138 Elements of Water Bacteriology. 



not occur from day to day and average results are gen- 

 erally reliable. It is wholly misleading, however, to com- 

 pare such results with the average examinations of an 

 imfiltered surface water. With surface waters daily vari- 

 ations are the rule and a low monthly average of colon 

 tests may include and cover up dangerous and significant 

 high numbers at particular periods. 



The general results of the studies of the colon tests 

 which have now been carried out ' in great numbers all 

 over the world may be summarized by a few further 

 citations. 



In America the fact that the number of colon bacilli in 

 a water measures the degree of its pollution is now univer- 

 sally accepted. The same conclusion has been estab- 

 lished in England by the elaborate investigations of 

 Houston and his pupils. Savage, for example, concluded 

 (Savage, 1902) from a study of a large number of water 

 supplies in Wales, that even in surface waters, exposed 

 to animal contamination from adjacent grazing grounds, 

 B. coli is not present in 2 c.c. unless other pollution is 

 present. In a more recent review of the whole subject, 

 the same author (Savage, 1906) concludes that "there is 

 no evidence or observations which have ever shown that 

 B. coli, reasonably defined, is present in any numbers in 

 sources which have not been exposed to some form of 

 fecal contamination." 



In Germany, Petruschky .and Pusch (Petruschky and 

 Pusch, 1903) examined a considerable series of waters 



