242 



STEPHEN DEM. GAGE 



numbers correspond fairly well with the probable difference in the 

 character of the two waters. In other words, our ratios indicate 

 that both Applied 216 and Applied 219 were more dangerous from 

 a sanitary standpoint than the difference between their total bacterial 

 counts and the count on the canal water would indicate. 



Bacteria-B. coll ratios on Merrimack River water. We will now 

 consider the ratio between the numbers of bacteria and the numbers 

 of B. coliy this being the one of the new factors here introduced upon 

 which we have the most extensive data. Routine determinations 

 of bacteria and B. coli have been made on samples of Merrimack 

 River water from the Intake during a period of seven years, and on 

 samples from the canal during a period of eight years. The average 

 monthly numbers of bacteria and B. coli in samples from these two 

 sources have been published annually in the reports of the Massa- 

 chusetts State Board of Health, and it is unnecessary to reproduce 

 them here. The ratios between the bacteria and B. coli computed 

 from the monthly averages for these two sources are shown in Tables 

 10 and ii. The use of the monthly averages tends to eliminate 

 abnormal values and fluctuations, and to give us results which are very 



TABLE 10. 



AVERAGE MONTHLY BACTERIA-JB. coli RATIOS ON SAMPLES FROM THE MERRIMACK RIVER AT THE 

 INTAKE OF THE CITY FILTER, 1899-1905, INCLUSIVE. 



nearly normal, and the fluctuations occurring from month to month 

 must be considered normal fluctuations. These fluctuations from 

 the normal and the effect of various factors in producing normal 

 fluctuations will be considered later. 



The numbers of bacteria and of B. coli in the Intake samples 

 have usually been somewhat greater than those in the canal samples. 



