69 



The Selective Action of Gentian Violet in Bacte- 

 riological Analysis. 



C. M. HiLLlARD. 



In 1912 Churchman^ reported a new differential test for the Schizomycetes , 

 depending upon the selective bactericidal action of gentian violet. The action 

 of this stain in high dilution upon various organisms planted in media con- 

 taining the dye was found to correspond closely to the Gram staining reaction, 

 the forms inhibited — "violet positive" — being in the majority of cases forms 

 that retain the stain; those growing — "violet negative" — usually being strains 

 that decolorize when treated by the Gram method. His actual results on 

 318 different strains of bacteria are shown in the following table: 



318 Strains. Gram positive. Violet positive. Violet negative. 

 182 165 (90%) 17 (10%) 



Gram negative. 



136 15 (11%) 121 (89%) 



The characteristic behavior of bacteria cultivated in the presence of the 

 dye in high dilution (1:100,000) is "so constant and clear cut that it must be 

 regarded as a fundamental biological characteristic." The Gram stain has 

 ever been an unsatisfactory test with certain groups of organisms, especially 

 the Coccacese. Differences in the age of the cultures, time of application of 

 the various reagents, and the temperature may influence the result. It is 

 sometimes extremely difficult to interpret the result of the stain as, some 

 individual cells will retain the stain and others in the same field or even in 

 the same chain as contiguous cells will have decolorized. As an instance of 

 discrepancy in interpretation of results we may cite Kligler's work^ who by 

 the same method (Churchman's) recorded 13 of 17 strains of saprophytic 

 cocci as certainly Gram negative, while four stained uncertainly, as opposed 

 to Churchman's recording of all 17 strains as Gram positive. In my own work 

 on 240 strains of streptococci I found 21 to stain irregularly and occasionally 

 successive stains of the same culture at different times would give totally 



'Jour, of Exp. Med., Vol. XVI. No. 2, p. 221. 

 ^ Jour, of Exp. Med., Vol. XVII, No. 6, p. 653. 



