XLII. J. M. SMAIL 



penetration into the sand layers of matters removed from 

 the water. "Regarding the percentage of wash- water, the 

 actual volume of water required per filter for cleaning does 

 not seem to be exactly proportionate to the amount of 

 accumulated matters in the sand layer." 



I am indebted to G. W. Fuller, m. Am. soc. ce„ for the 

 foregoing description of this plant. 



From personal observation of the system under working 

 conditions, I was impressed with the ease by which the 

 attendant controlled the rate of filtration and washing of 

 the Alters. One attendant is all that is required on watch 

 in the filter room, and two others in the machinery and 

 coagulating rooms. After examination of the raw water, 

 and the same after passing through the coagulating process 

 and filters, the change in the appearance of the water was 

 marked. The results of the chemical analyses and bacterio- 

 logical examination show that the water would be con- 

 sidered a good potable water. The concentration of a plant 

 capable of dealing with so large a quantity as 32,000,000 

 T.S. gallons per diem, within the area occupied, leads to 

 the consideration as to whether this system (judging by 

 the results shown) does not bear favourable comparison 

 with the English method of slow filtration, which requires 

 areas ranging from 0*42 acres to 1*33 acres per million 

 gallons (Imperial), exclusive of subsiding reservoirs. I ven- 

 ture to think that it does. The initial cost of a slow filtra- 

 tion plant to deal with the quantity dealt with at these 

 works would be higher than the American system. If 

 coagulation could be obtained at a comparatively low rate, 

 I think the working expenses would be lower per million 

 gallons treated. 



Cost of Filters. 

 The cost of construction depends upon local conditions of 

 site, cost of land, etc., together with the character of the 



