560 REUBEN L. KAHN 



weeks. A large number of red colonies were observed, but none 

 with a raetallic sheen. One can not, however, draw any definite 

 conclusions from this small number. 



RESUME 



It is suggested that: 



1. In the confirmatory test for B. coli, Endo slants be em- 

 ployed instead of Endo plates. 



2. The inoculations from the lactose fermentation tubes be 

 made in the butt as well as on the slant of the Endo tubes, to 

 permit aerobic and anaerobic growth. 



3. In the case of unknown waters, inoculations be made from 

 the fermentation tubes into Endo tubes after about fifteen 

 hours incubation, even if there be no gas present in the fermen- 

 tation tubes at that time. 



The advantages of the proposed modifications are: 



1. The production of more typical colonies of B. coli on the 

 Endo slant than on the Endo plate. 



2. The detection of aerobic and anaerobic growth in the same 

 test. 



3. Endo agar tubes can be kept for three to four weeks with- 

 out deterioration. 



4. The inoculation from the presumptive fermentation tubes 

 to Endo agar tubes before the production of gas in the former 

 will often save several days time in establishing the presence of 

 B. coli in an unknown water. 



REFERENCES 



Committee on Standard Methods of Water Analysis 1917 American Public 



Health Association, Boston. 

 Hasseltine 1917 Public Health Reports, 32, 1879. 

 Kahn 1918 J. Bact., 3, 547. 



Robinson and Rettger 1916 J. Med. Research, 29, 363. 

 Russell 1912 J. Med. Research, 20, 217. 



