700 A MANUAL OF BACTERIOLOGY 



in flood time or drought, may be avoided. Moreover, 

 storage alone usually markedly diminishes the number 

 of organisms, partly by subsidence, partly by lack of 

 aeration, and partly probably owing to the struggle for 

 existence going on among them. 



(2) Sand filtration. Efficient sand filtration removes 

 quite 99 per cent, of the organisms originally present. 

 The fine sand only has to be taken into account in estimat- 

 ing the removal of organisms and efficiency of a filter 

 bacteriologically. It probably should form a layer not 

 less than 3 ft. to 3 ft. 6 in. in thickness. 



The removal of organisms is less perfect when the rate 

 of filtration is rapid. At the Metropolitan Water Board 

 Works the rate of filtration varied from a minimum of 

 0-98 (Hampton) to a maximum of 2-53 (Stoke Newington) 

 gallons per square foot per hour during 1919-20. 



New, or recently cleaned, filter-beds allow a large 

 number of organisms to pass through. A filter-bed which 

 is not efficient at first becomes so when the surface film 

 forms, composed of sedimented particulate matter, and of 

 a zooglreal mass of bacteria and algse. The beds must be 

 cleaned occasionally by raking up and clearing away 

 the surface layer of sand, for as time elapses the rate of 

 filtration becomes slower and slower, though the bacterial 

 efficiency of the filter-beds does not appear to be reduced 

 by prolonged use. The normal bacterial efficiency is 

 rapidly regained after cleaning within two or three 

 days. 



(3) Sedimentation. Besides storage and filtration, 

 sedimentation in the presence of fine particles, either 

 naturally present or artificially added, may also effect a 

 marked removal of micro-organisms from water. Thus, 

 by the addition of alum, an old method of clarifying turbid 

 water, a large number of the organisms present are carried 

 down in the precipitate. 



