THE SIGNIFICANCE OF B. CO LI IN WATER. 85 



from other sources need not be condemned unless the 

 organism was found in 20 c.c. or less. When colon bacilli 

 were fotind in only greater quantities than 100 c.c. the 

 water might be considered as probably safe. Horrocks 

 (Horrocks, 1901), after a general review of English prac- 

 tice, concluded that "when a water-supply has been 

 recently polluted with sewage, even in a dilution of one in 

 one hundred thousand, it is quite easy to isolate the B. coli 

 from i c.c. of the water." "I would say that a water 

 which contained B. coli so sparingly that 200 c.c. required 

 to be tested in order to find it had probably been polluted 

 with sewage, but the contamination was not of recent 

 date." Chick (Chick, 1900) found 6100 colon bacilli 

 per c.c. in the Manchester ship canal, 55-190 in the pol- 

 luted River Severn, and numbers up to 65,000 per gram in 

 roadside mud. On the other hand, of 38 unpolluted 

 streams and rivulets, 31 gave no Bacillus coli a*hd the 

 other 7 gave i per c.c. or less. The Liverpool tap water, 

 snow, rain, and hail gave no colon bacilli. 



In the United States the colon test has been exten- 

 sively applied during the last few years to certain pol- 

 luted river- waters, in particular with the view of measuring 

 the purification attained by sand filtration and that 

 naturally occurring during the flow of a stream. A fairly 

 good idea has thus been obtained of the numerical dis- 

 tribution of the B. coli in the larger rivers. Fuller (Ful- 

 ler. 1899), for example, recorded the presence of colon 

 bacilli in 60 per cent of the i-c.c. samples taken from the 

 Ohio at Cincinnati. When this water was passed through 



