10 ROYAL SOCIETY OF CANADA 



Thus about 16 per cent of all samples analysed were infected with 

 B. coli, the infection when the river was frozen over being about 30 

 per cent, and 22 per cent when the river was free of ice. Dr. Amyot 

 also reports two analyses of river water from Hawkesbury, in both of 

 which B. coli was present in 1 cc. quantities. 



These figures serve to show that at least the water starts above the 

 city of Ottawa without much infection. 



The river water intake pipe which provides the water supply of 

 Macdonald College, at Ste Anne's, runs 150 feet into the river and 

 obtains water at a depth of 12 to 14 feet. The samples for analysis were 

 tsken at the pump, in sterilised Iwttles and immediately brought to the 

 laboratory and the analyses started. 



The river water is pumped directly through a Jewel Filter manu- 

 factured by the " New York Continental Jewel Filtration Company." 

 AiTunonia alum is used as a coagulant, but there is no settling tank, the 

 water being pumped directly into the distributing pipes. 



The samples of filtered water were obtained from the power house, 

 being taken from a faucet on the service pipe leading from the filter. 



The analysis was commenced immediately, and all possible pre- 

 cautions were taken to ensure proper sampling and correct results. 



The media employed were beef peptone gelatine and beef peptone 

 agar, prepared according to the directions of the Laboratory Committee 

 of the American Public Health Association. The colon tests were made 

 in our sesculin bile-salt-agar medium, and subcultures of black colonies 

 were frequently made on the set of media suggested by Savage and 

 Houston 2 for the necessary identification of B. coli. 



The two most striking features of the analyses were the large 

 fluctuation in the total numbers of bacteria and the almost constant 

 presence of B. coli in large numbers. 



The direction and force of the wind and rain undoubtedly tend to 

 produce higher bacterial counts. The river is wide, and after a strong 

 breeze (particularly if it is in a direction opposite to the river flow) 

 the water becomes very turbid and analyses made of such water invariably 

 give higher counts. Eain also stirs up the water, but its greater effect 

 on the bacterial content of the stream is due to the surface washijigs 

 falling into the river. November, September, June and April gave the 

 highest total counts, in the order named. 



The chief interest in these analyses from the hygienic standpoint 

 is the number of B. coli. The varying opinions which are held as to 

 numerical standards of B coli are well known, but assuming that Savage's 

 standard is about correct " that the finding of excrétai B. coli in 10 cc. 

 or less points to undesirable pollution, and that if B. coli is present in 

 smaller amounts — e.g., 1 cc. or less — the water can be safely condemned 



