156 BACTERIAL POISONS. 



germicidal property of the substance is demonstrated. 

 HANKIN closes this contribution with the following con- 

 clusions : 



(1) HALLIBURTON'S cell-globulin /? has marked germi- 

 cidal properties. 



(2) In this respect it differs from fibrin ferment. 



(3) The germicidal property of this substance seems to 

 be identical with that of serum as described by BUCHNER, 

 NISSEN, and NUTTAI.L. 



(4) The active properties of the serum are probably due 

 to this or to an allied body. 



In a more recent contribution HANKIN designates the 

 germicidal agents of the body as "defensive proteids." 

 He thinks it probable that blood-serum owes its activity 

 to these bodies and that the assumption of an "active con- 

 dition of the serum albuminate" made by BUCHNER is 

 unnecessary. He also thinks that BEHRING'S supposed 

 alkaline base exists in the form of an albumose. We know 

 of three albumoses which are alkaline in reaction. These 

 are the protomyosinose and deuteromyosinose of KUHNE 

 and CHITTENDEN, prepared by the digestion of myosin, 

 and the anthrax albumose of MARTIN. 



By a method similar to that which he had employed in 

 the preceding experiments, HANKIN has isolated a "defen- 

 sive proteid " from the blood-serum and the spleen of the 

 rat. This substance belongs to the globulins, and the nat- 

 ural immunity of the rat against anthrax is probably due 

 to its existence in the blood. 



STERN finds that the blood taken from different men, or 

 from the same man at different times, varies markedly in 

 its germicidal properties ; also, that the germicidal proper- 

 ties of the blood when kept at 42 are at least as great as 

 at the normal temperature of the body. These statements 

 are substantially confirmed by ROVIGHI. 



