CHEMICAL PRECIPITATION 



159 



ind ferric salts as precipitants is not only that the former act 

 LS reducers, diminishing the amount of free oxygen available, 



[but that ferrous oxide is soluble in alkaline liquors, while ferric 

 >xide is almost entirely precipitated, so that a liquid treated 

 'ith ferric or per-salts of iron after filtration or deposition may 

 contain no iron, whereas one from ferrous or proto-salts, such 

 LS copperas, retains iron dissolved in the ferrous state, and 

 >n exposure to the air gradually oxidizes and gives rusty 

 leposits. 



When iron salts have been added to sewage, I have observed 

 that a residue of the metal was always left in solution, with 

 ferrous salts from the solvent action of alkalies already men- 



ftioned, with ferric compounds from the well-known fact that 



[organic matter prevents their precipitation by alkalies. After 

 time, if not thoroughly aerated, a black deposit of sulphide 



[of iron is liable to form, and is often seen on sides of channels. 



[Aluminium salts have not these disadvantages. The presence 

 )f any of these chemicals in more than traces is injurious to 

 ish and hinders nitrification. 



In the Massachusetts experiments on the effect of different 

 imounts of chemicals in removing micro-organisms it was 

 found better to add the metallic salts first, and then an equiva- 



[lent amount of lime afterwards. Ferric sulphate gave the best 

 results as to removal of organisms and organic matter, cop- 



[peras or alum acting about equally in the second place. In 



|<:ost, their table gives the preference to copperas (ferrous sul- 



[phate) and lime. 



At the London County Council's Works, Mr. Dibdin adopted 



[Z grain of copperas and 4 grains of lime to i gallon sewage, 



ifter a long series of experiments with various amounts of 



iifferent precipitants.^ His conclusions are that the following 



■rules should, as far as practicable, be observed : (i) The sewage 



|should be diluted as httle as possible; (2) agitation after mixing 

 should be avoided ; (3) unless absolutely necessary, no pumping 

 [Should take place before precipitation, the reason apparently 

 being that the entanglement of air with the precipitate prevents 

 settling. He also infers that with lime iron is superior to 

 alumina and also cheaper, and that a large increase in the 

 ■quantity of chemicals yields no advantage. 



At York in 1896 the sewage was chemically treated with 



1 His results were given in detail in the second edition of the present work, 

 P- 141. 



