192 COLLECTED STUDIES IX IMMUNITY. 



effect complete hcemolysis as complements in amounts in which alone 

 they are entirely inert) we could furnish positive proof of the presence 

 of amboceptors. 



Buchner and Gruber have therefore described a total of seven 

 cases said to show pure alexin action; and these cases were held 

 by them to be sufficient to decide in the negative the entire question 

 of the complex nature of the normal serum haemolysins. Against 

 this we have in all these cases brought positive proof that the " alexin," 

 conceived by Buchner to be a simple unit, always produces its effects 

 through the co-action of two components, the existence of which is 

 demonstrable in different ways. We must therefore uphold Ehrlich 

 and Morgenroth's view, that normal and artificially produced hcemo- 

 lysins exert their action according to exactly the same mechanism. 



We do not yet possess a method generally applicable to demon- 

 strate the complex nature of the hcemolysin, and even a thorough 

 analysis, therefore, need not necessarily achieve the desired result in 

 every case. The method adopted by Muller 1 for demonstrating 

 the amboceptors in chicken serum, which is hasmolytic for rabbit 

 blood, is of interest in this connection. When the usual methods 

 failed he found that bouillon injections caused an increase in the 

 amount of complement in the chicken serum without affecting the 

 amount of amboceptor. This led him to recognize the complex 

 nature of the hsemolysin, a fact confirmed by the successful activa- 

 tion of heated chicken serum by means of pigeon serum. When 

 therefore in isolated cases the separation does not succeed accord- 

 ing to the methods heretofore employed, such results, the product of 

 incomplete methods, most certainly do not argue for a simple alexin 

 action. We hope that the employment of the lowest possible tem- 

 peratures in inactivation will result in increasing " completion ' 

 possibilities and make the demonstration of the complex consti- 

 tution of the hremolysins easier in difficult cases. At present this 

 demonstration has failed only in the case of eel serum (which, to 

 be sure, is very peculiar in its hsemolytic behavior), for thus far no 

 fitting complements have been found for this serum. In all other 

 cases of hcemolysis through normal sera, which have been investi- 

 gated for this purpose, according to our experiences positive proof 

 of the presence of the amboceptors has been furnished. 



The normal bactericidal sera also owe their bactericidal power 



1 T Muller, 1. c. 



