A GENERAL RESUME OF IMMUNITY. 499 



the progress of knowledge, and has really hindered the free de- 

 velopment of investigation. In offering explanations which seem 

 definitive, and schemata which satisfy the experimenter and 

 appease his curiosity, Ehrlich's theory has come to make certain 

 problems, which have scarcely been touched upon, regarded as 

 worked out. It must be admitted that its partisans have seemed 

 chiefly preoccupied with justifying it, and it would seem as if their 

 efforts were directed rather toward defending the theory than con- 

 trolling it. And the guiding thought, to my thinking fatal, which 

 they have endeavored to enforce, is the constant attribution to the 

 molecule of the antibody of separate atom groups for each of the 

 phenomena to which the antibody gives rise. To each manifesta- 

 tion that has been observed there must be a corresponding molec- 

 ular group or even a new molecule. Such a method is favorable to 

 the theory, inasmuch as each influence as noted finds material 

 evidence in a special group which becomes a sort of sub-stratum for 

 it. Its use, however, is doubtful for the same reason, for it gives too 

 easily the appearance of an explanation, when in reality it estab- 

 lishes no definite relation between the phenomena that have been 

 observed, since it satisfies itself in symbolizing these manifestations 

 by independent groups, each of which appears as a resting place for 

 such and such a property. These groups are simply evoked by the 

 theorist as he wishes and their very existence in each and every case 

 is far from sure. When we say that an agglutinin which unites with 

 bacteria and clumps them produces these results because it possesses 

 two groups in its molecule, one of which combines and the other of 

 which agglutinates, it facilitates explanation, and seems, indeed, 

 and in this its clanger lies, to resolve the question entirely, but it 

 is probably inexact. In the first place, it is not, strictly speaking, 

 the agglutinin which agglutinates, but rather, as I showed in 1S99, 

 the salt. There are antigens which, when united with their anti- 

 bodies, form complexes that have the characteristic of being 

 flocculable by electrolytes.* It is the complex which agglutinates 

 and there is no reason to localize the cause of agglutination in a 



* In this respect there are analogies between bacteria laden with agglutinin 

 and colloidal complexes, not only in respect to the action of salts but also to the 

 influence of electricity. According to Neisser and Friedemann, the complex mas- 

 tic-gelatin is flocculable by the electric current just as bacteria that have fixed 

 agglutinin are. 



