312 AMERICAN HOMES AND GARDENS August, 1907 



bacterial Beds of First and Second Contact at La Madeleine-les-Lille 



this process are excellent. Unfortunately it retains some of 

 the disadvantages of ordinary irrigation. The quantity of 

 water purified is hmited to three or four quarts per square 

 foot per day, so that the area required for a large city is 

 still very great. Besides, sandy soil is not found everywhere. 



The results obtained with intermittent filtration suggested' 

 to the English chemist Dibdin the idea of effecting the decom- 

 position of the or- 

 ganic matter by the 

 action of bacteria in 

 specially contrived ar- 

 tificial beds, instead 

 of allowing the action 

 to be controlled by 

 the nature of t h e 

 ground, as in the 

 preceding methods. 

 He hoped thus to ac- 

 celerate and regulate 

 t h e w o rk of the 

 bacteria, and conse- 

 quently to purify a 

 much larger quantity 

 of water in propor- 

 tion to the area em- 

 ployed. D i b d i n ' s 

 first experiments 

 were made in iSqi;. 

 His artificial bacterial 

 beds were composed 



of a mixture of coke and burned clay. The water, after 

 settling and being decanted, was caused to flood the first 

 bed, where it remained two hours. The partially purified 

 water then flowed to a second bed, where in two hours more 

 it was purified so well that it could safely be allowed to flow 

 into streams. After each flooding the beds were exposed to 

 the air for four hours to enable the bacteria to multiply. 

 But it was discovered that these beds soon became 

 clogged with filth, and then required two or three 

 weeks' rest to fit them for further service. To 

 avoid this inconvenience Cameron conceived the 

 idea of interposing between the sewage supply pipe 

 and the bacterial beds a series of ditches called 

 septic fosses, in which the putrefiable matter in sus- 

 pension could be deposited and then dissolved and 

 decomposed by anaerobic bacteria.* The water 

 which flows from the fosses holds almost no matter 

 in suspension, and Its dissolved organic matter is 

 very easily decomposed by the bacteria of the beds, 

 which, consequently, do not become clogged. 



The biological processes, as these methods of 

 Dibdin, thus modified, are now called, have been 



Septic Fosses at La Madeleine-les-Lille 



tested at Exeter, Veovil, Manchester and else- 

 where. The results obtained have generally been 

 good, and the study of these methods has greatly 

 developed in the last few years. With the energetic 

 and fruitful initiative of Dr. Calmette, director of 

 the Pasteur Institute of Lille, and the aid of a 

 large subsidy from the national fund for scientific 

 research, France has now, in turn, taken up the 

 study of the biological methods, and sufficient 

 work has already been done to show that these 

 methods are very practical and very advantageous. 



Let us see, then, how these methods should be 

 applied in practice, in accordance with the investi- 

 gations made at the experimental station for the 

 purification of water, at La Madeleine-les-Lille. 

 Artificial bacterial purification comprises three 

 operations : First, the separation of non-putrescible solid 

 residuum (sand, stones, fragments of metal, etc.); second, 

 the solution, in the septic fosses, of organic matter in sus- 

 pension, and its partial conversion into gases; third, the de- 

 composition of soluble organic matter in the bacterial beds. 



The first operation is purely mechanical. The sewage 

 first traverses a chamber with double gratings which retain 



floating bodies, and 

 then flows very 

 slowly into a cham- 

 ber containing a thin 

 layer of sand, on 

 which the gravel and 

 metallic particles are 

 deposited. This de- 

 posit is removed at 

 regular intervals 

 with hand or chain 

 scrapers. 



The water next 

 enters the septic 

 fosses. These are 

 long basins of rect- 

 angular cross-section, 

 lined with masonry 

 and subdivided by in- 

 complete transverse 

 partitions which 

 check the flow of the 

 water and facilitate 

 deposition of sediment. Their depth is ten or twelve feet, 

 their length and width such that they contain one day's out- 

 put of the sewer so that each gallon of water occupies twenty- 

 four hours in traversing the length of the system, which may 

 consist of several parallel fosses or one very long one. The 

 fosses discharge by overflow, and are consequently always 

 full. The muddy sediment which collects in the bottom of 



* Bacteria which, unlike the aerobic bacteria'of the filter beds, 

 do not require the presence of air. 



Bacterial Bed with Fiddian Rotary Distributor at La Madeleine-les- Lille 



