HJEMOLYS1NS AND ALLIED BODIES 449 



its nature to the variable structure of the cytophilic and com- 

 plemento-philic groups. To demonstrate this for the cytophilic 

 group, let us take an actual experiment : the serum of a rabbit, 

 inoculated with ox's erythrocytes, can lake both ox's and goat's 

 erythrocytes. The erythrocytes of these two animals can also be 

 laked by the serum of another rabbit previously treated with in- 

 jections of goat's erythrocytes. If, then, the cytophilic group of 

 the amboceptors be a simple body, it must fit the receptors both 

 of ox's and goat's erythrocytes equally well, and by mixing the 

 immune blood with either kind of erythrocyte it should be 

 possible to remove the amboceptors to an equal extent. Experi- 

 ment proves this is not the case. 1 For if rabbit's serum, immu- 

 nised against ox's erythrocytes, be mixed with ox's erythrocytes 

 its hsemolytic power, both for the ox's and the goat's erythro- 

 cytes, will be greatly reduced ; while, if it be treated with goat's 

 erythrocytes, the original haemolytic power for ox's erythrocytes 

 will remain almost as before, but there will be a very marked fall in 

 hsemolytic power towards goat's erythrocytes. The amboceptors 

 which react with these two kinds of erythrocytes cannot, there- 

 fore, be exactly alike, else, in the latter part of the experiment, 

 would the fall in haemolytic power of the serum affect ox's and 

 goat's erythrocytes to the same extent. An exactly similar result 

 follows if rabbit's serum, immunised against goat's erythrocytes, 

 be used ; goat's erythrocytes binding the amboceptors for both 

 kinds of erythrocytes, but ox's erythrocytes binding only those 

 which take part in the haemolysis of ox's blood. The serum of 

 both these rabbits must therefore contain a certain amount of 

 amboceptors common to both kinds of erythrocytes, and, besides 

 this, each kind of serum must contain an amboceptor reacting only 

 with the erythrocytes used for injection. In other words, the 

 amboceptor cannot be a simple body, but must be compounded out 

 of a number of partial amboceptors, each of which has slightly 

 different cytophilic groups. 



The production of these compound amboceptors is easily ex- 

 plained on the side-chain theory. An erythrocyte contains a number 

 of different receptors : some of these, when the erythrocyte is 

 injected into an animal, will anchor on to suitable cell receptors. 

 There will accordingly be a liberation into the blood of various 



1 i.e. rendered hasmolytic towards. 



2 F 



