^S WATER SUPPLY FOR DOMESTIC USE. 



of sand and gravel — about 2 feet of the former to 3 feet of the 

 latter. The greater portion of the London water supply is 

 filtered — the New River Company having 20 filter-beds with an 

 area of 16^ acres, the filtering medium being — sand 2ft. 3in., 

 gravel 3ft. to 5ft. 3in. 



The East London Waterworks have 31 filter-beds, with an 

 area of 29f acres, the filtering medium being — sand 2 feet, hoggin 

 6 inches, coarse gravel 1ft., equal to 3ft. 6in. 



The South war k and Vauxhall Waterworks use a filtering 

 material of sand 3ft., hoggin 1ft., fine gravel 9in., coarse gravel 

 9in., equal to 5ft. 6in. in all. 



The West Middlesex and Grand Junction Waterworks use 

 a filtering medium of sand and gravel 5ft. 6in. deep ; the 

 Lambeth Waterworks, one of 7 feet ; and the Chelsea Water- 

 works filter-beds 8 feet deep. The filtered water is conveyed 

 to covered tanks, from which it is drawn off by mains to be 

 distributed. 



The filter-beds require cleaning about once in six weeks or 

 oftener, according to the condition of the water, and may need 

 renewing every six months. 



The effectual filtration of water depends on — (1) a sufficient 

 area of properly constructed filter-beds, cleaned and fresh sanded 

 from time to time as required ; (2) the rate of filtration being 

 controlled and regulated; (8) and the water, if at all turbid, 

 being first received into subsidence tanks, in order that the sus- 

 pended matters may be deposited before the water is run on to 

 the filter-beds, thereby maintaining their efficiency for a maxi- 

 mum period by preventing unnecessary fouling. After a filter- 

 bed has been in use for a short time, a film of gelatinous material 

 forms on the surface of the sand. When this film becomes 

 thick enough to interfere with the usual rate of filtration it is 

 skimmed off, and the filter-bed allowed a certain period of rest. 



This gelatinous film forms a very important function in 

 relation to bacterial organisms, as it effectually intercepts by far 

 the greater number contained in the water passed through the 

 filter. The bacteria become attached to and entangled in the 

 colloidal mass, and are consequently prevented from passing 

 down into the deeper beds of sand and gravel. Professor Ray 

 Lankester describes this filtration as like the dialyses through a 



