358 IVAN C. HALL AND LILLIAN J. ELLEFSON 



showed acid colonies, B. coli could be isolated but twice. The acid 

 colonies in each of these cases proved to be gram positive cocci. 

 In 3 samples, plates of which were acid after streaking from 

 positive presumptive tests in plain lactose broth, gram positive 

 cocci were responsible in each case for the acid; gas forming 

 aerobes could not be isolated, but from one came a gram negative 

 coliform bacillus giving acid on litmus lactose agar and not 

 liquefying gelatin. This and one other sample of the three 

 yielded B. coli from the presumptive test in gentian violet lactose 

 broth plated on plain litmus lactose agar. From 18 samples 

 tested in lactose broth with 1-20,000 gentian violet and streaked 

 out on gentian violet litmus lactose agar B. coli was isolated 

 in each case; from 17 samples tested in plain lactose broth and 

 streaked out on gentian violet litmus lactose agar B. coli could 

 not be isolated in two cases. In one of them only a large Gram 

 positive bacillus, acidifying the plate and apparently not com- 

 pletely inhibited by the gentian violet at a concentration of 

 1-100,000, was isolated; it did not produce gas from lactose and 

 the subculture from the first presumptive test into a second 

 showed that we had lost the gas former; B. coli was isolated 

 from the presumptive test of this sample in lactose broth with 

 1-20,000 gentian violet. In the other only streptococci appeared 

 on the plate, which was acidified by them; they too were not 

 inhibited by the weaker concentration of gentian violet in the 

 plate and the gas former was lost when the second presmnptive 

 test was made; B. coli was isolated from the presumptive test on 

 gentian violet lactose broth in this case also. 



The single heated sample showing gas in the presumptive 

 test yielded B. coli from the lactose broth with, as well as with- 

 out, gentian violet. The organism had apparently been able to 

 survive the heating. We cannot of course be certain that 

 sporulating anaerobes were absent from this specimen any 

 more than we can be sure that they were absent from the un- 

 heated specimens yielding gas in the presumptive test. 



From a comparison of our work on water with that on milk 

 we are inclined to conclude that the so-called "attenuated" B. 

 coli is less frequent in milk than in water; a larger proportion 



