however, indicates that the peculiarity of their 

 behavior is not due to their having less than the usual number of com- 

 bining groups but to other features of the molecule. 



Formation of Antibody 



We must now ask ourselves : How does the body manage to pro- 

 duce relatively large amounts of globulin molecules, so precisely 

 adapted to combining with definite chemical groupings? 



It is not easy to answer this question. A number of hypothetical 

 mechanisms of antibody formation have been proposed, of which we 

 may mention (a) the cast-off receptor theory of Ehrlich, (b) the 

 template theory of Haurowitz, (c) the template theory of Pauling, 

 (d) the "trained enzyme" theory of Burnet, and (e) the "natural 

 selection" theory of Jerne, which is supported by Talmage and 

 Lederberg. 



(a) According to the theory of Ehrlich, antibodies are simply 



