170 ELEMENTS OF WATER BACTERIOLOGY 



mainly on two considerations, the inaccuracy of the 

 method itself and the difficulties in its interpretation. 

 They hold on the one hand that the errors in the enrich- 

 ment process and the consequent lack of correspondence 

 between duplicate determinations are so great that the 

 whole process is worthless. It is of course true that 

 chance errors of distribution and the overgrowth which 

 often occurs in the lower tubes of a dilution series do 

 occasionally lead to individual erroneous results. If 

 several dilutions are made with duplicates in each 

 dilution, and particularly if reliance is placed, as it 

 should be placed, not on single determinations, but 

 on the average of several tests, results are, however, 

 obtained which are in accord with each other and with 

 the results of practical epidemiological experience. 



The other objection brought forward by many Ger- 

 man sanitarians is that the wide distribution of the colon 

 bacillus leads to its presence even in considerable num- 

 bers in waters which are really of good sanitary quality. 

 It is undoubtedly true that colon bacilli are often 

 found in surface waters which receive no sewage but 

 which are polluted only with the wash from roadways 

 or cultivated land. Even dust blowing in from a road- 

 way may perhaps contribute an appreciable pollution, 

 as we have pointed out above. Gartner (19 10) cal- 

 culates that in the soil of a cultivated field 100 

 meters square there are 15,000,000,000 colon bacilli 

 and points out that it is no wonder that a rain should 

 wash a few of them through into neighboring wells. 

 Konrich (1910) goes so far as to say that to reject on 

 principle all water containing B. coli in i c.c. portions 



