412 W. W. WILLIAMS AND W. BURDICK 



duction without implying that the organisms are necessarily 

 killed. He finds that this gentian reaction is much more definite 

 and constant than the Gram stain. 



Next came the attempt of Petroff (1915) to devise a simple, 

 practical and rehable method for the isolation and cultivation 

 of the tubercle bacillus from the sputum and feces. He experi- 

 mented with gentian violet, methyl violet, methylene blue, 

 crystal violet, and fuchsin added to a meat-juice-glycerin -egg 

 medium. He found that gentian violet was the most favorable 

 stain on account of its inhibitory action on many organisms and 

 reported that he obtained sixty-nine positive cultures from sixty- 

 nine specimens of sputum from practically all stages of tuber- 

 culosis. Six of these specimens were negative by direct micro- 

 scopic examination. 



Shortly after the publication of Petroff's article, we began 

 the use of his method. We were soon impressed with the in- 

 hibitory effect of the sodium hydroxid and gentian violet which 

 he used, but not with the medium as a whole. It did not seem 

 to contain enough moisture to prevent rapid drying; it required 

 inspissation on three successive days, with the result that fre- 

 quently an uneven surface was obtained in spite of the utmost 

 care; also, it occasionally became contaminated with spore 

 bearing organisms which would usually cause liquefaction of 

 the medium with consequent destruction of hopeful cultures. 



It seemed to us that if a gentian violet medium could be made 

 which did not require inspissation, gave a smooth surface, con- 

 tained sufficient moisture, and could be sterilized in an auto- 

 clave, we should have a nearly ideal culture medium. 



While working with the medium devised by Besredka (1913) 

 for growing tubercle bacilli to make his tuberculin, which serves 

 as a good antigen in tuberculosis complement fixation tests, 

 it occurred to us to utilize it in an attempt to overcome the 

 defects of Petroff's medium. We believe we have accomplished 

 this by making a medium as here described. 



