408 STUDIES IN IMMUNITY. 



serum, whether normal or immune, may be endowed with sensitizing 

 power for certain corpuscles and consequently may, in their presence, 

 produce alexin fixation, although as serum it tends simply to inhibit 

 such a fixation. These two forces in the same serum oppose one 

 another and the antagonistic property will be the more evident 

 the weaker the sensitizing property. Under such conditions it 

 might well happen that there would be no hemolysis, in other 

 words, it might appear as if no sensitizer were present. So, for 

 example, it is generally considered that normal rabbit serum (56 

 degrees) does not sensitize bovine corpuscles; in a mixture of 

 bovine corpuscles (0.05 c.c.), heated rabbit serum (0.5 c.c.), and a 

 little guinea-pig alexin (0.1 c.c) there is indeed no hemolysis. But 

 if the corpuscles are first subjected to rabbit serum, then centri- 

 fugalized, the supernatant fluid decanted, and an equal amount 

 of normal salt solution added, followed by alexin,* we find that 

 slow but complete hemolysis takes place. Normal rabbit serum, 

 then, is sensitizing for bovine corpuscles, although weakly so. 



Conversely, when corpuscles that are strongly sensitized by this 

 serum are chosen, as Sachs has done,t there is no danger of losing 

 sight of the antagonistic property. It is found, for example, that 

 normal rabbit serum sensitizes goat corpuscles strongly. We mix 

 0.6 c.c. of rabbit serum (56 degrees) with 0.2 c.c. of salt solution 

 containing 25 per cent of washed goat blood; half an hour 

 later 0.05 c.c. of fresh guinea-pig serum is added. Hemolysis is 

 nearly complete in one and a half hours at 37 degrees, whereas 

 in a control with salt solution replacing the rabbit serum there 

 is no hemolysis. 



In this case the antagonistic property, although present, is con- 

 cealed. Although it fails to prevent hemolysis, it does retard it 

 notably. If we make a mixture of goat blood and rabbit serum like 

 the preceding and, after contact, centrifugalize, remove the serum, 

 and replace it with salt solution, we find that the addition of alexin 

 will produce a much more rapid hemolysis; it is complete, indeed, 

 in from 15 to 20 minutes. By removing the serum we have eliminated 

 the antagonistic property and hemolysis is accelerated. As is evi- 



* A control is made at the same time with corpuscles that have not been treated 

 with rabbit serum to which salt solution and alexin are added. 



t Analogous results have been obtained in bacteriolysis by Pfeiffer and 

 Fried berger. 



