UNION OF IMMUNE-BODY 9 



to those just stated. The apparent non-specificity is thus 

 shown to be really due to the molecules of immune-body 

 being of more than one kind, corresponding to different 

 receptors (antigens) which give rise to them. All the results 

 obtained by absorption methods go to show that when an 

 anti-serum acts on heterologous corpuscles or bacteria, this 

 is due to their possessing some receptors similar to those in 

 the corpuscles or bacteria used for injection. The number 

 of antigens, as tested by their combining affinities, would 

 appear to be practically limitless, and each anti-substance 

 combines only with those similar to the one which has 

 been used in treating the animal. 



ON THE MODE OF UNION OF IMMUNE-BODY WITH 

 RED CORPUSCLES 



The union of immune-body with red corpuscles, as 

 Ehrlich and Morgenroth showed in their first communica- 

 tion, takes place at C. This can be readily shown by 

 placing red corpuscles in a mixture of complement and 

 immune-body at C. and, after a time, centrifugalizing 

 and pipetting off the fluid. If the separated corpuscles 

 are then washed in salt solution, and the usual suspension 

 is made, no lysis occurs when they are incubated at 37 C. ; 

 whereas lysis occurs at once on the addition of complement. 

 The presence of complement in the fluid which was pipetted 

 off can be of course shown by adding it to sensitized cor- 

 puscles, when lysis occurs. A method is thus given for 

 separating immune-body from complement when the two are 

 present in a mixture, and it can be similarly applied in some 

 cases for freeing a normal serum of a natural immune-body 

 when it is present (vide infra, p. 28). 



The amount of immune-body which may be taken up or 

 absorbed by red corpuscles is conveniently estimated in the 

 following manner. Varying multiples of the minimum hsemo- 



