22 



ROYAL SOCIETY OF CANADA 



As under natural conditions the numbers of B, coli were consider- 

 ably less; on August 7th, we transferred small quantities of water from 

 flasks 5 to 10 into 6 two-litre flasks of river water; four of these were 

 kept at 10° to 16° C. and two at 22° to 27° C. The number of B. coli 

 in each flask was determined daily for a month and after that at week 

 intervals. 



Number of B. Coli per cc. in River Water — Kept in Flasks. 



The flasks used were of Jena glass and had been carefully washed 

 with potassium bichromate and sulphuric acid, rinsed with distilled 

 water, and sterilised to remove any traces of ammonia. 



As it was necessary to remove larger quantities of water to de- 

 termine the presence of B. coli the test was continued, but liquid aesculin 



