THE MECHANISM OF THE ACTION OF AMBOCEPTORS. 219 



of analyzing hsemolysins. For, entirely aside from the fact that 

 under these circumstances the attempt to activate the centrifuged, 

 and presumably " sensitized/' blood-cells necessarily fails, it will 

 be seen that the occurrence of this complication considerably limits 

 the application of the second method employed to discover the com- 

 plex nature of hsemolysins, namely, separation in the cold, a method 

 already markedly restricted. This method, it will be recalled, depends 

 upon the fact that at C. usually only the amboceptors are bound 

 to the blood-cells, not the complements to the amboceptors. In 

 the case just described, however, the union of amboceptor and cell 

 depend on the combining of amboceptor and complement. How, 

 then, can a separation of the two components be effected if, on the 

 one hand, the conditio sine qua non for the union of amboceptor 

 and cell, a condition which obtains here, cannot be fulfilled at low 

 temperature, and if, on the other, it in itself precludes any sepa- 

 ration whatever? No winder, therefore, that Gruber 1 failed with 

 the cold separation method in just this case (guinea-pig blood + active 

 ox serum). 



The two atypical cases here described are, however, peculiarly 

 adapted to throw light on the mechanism of hsemolysin action. In 

 the first case the fact that blood-cells "sensitized' in the usual 

 manner withstand the action of the complement is hard to explain 

 in accordance with Bordet's view. But the behavior shown in the 

 second case becomes entirely inexplicable if, like Bordet, we believe 

 the action of hsemolysins to consist in this, that the amboceptors 

 (substance sensibilatrice) sensitize the blood-cells and so render 

 them vulnerable to the action of the complements (Bordet's alexins) 

 exerted directly upon them. For here we have demonstrated that 

 a sensitization does not take place; the amboceptor by itself is not 

 at all bound, and becomes effective only on the addition of comple- 

 ment. If, however, we were to assume that in our case the com- 

 plement nevertheless attacks the cell directly so that then the ambo- 

 ceptor can be found, w 7 e should arrive at a theory as unlike Bor- 

 det's as that held by Ehrlich and Morgenroth. But such a theory, 

 strange to say, would apply only to this and perhaps a few other 

 cases, that is, only to a few exceptions. Although superfluous, a 

 suitable experiment was also made in this case and, as might have 



1 Gruber, Zur Theorie der Antikorper. Munch, med. Wochenschr. 1901, 

 l\ T o. 49. See also H. Sachs, 1. c. 



