STUDIES ON H.EMOLYSINS. 113 



to an immune body acting on rabbit blood, passed through with 

 the greatest difficulty; the other, fitting an immune body acting 

 on guinea-pig blood, passed through in part completely isolated. 

 We were further able to show that heating the serum of a buck 

 treated with sheep blood caused all the complements excepting 

 one to disappear. The one which withstood the heat fitted the immune 

 body developed by the immunization. We were able to demon- 

 strate the same thermostabile complement in greater or smaller amounts 

 in the serum of normal goats and calves. To again call attention 

 to these experiments is not superfluous, for Gengou (Annal. I'lnst. 

 Pasteur, 1901) in spite of these proofs of the plurality of comple- 

 ments, still maintains that the serum of each species contains only 

 a single simple complement, " the alexin." 



It would be natural to conclude that there is a plurality of com- 

 plements from the manifold variations observed in the comple- 

 tion of various inactive sera by normal sera. The commonest, 

 example of this, probably known to every one having experience 

 in this field, consists in the fact that a certain immune serum can 

 be activated by two different sera serving as complement, whereas 

 other immune sera can be activated by only one of these sera. Never- 

 theless from our standpoint we cannot regard this method of proof 

 as at all conclusive because it rests on the assumption that for a 

 certain species of blood a serum contains only a single interbody 

 (or immune body). In our fourth communication we have already 

 shown that this assumption does not hold, even for the interbodies 

 of normal sera. 



The assumption of a plurality of complements in normal sera is 

 supported by the fact that by injections of a normal serum (which, accord- 

 ing to our view, possesses various active substances which may be 

 present as complements, or, at times, in the form of complementoids) 

 antisera are formed which act against the complements of various other 

 sera. In a number of different animals by injecting various sera we 

 have succeeded in obtaining anticomplements acting not only against 

 the serum originally employed, but also against certain comple- 

 ments of rabbits and guinea-pigs. According to Bordet's experi- 

 ments it is possible, by injecting a rabbit with guinea-pig serum, 

 to obtain an isolated anticomplement against a complement (able 

 to act in this particular case) present in guinea-pig serum. From 

 this it follows that in these sera, since they excite the production 

 of different anticomplements, at least two different complements 



