126 ELEMENTS OF WATER BACTERIOLOGY 



Hale and Melia (1910) have also made a valuable 

 comparison of a number of presumptive tests as applied 

 during a period of 2 years to 85 samples of Manhattan 

 water (surface) and 160 samples of Brooklyn water 

 (largely ground-water). Their principal results are 

 tabulated below. 



RESULTS OF VARIOUS PRESUMPTIVE TESTS 



(HALE AND MELIA, 1910) 

 245 Samples of New York Water 



This table indicates very clearly the fallacies of the 

 dextrose broth tube. Counting all gas formers in this 

 medium as positive indicates much too high a value 

 and including only the tubes showing the standard gas 

 formula gives much too low a value. This conclusion 

 is based on the assumption, warranted by the results 

 of many workers, that incubation in dextrose broth 

 followed by reinoculation into lactose bile (fourth line 

 of the table) gives with reasonable accuracy the real 

 number of colon bacilli present. By this standard 

 the use of plain lactose bile with these waters is seen 

 to give results which are also much too low; but lac- 

 tose peptone bile approximates closely to the truth. 



