A GENERAL RESUME OF IMMUNITY. 501 



tination depends. In short, the essential phenomenon with agglu- 

 tinins, as with other active substances in sera, is its union with the 

 antigen; as far as the agglutination itself, which follows this union, 

 is concerned, it is only a secondary phenomenon on which we can- 

 not depend in considering agglutinins as functionally different 

 in molecular structure from the other antibodies. Such remarks on 

 agglutinins also apply to precipitins or the sensitizers which Ehrlich 

 endows with a complementophilic group. It is particularly be- 

 cause this investigator's theory suggested an artificial classification 

 of the antibodies, that it appears to me to have exerted a harmful 

 influence on scientific progress. 



Everyone agrees, naturally, that the numerous antibodies which 

 the study of immunity has brought to our knowledge and which 

 are active on such different elements as bacteria, cells, toxins, and 

 the like, should not be considered as identical, inasmuch as they may 

 be distinguished as regards specificity, or, in other words, since they 

 unite with different antigens. But in addition to this incontestable 

 difference of specificity, Ehrlich has imagined another one which is 

 more far reaching and which deals with the molecular structure 

 of the antibody. Indeed, he classes these antibodies in ac- 

 cordance with their molecular structure into three genera; anti- 

 toxins with a single combining group; agglutinins and precipitins 

 with a single combining group but with an additional functional 

 group which brings about agglutination or precipitation ; and finally 

 sensitizers which have two combining groups in their molecule, 

 uniting on the one hand with the cell that is affected and on the 

 other with the alexin (complement), and hence the name of ambo- 

 ceptor. 



In every instance according to this classification the phenomena 

 observed are attributed to special properties in the antibody and 

 never to those in the antigen. As a matter of fact, these phenomena 

 should be related, not as regards antigen or antibody considered 

 separately, but as regards the complexes which result from their 

 union, and it is evident that the special properties of the antigen 

 must affect markedly and perhaps to a preponderating degree, the 

 qualities of such complexes. I have just mentioned this idea as 

 applied to agglutination, and I have frequently defended it in 

 respect to sensitizers. The sensitizer, on uniting with the cell 



