GENERAL PROPERTIES OF COMPLEMENTS 215 



bining relationships. Even the complements of different 

 animals may combine with the same molecules, the com- 

 plement of one animal keeping out of combination the 

 complement of another, and vice versa. It was considered 

 for a time that the complement of each species had a rela- 

 tively specific group, as the supposed anti-complement had 

 always the maximum effect on the complement used in its 

 production, and only slight effects on the complements of 

 allied species. As has been shown above, however, this 

 anti-complement action may be due simply to the deviation 

 by the serum receptors + the anti-substances, and the 

 existence of true anti-complements is somewhat doubtful. 

 The result of the work on this subject, in fact, has been to 

 emphasize the community of the combining affinities of 

 complements of different animals. Whilst therefore we 

 consider that the view as to the uniformity of alexine 

 (complement) molecules of a given species of animal 

 is not tenable, we think that Ehrlich has pushed 

 too far the conception of special combining affinities of 

 complement. Differences in degree of combining affinity 

 have been almost completely overlooked by writers on the 

 subject ; and when we use a powerful complement absorber, 

 say a serum precipitate produced by the action of a precipitin 

 on the homologous serum, the variety of complements from 

 different animals which may be absorbed is extraordinary. 



Complements of different animals, however, differ widely 

 in their zymotoxic groups ; that is, in their relative toxic 

 action. Thus it is that in order to bring about a particular 

 effect, e.g. complete haemolysis, a much greater amount of 

 the complement of one animal than of another has to be 

 brought into union, and hence a larger amount of immune- 

 body comes to be necessary ; and in some instances a rela- 

 tively large amount of complement may enter into union 

 with a corpuscle without producing any marked lysis. 

 This factor of relative toxicity is also seen when we use/ 

 the same complement and test it on the corpuscles of 



