ELIMINATION OF SPURIOUS TESTS FOR B. COLI 



341 



agar, as against 12 out of the 21, And of the plates from the 

 gentian violet presumptive tests, 15 samples yielded B. coli 

 whereas only 12 of the samples tested on plain lactose broth 

 yielded B. coli. The discrepancy in favor of gentian violet in 

 the presumptive test is still further shown in the greater pro- 

 portion of cultures secured in the first test, 73 per cent from the 

 dye series, as against 67 per cent from the non-dye presumptive. 

 Strangely enough, the odd sample failing to show a positive 

 presumptive test in the presence of gentian violet yielded B. coli 

 from the non-dye tube. Looking ahead to the data formulated 

 in table 3 in which gentian violet 1-20.000 was used we find that 

 this particular sample yielded B. coli from the test, with, as well 



TABLE 3 



Tests for B. coli and sporulating gas-forming anaerobes in 20 samples of water- 

 using 1-20,000 gentian violet in the ■presumptive test 



as without, the dye. We cannot explain this anomaly. Our 

 failure to isolate B. coli from five unheated samples showing 

 gas in the presumptive test led us to increase the concentration 

 of dye to 1-20,000. The unused residues of 20 of the first 21 

 samples were therefore re-examined with this change in technic 

 along with new -samples secured. Table 3 summarizes the 

 results of this second series of tests with these 20 samples. Part 

 of the differences in findings are no doubt due to the greater age 

 of the samples when the last tests were made. 



This tune as in the first instance the selective inhibition of 

 sporulating gas formers is shown in the case of the heated sample 

 tested in lactose broth containing gentian violet ; not one showed 



THE JOURNAL OF BACTERIOLOGY, VOL. Ill, NO. 4 



