570 COLLECTED STUDIES IN IMMUNITY. 



presence of other substances with different properties. In the case 

 described above there is an absence of antilytic action with a certain 

 excess of the antiserum. If we look at the subject from this stand- 

 point, we shall have to assume that the antiserum contains two sub- 

 stances, 1 one of which, of course, is the effective antiamboceptor. 

 The other substance would then be the cause of the inhibition of 

 the antiamboceptor action. Furthermore, since this inhibition is 

 only brought about by large quantities of the serum, this substance 

 would be present in the serum in much smaller amounts than the 

 former. The simplest explanation of the action of this substance 

 seems to be somewhat as follows: We must assume that this sub- 

 stance's point of attachment is a complementophilic auxiliary group 

 in the amboceptor. The occupation of this group so affects the 

 amboceptor molecule that the simultaneous presence of antiambo- 

 ceptor no longer prevents the combination with complement. Such 

 a behavior would be analogous to an observation published by Ehr- 

 lich and Marshall. 2 At that time, by means of a differentiating 

 method made available for one particular instance 3 by Marshall 

 and Morgenroth, it was shown that the amboceptor anchored to 

 the cell, although it could deprive native guinea-pig serum of all its 

 complement functions, was unable to absorb the non-dominant 

 complements if the dominant complement had first been neutralized 

 by the partial anticomplements of Marshall and Morgenroth. In 

 other words, an anchoring of the non-dominant complements was 

 only possible after the corresponding complementophile group 

 of the amboceptor had combined with the dominant complement. 

 In our case we would be dealing with an influence entirely similar 

 in principle, except that here the influence is reversed, i.e., the affinity 

 of the amboceptor to the antiamboceptor is reduced by the occupa- 

 tion of the auxiliary group. We believe that we can show directly 

 that the antiamboceptor is bound in either case, but that where .the 

 auxiliary group is occupied, the union of amboceptor and antiambo- 



1 We can of course assume a priori that an antiamboceptor serum directed 

 against the complementophile groups will possess a multiplicity of partial 

 antiamboceptors, for the amboceptors which take part in the immunization 

 possess a large number of different complementophile groups, and against 

 each of these a particular antibody is conceivable. 



2 Ehrlich and Marshall, 1. c. 



3 H. T. Marshall and J. Morgenroth, Uber Differenzierung von Comple- 

 menten durch ein Partialanticomplement. Centralblatt f. Bact. 1902, Vol. 31, 

 No. 12. 



