ILEMOLYSINS AND ALLIED BODIES 447 



presence of amboceptor and complement. The amboceptor was 

 found confined to the sediment (that is, bound to the erythro- 

 cytes) ; the complement was confined to the supernatant fluid. 

 This experiment shows us, so far, that the amboceptor has a 

 very marked avidity for. the erythrocytes, combining with them 

 even at C. This is then evidently the first stage in the process 

 of haemolysis, combination of amboceptors and erythrocytes. 

 (6) To study the relation of the complements to this compound, 

 it was necessary to work at a temperature at which haemolysis 

 might result, and to stop the process short in its middle, that is, 

 when the amboceptors had united with the erythrocytes, but some 

 of the complement was still free. Now, Ehrlich and Morgenroth 

 found that haemolysis was completed in fifteen minutes, at 37 C. 

 If they stopped the process after ten minutes and then quickly 

 centrifugalised the mixture, the resulting sediment, suspended in 

 saline, was found to contain all the amboceptor and sufficient 

 complement to produce moderate haemolysis, this being readily 

 completed by adding some normal serum. The supernatant fluid, 

 on the other hand, was found to contain complement but no 

 amboceptor. 



From these results Ehrlich and Morgenroth argued that the 

 amboceptors must possess two binding groups, one of which, the 

 cytophilic group, has very marked combining affinities for erythro- 

 cytes the combination occurring even at C. (Expt. 3a) ; and 

 the other, the complemento-philic group, only weak affinities for 

 the complement no union resulting at C. and only partial 

 union after ten minutes at 37 C. (Expt. 36) the latter union 

 only occurring when the cytophilic group had been fitted to an 

 erythrocyte. 



It should be pointed out here that these experiments, although 

 they can be best explained by Ehrlich's theory, and indeed were 

 devised to prove this, do not disprove Bordet's. When we 

 enter into the subject more fully, however, we shall encounter 

 results (the production of anti-bodies, &c.) which can be explained 

 best if Ehrlich's theory be accepted. 



Perhaps the strongest evidence in favour of the side-chain 

 theory is that furnished by Preston Kyes ( 6 ) and Hans Sachs in 

 their studies on snake venom. As we shall see later, cobra venom, 

 regarded as a haemolytic agent, contains amboceptors alone. If 

 it be mixed with the thoroughly washed erythrocytes of the 



