348 IVAN C. HALL AND LILLIAN J. ELLEFSON 



after three and four days respectively and failed to give aerobic 

 growth on lactose or plain agar with or without gentian violet; 

 they were found to contain apparently pure cultures of sporulat- 

 ing anaerobes. There has not been time to identify these organ- 

 isms with certainty, but finding them in the positive presump- 

 tive test with gentian violet indicates that for the best results 

 we shall have to increase the concentration of dye somewhat 

 over that with which we have hitherto experimented, although 

 even a concentration of 1-20,000 gives nearly perfect results, as 

 we have shown. A comparison of these cultures, numbers 

 210, 215 and 216 as to their resistance to gentian violet, in table 

 1, has shown that they are not unique in this respect. 



Among the unheated samples B. coli was isolated from both 

 the standard and dye presumptive tests eighteen times. In one 

 sample tested in dye broth another instance of supposed B. 

 cloacae turned out to be a mixture of B. coli and B. proteus. 



One sample yielded ''attenuated" B. coli with difficulty from 

 the dye test and not from the plain lactose broth test. It is 

 safe to say that the organisms could not have been isolated from 

 plain litmus lactose agar without the dye, since the culture in 

 the non-dye tube was heavily overgrown with a member of the 

 hay bacillus group, which did not appear on the plates, however. 



Two samples yielded B. coli from the standard test and not 

 from the dye test; the gentian violet litmus lactose agar plates 

 made from the last failed on two successive occasions to show 

 aerobic growth. Unfortunately the gas formers, probably dye 

 resistant anaerobes, (though these samples were not the same 

 as those from which we recovered the pure cultures of anaerobes 

 in the heated fractions) were lost before the next subculture was 

 successfully made. We were surprised to find that these last 

 samples were originally heavily contaminated, yielding B. coli 

 in 0.001 cc, according to the record of the Bureau of Sanitary 

 Engineering. The result seems to support the idea expressed by 

 Mr. Frank Bachman (1917) that it is such samples from which 

 B. coli disappear most readily on standing. 



One sample failed to yield B. coli in any of our tests; anae- 

 robes were present however, giving positive presumptive tests 



