Presumptive Tests for B. Coli. 145 



lated with water containing many streptococci or other 

 sewage forms. 



On the other hand, with some waters, positive presump- 

 tive tests may be obtained when colon bacilli are not 

 present. According to Clark and Gage (1903) there 

 are fifty-eight well-described species of bacteria which 

 give the presumptive test in dextrose-broth, of which 

 23 are widely separated from the B. coli group. A 

 recent unpublished investigation by Winslow and Phelps, 

 indicates that the result of the dextrose broth test is 

 markedly influenced by the factor of temperature. Their 

 work consisted in the examination of 185 samples of 

 water from 90 different sources, ponds, brooks, pools, 

 wells and springs in five different states (Maine, New 

 Hampshire, Massachusetts, Michigan and Virginia) at 

 three different seasons of the year. All the waters exam- 

 ined were, as far as could be determined, free from spe- 

 cific pollution, although washings from roads or pasture- 

 land might have had access to some of them. Most 

 of the sources were undoubtedly unpolluted, and the 

 examination of 119 samples for B. coli yielded only 12 

 positive results. The presumptive test, however, was 

 obtained in a large proportion of the cases, and much 

 more often in summer than in winter or spring, as 

 indicated in the table on page 146. 



