CHEMICAL ACHIEVEMENT— PAULING 237 



attraction between the haptenic group and the antibody, as a result 

 of the smaller van der Waals attraction of a hydrogen atom and of 

 a chlorine atom for surrounding atoms. On the other hand the cavity 

 in the anti-m-azobenzoic acid antibody is required only to be large 

 enough to receive a hydrogen atom, with van der Waals radius 1.2 A. 

 There might well then be a considerable amount of steric strain if the 

 4-chloro-3-azobenzoic acid haptenic group were to be forced into this 

 cavity in the antibody, and the steric strain might be great enough to 

 decrease the combining power to such an extent that no precipitate 

 would be observed by the investigators. 



This experimental result indicates that the fit of antibody to anti- 

 gen is, in some cases at least, a very close one, so that a difference in 

 atomic radius of 0.6 A. is significant. Our quantitative investiga- 

 tions in Pasadena provided a large amount of evidence substantiating 

 this conclusion. One extensive series of investigations was made of the 

 combination of antisera homologous to the o-benzenearsonic acid 

 haptenic group, the m-azobenzenearsonic acid group, and the ;?-azo- 

 benzenearsonic acid group. It was found that in each case the sub- 

 stituted benzenearsonic acids with the substituent in the same position 

 as the azo group of the innimnizing azoprotein combine more strongly 

 with the antibody than those with the substituent group in a different 

 position, and the conclusion was reached from the values of the hapten 

 inhibition constant that the surface configuration of the combining 

 regions of the antibody molecules approximates that of the haptenic 

 group to within closer than 1 A. A similar conclusion has also been 

 reached by a study of the effect of electrical charge. The ratio of 

 inhibiting powers of two similar haptens, one containing a positively 

 charged group, the trimethylammonium ion group, and the other an 

 uncharged group with the same size and shape, the tertiary butyl 

 group, with antiserum made by injecting rabbits with sheep serum 

 with attached p-azobenzene-trimethylammonium ion groups could 

 be interpreted to show that the positive charge of the charged hap- 

 tenic group interacts with a negative charge in the antibody 7 A. 

 away. Inasmuch as the positive charge in the phenyltrimethylam- 

 monium ion may be considered to be at the center of the nitrogen 

 atom, and the radius of this ion (the distance from the center of the 

 nitrogen atom to the surface of the methyl groups) is 3.5 A., and inas- 

 much as the minimum distance of approach of a negative charge to 

 the surface of the antibody may be taken as the radius of an oxygen 

 atom, 1.4 A., the mininunn distance of approach of a positive charge 

 in the hapten and a negative charge in the antibody is calculated to be 

 4.9 A. The fact that the value calculated from the hapten-inhibition 

 data is only 2.1 A. greater than this again indicates that in general 

 there is a very great complementariness in structure and closeness of 

 fit of antibody and antigen. 



