H^EMOLYTIC AND OTHER SERA 537 



for example, found that when a fresh serum is passed through a 

 Berkefeld filter, complement is largely retained in the pores of 

 the filter, whereas immune-body passes through practically 

 unchanged : and that if a mixture of complement and immune- 

 body be made and filtered at a temperature of 37 C., the 

 amount of immune-body which passes through is not diminished, 

 whereas it would be if it had united with the retained comple- 

 ment. Accordingly by this method there was obtained no 

 evidence of the direct union of immune-body and comple- 

 ment. Bordet holds that the immune-body acts merely 

 as a sensitising agent hence the term substance sensibili- 

 satrice and allows the ferment-like complement to act. It 

 is quite evident from his writings, however, that he does not 

 mean, as is often assumed, that the immune-body causes some 

 lesion in the corpuscle which allows the complement to act, but 

 simply that it produces in the molecules (receptors) of the red 

 corpuscles an avidity for complement. All that we can say 

 definitely at present is that the combination of receptor -f 

 immune-body takes up complement in firm union while neither 

 does so alone ; whether the immune-body acts as a link be- 

 tween the two or not must be left an open question. Even 

 after the corpuscles are laked with water the receptors are not 

 destroyed. Muir and Ferguson have shown that they can still 

 take up immune-body and, through its medium, complement, 

 just as the intact corpuscles do. Ehrlich and Morgenroth showed 

 that in some cases the red corpuscles can take up much more 

 immune-body than is necessary for their lysis, and Muir found 

 in one case studied, that each further dose of immune-body led 

 to the fixation of more complement, so that as many as ten 

 times the htemolytic dose of complement might thus be used up. 

 It is a matter of considerable importance that the union of 

 immune-body and red corpuscles can be shown to be a reversible 

 action. If, as was found by Morgenroth and Muir indepen- 

 dently, corpuscles treated with several doses of immune-body 

 and then repeatedly washed in salt solution be mixed with 

 untreated corpuscles and allowed to remain for an hour, then 

 sufficient immune-body will pass from the former to the latter, 

 so that all become lysed on the addition of sufficient complement. 

 The combination of complement, on the other hand, is usually 

 of very firm nature. It has been a disputed point whether there 

 are several distinct complements in a normal serum with 

 different relations to different immune-bodies, for which Ehrlich 

 and his co-workers have brought forward a certain amount of 

 evidence, or whether, as Bordet holds, there is a single comple- 



