70 



ojjposite results. The violet reaction is in striking contrast to this "notori- 

 ously uncertain" staining test and though not assuming to be a i)arallel or 

 substitute test, it is a valuable differential reaction. 



Work on various other staining agents has shown many to exhibit a defi- 

 nite selective inhibitive action. The Conradi-Drigalski medium for isolation 

 of B. typhosus from water, stools, etc., has as its basis the restraining action 

 of the crystal violet towards various cocci and bacilli, without influencing at 

 all the growth of the typhoid-colon group. Krumwiede' and Pratt'' and 

 Churchman'' have made observations on the growth of bacteria on media 

 containing various closely related aniline dyes and have found their action 

 to correspond closely to that of the gentian violet. Smith^ has shown the 

 violet test to be specific for certain of the phytopathological bacteria. 



Aside from the significance of this test as a classificatory feature of great 

 value it might be expected to have some practical application in sanitary 

 bacteriological analysis, as most of the intestinal bacteria that we presume 

 indicate pollution by sewage are Gram negative and, therefore, with few 

 exceptions, are violet negative. Many of the connnon saprophytic bacteria 

 found normally in water and in milk are Oram positive and so would in the 

 majority of cases fail to grow in the presence of the stain. Churchman in 

 his work on the collection at the American Museum of Natural History found 

 the following organisms to be Gram negative and with two exceptions also 

 violet negative: 



3 strains of B. coli communis. 



5 strains of B. culi cottrnii/nior. 



5 strains of B. paracoli. 



2 strains of B. coli varietas. 

 14 strains of B. typhosus. 



18 strains of B. parntyphosus. 

 5 strains of ]'ibrio chuh-rit. 



3 strains of B. dysenterice. 

 5 strains of B. enteritidis. 

 3 strains of B. cloaca. 



Curiously enough B. vH'lchii and />'. .s/w/of/cwc.s, both (!rani jiositive, proved 



t () be \'i()let n('gati\'e. Sulililis, nn/coidcs, inccjathcriinti . liodi ritnis, ni'srnlcricus 

 and many of the sapi'i>i)hyt ic cocci are NJolet posit ixc. 



»Ztsehr. of Hyg., Vol. XXXIX, p. 283. 

 'Proc. N. Y. Path. Soc, Vol. XIII, p. 43. 

 'Jour. Exp. Med., ^'ol. XVII, No. 4, p. 373. 

 « PhytopatholoKy, Vol. II, No. 5, p. 213. 



