Charles Todd 127 



When one of the sera wa« tested on a series of corpuscles of 

 different normal cattle, it was found that the haemolytic power 

 of the serum wtis (juantitatively very different for the corpuscles of 

 different individuals; being very highly haemolytic for some, less so 

 for others, and again comparatively slightly so for others. In a similar 

 test made on the corpuscles of the same individuals but with a second 

 serum the same variations were observed but they did not coincide 

 with those obtained with the first serum, that is to say corpuscles 

 which were highly haemolysed by the first serum were often com- 

 imratively slightly affected by the second and vice versa. When the 

 tests were extended to other sera similar results were obtained — in 

 short no two sera appeared to be absolutely alike. 



It was next found that if one of these isolytic sera is exhausted 

 with the corpuscles of a particular individual ox (/I), it remains 

 haemolytic for the corpuscles of many other individuals, but loses its 

 haemolytic power for the corpuscles of some other individuals as well 

 as for those oi{A). 



If now a second isolytic serum is exhausted with the corpuscles 

 of the same individual (A) and then tested, its action on the various 

 corpuscles is not exactly parallel to that of the first serum and often 

 shows marked differences. This result is to be expected as it was 

 shown by Ehrlich and Morgenroth that two goats each injected with 

 similar doses of the same goat's blood at the same time gave quite 

 diff'erent isolysins. 



As a matter of fact the isolysins formed depend upon two distinct 

 factors : 



(a) The individuality of the injected corpuscles. 



(6) The individuality of the animal into which they are injected. 



When we consider the enormous number of variations possible in 

 each of these factors, we see the almost unlimited possibilities in the 

 resulting sera. 



In view of the above it should be possible by taking a mixture 

 of a sufficiently large number of immune sera and exhausting this 

 with the corpuscles of one individual, to obtain a serum which has no 

 haemolytic action on these corpuscles but haemolyses those of all other 

 individuals of the same species and might therefore be made use of for 

 the identification of this particular individual. 



To test this a mixture was made of between 60 and 70 of these 

 isolytic sera. This mixture was then exhausted with the corpuscles 



9—2 



