342 IMMUNOLOGICAL STUDIES 



produced reacts not only with the original antigen but also with 

 other proteins, such as those of chicken serum, that have been 

 coupled with arsanilic acid. The antiserum reacts also with ordi- 

 nary horse serum, with arsanilic acid alone (no visible precipi- 

 tate ) and with "haptens" structurally similar to the arsanilic acid, 

 but not with chicken serum; nor does an antiserum against horse 

 serum react with chicken serum. From the many experiments of 

 this type with which the immunological literature abounds it is 

 now clear that relatively small parts of a large complex molecule 

 can act as determinants of the specificity of antigens and anti- 

 bodies. 



Immunochemical research has not, as yet, provided very much 

 information as to the size, number, and diversity of structure of 

 the determinant groups of various native proteins and other large 

 molecular substances of interest to biologists. Landsteiner (1942) 

 showed that products of the hydrolysis of silk protein with 

 molecular weight of about 600 possessed such activity. Another 

 example of investigations in this direction that may be mentioned 

 is the current work of Kabat and Leskowitz ( 1955; cf. Watkins 

 and Morgan, 1955) showing that certain simple sugars possessed 

 the ability to react specifically with antibodies against human 

 blood group substance. They suggest that the determinant struc- 

 tures of the blood group mucopolysaccharidic antigens are prob- 

 ably no larger than tri- to penta- or hexasaccharides. 



One of the aims of embryologists in applying immunological 

 methods is the determination of the time of appearance of specific 

 constituents of the adult organs. However, this approach is beset 

 with various difficulties, which have not apparently been appre- 

 ciated by all the workers in the field. The prime difficulty is in 

 knowing what is being detected by the antiserum that is used 

 as the test reagent. As indicated above antibodies can detect cer- 

 tain specific structures of a large molecule. A particular protein 

 may possess a number of such structures and they may be of 

 several sorts, each of which may induce the formation of distinct 

 antibodies in the serum, as is illustrated in the recent work of 

 Francis et al. (1955) with "multi-haptenic" antigens. If some of 

 the determinant groups are common to different proteins of an 



