BACTERIAL PURIFICATION 229 



burnt ballast works better than clinker. The contact beds in 

 this case are 4 feet 6 inches deep, and the flow of sewage 

 400,000 gallons per day. The amount of nitrate in the effluent 

 has been as high as 5*2 parts per 100,000.^ 



In some places where the beds have been made by the simple 

 method mentioned in p. 216, care has not been taken that the 

 work was sound, and earth and clay from the sides have washed 

 into the filter. 



From the enormous waste heaps in the neighbourhood of 

 various collieries and iron works, Wake and Hollis, Darlington, 

 separated by special machinery tap cinder, coke breeze, clinker, 

 broken bricks, and " carbonaceous iron sand "; the latter was 

 proposed by them as a suitable and cheap material for bacteria 

 beds.^ Its content of iron oxides, with a trace of manganese, 

 suggested the action of these substances as carriers of oxygen 

 to the organic matter (p. 179). But, as in coke, which also 

 contains iron, and often manganese, and in many patented 

 materials prepared by ignition, the density and insolubility 

 make chemical action very slight, and quite subordinate to the 

 use as strainers and bacterial surfaces. 



Dibdin and Thudichum compared "carbonaceous iron sand" 

 with other materials in some experiments on triple filtration of 

 crude sewage. In each set the size of the grain was : — 



First bed, passed by J-inch mesh and rejected by J-inch. 

 Second „ ,, ^-inch ,, „ ,, ^-inch. 



Third ,, „ |-inch ,, „ „ ^V-i^^ch. 



After five weeks' running with two fillings per day, the 

 effluents were analysed. The composition of the crude sewage, 

 calculated to parts per 100,000, is given as: Free NH3 13*53, 

 Albumd. 0740, Oxygen consumed 9*14. The effect in reduction 

 of these figures, or " percentage of purification," the water 

 capacity at this stage, and the nitrogen oxidized, is shown in 

 the annexed table. From it we may gather the following 

 conclusions : — 



I. The iron sand shows in this case a slight superiority over 

 the coke breeze alone, but, in the words of the report, " it was 

 practically identical in work effected with the other materials, 

 and there was no specific advantage in the use of any one 



^ Sanitary Record, April 20, 1905, p. 333. 



2 An analysis by W. F. K. Stock gives— moisture 675, FeO 30-41, FeoOg 10-33, 

 carbon 7-53, rough sand 16-70. 



