274 SEWAGE AND ITS PURIFICATION 



scheme for conveyance of the present effluent through a tunnel 

 to the tidal part of the Mersey. 



The experimental plant of 1899 consisted of three independent 

 sections. 



I. Five " Bacteria Beds," with sloping sides/ filled with 3 ft. 

 of clinker of the following sizes : A, 3 in. to i in. ; B, i in. to 

 ^ in. ; C, I in. to J in. ; D and E (extra bed), J in. to ^ in. 



Distribution by perforated wooden troughs, collection by drain 

 pipes below. Area of A to D J^ acre. A and B were started 

 with settled sewage in September, 1898, run on at first once, 

 then twice, and finally three times a day. A month later raw 

 sewage, screened through a grid, was applied at a rate of four fill- 

 ings per day, each cycle including : filling f hour, resting full 

 2 hours, emptying f hour, resting empty 2J hours. After a 

 week the surface of bed A showed signs of clogging, so settled 

 sewage was again used. Subsequently the time of filling A was 

 shortened ^ hour, and of emptying to J hour, to give a longer 

 period of rest, and also because it was thought that by *' rush- 

 ing" sewage on to the filter a larger amount of air would be 

 entangled with it. The result would belong to the partially 

 aerobic class, but with as much as possible avoidance of suspended 

 solids. Samples were taken at short intervals and mixed for 

 analyses. Beds C and D were first used together, being filled 

 twice a day with settled sewage, with J-hour filling and two hours 

 contact. As A was too coarse, it was changed to |" to J'' mesh. 



II. Septic Tank System. — Concrete tank 40 x 12 ft., with 

 arched roof 9 ft. high, and air-tight manholes ; rest of construc- 

 tion as at Exeter. Six filter-beds with vertical concrete walls ; 

 total area 196 sq. yds., and depth 4 ft., containing from the 

 bottom upwards — i ft. clinker between 3 in. and i in. mesh : 

 2 ft. 9 in. clinker between f in. and | in. mesh : 3 in. residue 

 from above, passing at J in. mesh. 



Raw unscreened sewage was passed through, but not so 

 regularly as at Exeter. 



III. Two Roscoe Filters, first used in December, 1895, were 

 12 ft. 6 in. by 18 ft., or 25 sq. yds., with 3 ft. of rough clinker, 

 graded coke or cinders, and a covering of i ft. of clean gravel. 

 They were worked like the bacteria beds, but with chemically 

 precipitated effluent. 



1 Sloping sides obviously give greater surface and aeration at a sacrifice of 

 capacity which can be calculated from the angle of inclination of the sides. 

 Hence the adoption of this shape, lately recommended, would largely increase 

 area, and would only be justified by increased efficiency, on which further experi- 

 ments are needed. 



