BACTERIAL PURIFICATION 



217 



[.within 6 inches from the surface, and the sewage remains in 

 contact for two hours, after which the valve is opened and the 

 iliquid flows on to the fine bed, in which it is similarly treated. 

 [The coarse-grain filters are charged three times daily, with an 

 interval of not less than two hours' rest after being emptied. 

 jThe filling and emptying are controlled by Adams' siphons 

 [{Chapter XII.). An additional rest, of about one week in six, 

 is given to each bed. 



After three months' working, Mr. Dibdin was able to give a 

 ^satisfactory report. The oxygen consumed by the organic 

 latter was reduced by the tank 66 per cent., and by the filter- 

 fbeds 86*5 per cent. The solids in suspension were reduced by 

 :he tank 95 per cent., and by the filter 99*6 per cent., while the 

 [filtrate was practically clear, had no objectionable odour, and 

 [did not putrefy on keeping. The process has continued to the 

 [present time with satisfactory results, except when the filters 

 [were overtaxed, *' some of them," as Mr. Dibdin reports, " having 

 [been purposely worked up to a rate of nearly 3,000,000 gallons 

 )er acre per day, with the result that the bacterial action was 

 [evidently checked, as shown by the decrease in the production 

 )f nitrates, and an increase in the quantity of organic con- 

 stituents in the effluent. As the result of careful watching, 

 lowever, no permanent harm was done, as the filters were 

 immediately restored to their usual condition, when they pro- 

 fceeded to give good results." 



Here, again, we gather that when there is reliance on pre- 

 jumably aerobic filters and organisms for combined liquefaction 

 [and nitration, indiscriminately, in the same receptacles, the 

 result is apt to be variable, and to depend on " careful watch- 

 ing," an inference that is borne out by Mr. Dibdin's analyses 

 luring 1896 and 1897. 



Sutton System. 

 Parts per 100,000. 



