PURIFICATION OF WATER FOR DRINKING PURPOSES 133 



mixed with water whicli has been stored for a longer 

 or shorter period in this service reservoir. Notwith- 

 standing this disturbing factor, the contrast between 

 the water before and after filtration is sufficiently 

 marked. The average number of micro-organisms 

 found in the unfiltered water taken from the surface 

 of the four filter beds amounted to 1,498 per c.c, 

 whilst that in the water from the filter well was only 

 85 in the same volume. 



On August 13, 1892, the unfiltered water con- 

 tained rather less bacteria than on the previous occa- 

 sion, and that on the New West filter bed contained 

 markedly less than a.ny of the others, which is probably 

 to be accounted for by the fact that the water in this 

 filter bed was in a more quiescent state than that in 

 the others, in consequence of the unfiltered water having 

 to pass through the other beds before reaching this 

 one, which forms as it were the ^dead end' of the sys- 

 tem, and in this manner it furnishes further evidence 

 of the removal of micro-organisms through subsidence. 

 The average number in the unfiltered water was 887 

 per c.c. Samples were taken of the filtered water 

 issuing from each of the four filter beds, and in each 

 only a very small number of micro-organisms was 

 found (maximum 25, minimum 17, average 21). These 

 filters had been working from twelve to thirty-five days 

 since being last cleaned, but the figures show that 

 there was no difierence in their efficiency. The number 

 of micro-organisms in the water of the general filter 

 well was also practically the same as in that from the 

 individual beds, and this was to be anticipated, as 

 I was informed by the resident engineer, Mr. Loam, 

 that the tendency at the time would be rather for 

 water to pass from the filter well into the small service 

 reservoir than in the opposite direction. 



