Gram Reaction and Gentian Violet Reaction. 17 



has so far been unsatisfactory. Fractional crystallization leaves 

 the active substance in a syrupy filtrate. Precipitation with 

 AgS0 4 + Ba(OH) 2 , of histidine and a syrupy material giving 

 the histidine diazo test, leaves most of the active material in the 

 nitrate. The addition of an excess of baryta causes some, but 

 not all, to be thrown down together with impurities as the silver 

 compound . Phosphotungstic acid apparently destroys the activity 

 of the compound, although this requires further verification. 

 It has not been possible to obtain active material either from the 

 phosphotungstic precipitate or filtrate. The phosphotungstic 

 acid has been removed both by baryta and by extracting with 

 amyl alcohol and ether. 



In attempting other methods, also, the activity has gradually 

 diminished and been entirely lost, and it may prove impossible 

 to obtain the material in pure form by methods at present avail- 

 able. It is hoped, however, that further work with larger quan- 

 tities of material will result in the separation of this compound, 

 which may prove to be of more general interest than simply from 

 the standpoint of bacterial nutrition. 



To sum up : the experiments here reported indicate that casein 

 and certain other proteins contain a hitherto undescribed com- 

 ponent, which also occurs in an infusion of beef and beef heart. 

 It is essential to the growth of the hemolytic streptococcus and 

 probably the pneumococcus, and is absorbed from the beef 

 infusion by charcoal, and precipitated from the casein in an impure 

 form by mercuric sulphate. The chemical nature of the substance 

 has not yet been determined. 



7 (1589) 



The cause of the parallelism between the gram reaction and the 

 gentian violet reaction. 



By John W. Churchman. 



[From the Laboratory of Bacteriology, Cornell Medical School, New 



York City.] 



In previous studies, published at intervals since 1912, it has 

 been shown that a striking parallelism exists between the Gram 



