STUDIES ON H.EMOLYSINS. 107 



hand, will be completely inhibited in its action. For this contains 

 one complete solvent dose of immune body a which is neutralized 

 by the anti-immune body, and only one-tenth of a solvent dose of b 

 which, although not affected by the anti-immune body, is so slight 

 as not to cause any appreciable solution. Only when larger amounts 

 of immune body are employed in which b becomes active does solution 

 occur, and this becomes complete only when that quantity is reached 

 which contains 10a+16. Naturally, if the ratio is 1:20, a quantity 

 will be required which can be represented by the formula 2Qa + lb. 



These explanations will perhaps suffice to make the above-men- 

 tioned peculiarities in the action of the anti-immune body com- 

 prehensible. They will perhaps also make clear that between the 

 dose of immune body whose action is completely inhibited by the anti- 

 immune-body serum and the dose which causes complete solution a large 

 number of intermediary stages exist in which the degree of solution 

 gradually increases. 



In reality the circumstances are much more complicated than 

 this ; for with the increase in the dose of immune body a large number 

 of new immune bodies, similarly superposed, come into action 

 immune bodies which find few or no corresponding anti-immune 

 bodies in the antiserum. 



This brings us to another important question: 7s it possible by 

 means of the anti-immune body to demonstrate the difference of the 

 immune bodies produced by injections of ox blood into different species? 



Our first experiments were undertaken with the serum of goats 

 which had been immunized with ox blood. As will be seen by the 

 following figures, our anti-immune body (derived by injections of 

 an immune body obtained from rabbits) in this case exerts no action. 

 The varying amounts of immune body mentioned are mixed with. 

 0.4'cc. anti-immune body and then with 1 cc. 5% ox blood suspension 

 and 0.5 cc. normal active goat serum to activate the mixture. In 

 the control test 0.4 cc. inactive normal goat serum are used instead 

 of the anti-immune body. (See Table VIII.) 



That this serum differed markedly with respect to its content of 

 individual immune bodies was already shoAvn by the fact that, in 

 contrast to the serum of immunized rabbits, it did not possess a 

 hsemolysin acting on all goat blood-cells in general, since such a hfemoly- 

 sin would have exerted a most injurious action in the form of an auto- 

 lysin. As a matter of fact, the law already mentioned under the name 

 '" horror autotoxicus " applied here also, and hence merely an isolysin 



