I 



BACTERIAL PURIFICATION 



269 



the organisms gradually chose their own conditions, and allied 

 groups gather together at different levels as coatings on the 

 filtering material. The advantage of separating the organisms 

 appeared early from a remark of Jordan and Richards, in the 

 Massachusetts Report of 1890 (ii., 877), that " in the filter- 

 tanks at the Lawrence Experiment Station, speedy nitrification 

 was always coincident with a marked decline in the numbers of 

 bacteria. The more complete the nitrification, the fewer were 

 the bacteria in the effluent." In the latter sections the nitrify- 

 ing organisms should be almost alone, and therefore able to 

 exert their full activity. In this way Moncrieff secured a much 

 higher nitrification than was obtained by the other processes. 



This he has accomplished by spreading the " tank effluent " 

 by tipping troughs or distributors over the uppermost of a series 

 of " nitrifying trays." (See Fig. 29.) In experiments at Ash- 

 tead with a domestic sewage, nine perforated trays, each having 

 an effective area of i square foot and containing 7 inches of coke 

 broken to i inch in diameter, were supported vertically over 

 one another at about three inches apart. It required only from 

 eight to ten minutes for the liquid to pass through all the trays. 

 In 1898, after the apparatus had been running continuously for 

 three months at a flow equal to one million gallons per acre 

 per 24 hours, I collected on two occasions samples from the 

 different trays and examined them separately. 



The results of these analyses of the tank effluent and final 

 filtrate from the ninth tray are given in the table : — 



Parts per 100,000. 



