BACTERIAL PURIFICATION 233 



H *' I. That the initial capacity of a contact bed is practically 

 uninfluenced by the grade of material with which it is filled.^ 



" 2. That there is a rapid decrease in capacity during the 

 earlier period of working [before the resolving bacteria become 

 established and active]." (See also ante, p. 224.) 



After noticing the increase of capacity during a period of rest, 

 the report concludes that coarse cinders, 3 inches to i inch, 

 permit too free access of sludge to the body of the filter and 

 even into the drains, while " if the material is too fine the 

 beds soon become quite impervious to sewage." With bed C, 

 I to J inch, followed by D, J to ^ inch, they obained better 

 I results, but their final opinion is that the most suitable material 

 ! for bacterial beds consists of clinkers passing through i^-inch 

 mesh and rejected by ^-inch. 



It was concluded that " contact beds, after a comparatively 

 short space of time, acquire a practically constant capacity," 

 usually found to be about 33 per cent. ; that suspended matter 

 must be removed as far as possible by sedimentation, and that 

 ' any not so removed should be retained on the surface of the 

 bed ; that the surface must be raked or forked over about once 

 a month ; and that periodical intervals of rest must be allowed. 



At Sutton Mansergh reported that in three years the water- 

 content of the coarse filters had diminished from 32 to 19 per 

 cent., while the fine filters had not sensibly lost capacity. At 

 Hamburg, with single contact without previous sedimentation, 

 Dunbar and Thumm found that with one filling per day, after 

 700 fillings the original capacity of 33 per cent, had fallen to 

 20 per cent., and with two fillings an original 40 per cent, had 

 fallen to 14. Resting, forking, flushing, and altering the direc- 

 tion of supply, did not prevent the clogging, and the only 

 remedy was removal of the material and washing against 

 screens, by which about 10 per cent, of the medium was lost. 



The London County Council investigated in i8gg the effect 

 of "double treatment" — that is, by an extra coke bed. Their 

 " single treatment" meant two coke beds, the first correspond- 

 ing to an anaerobic tank, and about 4 feet deep ; the second of 

 6 feet thickness, called the " primary bed, for the first stage of 

 double treatment"; while the third, also 6 feet, was called the 

 " secondary coke bed," corresponding, however, to what is 

 commonly named triple treatment. The coke in all was of 

 uniform size, " about as large as a walnut." The primary and 



^ This is only the case within certain limits. See p. 226. 



