ZINSSER AND MUELLER 



731 



These relations are most easily understood from a schematic table such as the 

 following one which we take from the work of Avery and Heidelberger. 



Type I 



KE^ 



Type II I P (- ( Sii ^ 



Type III 



A. Intact cell (PS) 



pH I Anti-S serum Type I specific 



3 Anti-S serum Type II specific 



Anti-S serum Type III specific 



Protein 



I 



B. Isolated constituents of cell (P), (S) 



^ Anti-P serum, not type-specific 



Carbohydrate 





No antibody response 



Fig. I. — Under A the three type figures illustrate how the individual type specificity depends 

 upon the conformation of the carbohydrate substance, the antibody produced in each case fitting 

 this part of the complete antigen. Under B the formation of antibody by the separated protein alone 

 is indicated. (Scheme taken from Avery and Heidelberger, /. Exper. Med., 42, 372. 1925.) 



In order to separate the nucleoprotein and soluble specific substances from each 

 other, it is not necessary to disrupt the cell by chemical extraction and solution, but 

 in organisms like the pneumococcus, in which bile autolysis can be so easily produced, 

 the mere solution of the cell separates the two constituents. This indicates an ex- 

 traordinarily loose combination between the nucleoprotein nucleus and the carbohy- 

 drate associated with it. That the antibodies produced by the separated nucleopro- 

 teins are far less strictly specific than is the complete antigen or the carbohydrate 

 soluble specific substance has been demonstrated by Avery and Heidelberger, our- 

 selves, and by Lancefield for the streptococci. The nucleoproteins seem to possess a 

 more general species specificity, overlapping for types of the same organism. 



In any injection or immunization in which suspensions of whole bacteria are used 

 a certain amount of solution of the bacterial cells takes place either in vitro or after 

 injection. In consequence, therefore, both types of antibodies will be produced, as is 

 apparent in our table. 



In regard to what we consider the special importance of the antigenic properties 

 of autolysates, such as those produced with pneumococci, the point of greatest sig- 

 nificance is the fact that they serve to sensitize with a potency far above that of any 

 artificially produced bacterial fractions and are still more noticeably potent when 



