CONCERNING ALEXIN ACTION. 



185 



hand, and amboceptor and complement, on the other, may vary 

 from case to case. Whereas in Case II the decanted fluids were 



TABLE II. 

 ABSORPTION OF THE RABBIT SERUM BY SHEEP BLOOD AT C. 



absolutely inactive against native blood (i.e., all the amboceptor 

 had been bound at C. by the blood-cells) in Case I the decanted 

 fluids were then still active when the amounts of serum added were 

 less than the solvent dose. This indicates that in this case the affinity 

 of the amboceptor's cytophile group for the receptor of the cell is 

 relatively slight at C. In like manner the columns B of the tables 

 show a certain difference of affinity between amboceptor and com- 

 plement. In Case I the decanted fluid still contains the entire com- 

 plement; in Case II, on the other hand, a portion of the complement 

 must have combined with the amboceptor, for the decanted fluid 

 shows a distinct loss of complement. The separate examination of 

 the sediments of the specimens to which active serum was added 

 agrees with this; in Case I these sediments mixed with physiological 

 salt solution and placed into the incubator showed no trace of solution, 

 while in Case II the sediments of the first three tubes showed mod- 

 erate, little, and trace of solution respectively. 



Both normal hcemolysins (Buchner's negative Cases I and II) therefore 

 correspond in their main behavior. They consist of two components 

 (readily separable by the "cold method") which in their mutual relations 

 manifest a certain variation in the behavior of their receptors. 



The conditions in these two combinations were favorable for 

 analysis of the mode of action by means of our method. In the 

 study of Buchner's third negative case, however (guinea-pig blood 

 and dog serum), difficulties presented themselves which at first ap- 



