156 BACTERIA IN WATER 



the same is true when they are placed in natural conditions, 

 which involve their living in the presence of other organisms, is 

 unknown. In the case of such organisms we therefore seek for 

 the presence of indirect bacteriological evidence which might 

 point to the contamination of a water by human excreta. If 

 this be found we deduce that the water is dangerous, as organisms 

 from any case of intestinal disease occurring in the catchment 

 area may find access to it. The criterion here adopted is the 

 determination of the numbers of b. coli present in the water. 

 Klein and Houston point out that, in crufle sewage, members of 

 the coli group are practically never fewer than 100,00.0 per c.c, 

 and their detection is relatively easy by the methods to be 

 described later. In these circumstances, all modern work tends 

 to taking the presence of b. coli in a water as the best indirect 

 evidence of the possibility of disease organisms of intestinal 

 origin being likely to gain access to that water. It must, how- 

 ever, be at once clearly recognised that the presence of members ' 

 of the coli group is only an indication, and so far as the pota- 

 bility of any water is concerned, evidence is wanting that these 

 organisms, however undesirable, are under ordinary circumstances 

 actually harmful to man. 



The difficulty, however, is that (except in the case of water 

 from artesian wells), if a sufficient quantity be taken, evidence 

 of the presence of b. coli will always be found. This arises from 

 the fact that the organism is as numerous in the excreta of birds 

 ■ and other animals as in those of man, and it is impossible in the 

 present state of knowledge to distinguish between organisms 

 coming from these different sources. Thus in the moorland 

 waters so much used for urban supplies, there may be a high 

 content of b. coli — for example, 100 or more per c.c. — with- 

 out the least evidence of more than the most infinitesimal fraction 

 of these being derived from human sources, and the consumption 

 of such a water, even in an unfiltered condition, may be perfectly 

 safe.. On the other hand, a heavily contaminated surface well 

 may show no b. coli. to be present. There is thus the greatest 

 difficulty in the interpretation of bacteriological results in deal- 

 ing with raw waters, and it is impossible to set up any 

 standards of the bacteriological purity of a water based on the 

 estimation of the numbers of b. coli present alone. In any 

 particular case the results must be considered along with those 

 of chemical analysis and with the inspection of the locality. 

 The difficulty is greatest when dealing with water derived from 

 sewage-contaminated rivers, from agricultural land, and from 

 surface wells. With regard to the- .firstj two sources, the water 



