ISOLATION OF THE COLON BACILLUS. 65 



tion of the colon bacilli originally present were lost during 

 the incubation of the large samples, so that waters of bad 

 quality actually appeared to better advantage by the use 

 of 100 c.c. with preliminary incubation in phenol broth." 



Similarly Whipple (Whipple, 1903) notes that 2.9 per 

 cent of some samples of water examined by him gave 

 positive tests with .1 c.c. but not with i c.c., while 4.3 per 

 cent gave positive tests with .1 c.c. or i c.c. and negative 

 tests with 10 c.c. Again, in another series of samples 

 examined, of those which gave positive samples in smaller 

 portions 5.3 per cent were negative in 10 c,c., 4.7 per cent 

 in loo c.c., and 7.7 per cent in 500 c.c. 



In all practical processes of examining water for B. coli 

 one essential step is the isolation of pure cultures upon 

 the litmus-lactose-agar plate, whether the plate be inocu- 

 lated from the water direct or from a preliminary enrich- 

 ment culture. In the first case a measured quantity of 

 water must be added. In the second case, since the 

 enrichment- tube was inoculated with a known amount of 

 water all further work is purely qualitative and it is only 

 necessary to obtain such a number of colonies upon the 

 lactose plate that the isolation of a pure culture shall be 

 easy. In practice the following procedure has been 

 found generally successful. After the dextrose-tubes have 

 been incubated for twelve to twenty-four hours at 37, 

 from those which show gas, one loopful is carried over 

 to a tube containing 10 c.c. of sterile water, and of this 

 water one loopful is taken for the inoculation of the plate. 

 Ordinarily this will give colonies which are sufficiently 



