CHEMICAL PRECIPITATION i6i 



have often caused inconvenience by developing luxuriantly in 

 effluents from treatment by metallic precipitants, or from 

 various chemical trades. In October, 1902, I examined the 

 growths at Kingston in the channels and on the top of the 

 continuous filter. In different parts of the latter there were 

 gelatinous grayish-white flocculi, and reddish and nearly black 

 ^ slimy coatings. These consisted of Sphc^rotilus nutans (p. 79), 

 Crenothrix ochracea, and abundance of a cyanophycaceous alga, 

 probably Sirosiphon. The growths in the channel were of 

 similar character. The flakes of both kinds stained with log- 

 wood a deep purple, like an iron-alumina reaction. Iron and 

 aluminium were found in solution in the effluents in larger 

 amounts than usual, and sometimes one and sometimes the 

 other predominated. 



The real strength of all precipitants must be periodically 

 ascertained by analysis. Where they do not deteriorate on 

 keeping, a large weighed quantity can be dissolved in a definite 

 volume of water or effluent, stored in covered tanks, and drawn 

 off by suitable measuring arrangements in proportion to the 

 flow and strength of the sewage. The whole process must be 

 quantitative, and it is hardly necessary to warn against the 

 practice of turning so many hundredweights of crude chemical 

 into a tank of raw sewage, stirring roughly, and taking little 

 notice of imperfect solution or admixture. Many inventions 

 have had for object the automatic supply of precipitants to 

 sewage according to the flow,^ but the fault of these appliances 

 has been that the sewage varies so much in composition that 

 the chemicals will be sometimes in excess and sometimes in 

 deficiency. 



For separating the clarified liquid from the precipitate, either 

 siphoning, a floating arm drawing off from cocks at different 

 levels, or letting out the sludge at the bottom, is applied, with 

 a large number of patented modifications. 



Settling tanks may be constructed on the intermittent system, 

 in which the liquid is allowed to rest quiescent for a certain 

 number of hours, and the clear portion is then decanted. A 

 more usual method is continuous sedimentation, when the whole 

 runs very slowly through a tank of sufficient depth to allow the 

 solids to gravitate, while the clear solution overflows from the 

 top. Santo Crimp stated that the minimum size of the tanks 

 should be such as to hold two hours' sewage flow during the 

 1 See Colonel Moore's "Sanitary Engineering," 1898, p. 443. 



II 



