BACTERIAL PURIFICATION 267 



the filter-beds, therefore there has been no difficulty in estab- 

 lishing nitrification." This remark of course applies to the use 

 of a close-textured filtering material afterwards, in this case 

 gravel and sand. But the preliminary resolution of the solids is 

 very imperfect, and the sludge has to be removed from the 

 chambers once a month, amounting to 6 or 8 loads, the sewage 

 being 40,000 gallons daily. At intervals a sediment has also to 

 be removed from the surface of the filter-beds, and every few 

 weeks they require raking, or clogging and offensiveness would 

 occur. 



" Hydrolytic Tank.*' Observations in glass bottles or tanks, 

 such as those of Mouras about 1880 (see Chap. IX.), Kenwood 

 and Butler and other investigators, and those commenced by 

 Travis with a large glass model at the Hampton Sewage Works 

 in 1904, are often very useful, but, besides the difference of 

 scale, there are other divergences from actual practice, such as 

 the usual working in light instead of in darkness, and with a 

 smooth, almost sterile, instead of a rough fertile, surface. 

 Travis's patents, 11073 of 1903 and 4980 of 1905, have been 

 for some time in operation at Hampton. The sewage passes 

 through a screen and detritus chamber into a tank divided 

 longitudinally into 3 parts by flagstones sloping outwards, so 

 that the centre compartment resembles in section an inverted 

 funnel, and the lateral ones are wedge-shaped. Entering first 

 these side " sedimentation chambers," the sewage deposits a 

 great part of its remaining solids, which pass, with a portion of 

 the liquid, through holes at the bottom of the wedges, into the 

 central " Hquefying " or " reduction " chamber, in which fer- 

 mentation goes on, much gas is evolved, and the sludge accu- 

 mulates. The tanks are in duplicate ; every fortnight the flow is 

 diverted from one to the other, and the sludge is run out from the 

 full tank through valves at the bottom. From the tank the 

 sewage passes over weirs into an upward anaerobic filter, which 

 at first was practically the same as Moncrieff's cultivation tank 

 (p. 268), and later was fitted with inclined plates, like some of 

 the machines used for removing lime-deposit in water-softening. 

 In these " hydrolysing chambers " (four in number, arranged in 

 sequence) the sludge is again removed frequently through open- 

 ings below. The plant is roofed, and the gases are drawn off 

 by a fan into a chimney.^ 



^ Jones and Travis, Proc. Inst. Civ. Engineers, 3599, 1905 ; Travis, Trans. Civ. 

 and Median. Eng. Soc, May 3, 1906. 



