MODERN METHODS OF WATER PURIFICATION. 343 



British and most European waters; waters possessing" turbidity 

 not exceeding" a proportion now coming" into definition and 

 recognition, which proportion is not one only reckoned by so 

 many grains per gallon, but where the nature and he fineness 

 of division of the particles are the most important ctors. This 

 means, shortly stated, that for many, probably most waters, 

 natural subsidence and slow sand filtration is the most reliable 

 •and proved method. Further, in the case of waters in which it 

 fails to give satisfactory results, it may fail, not necessarily 

 from any defect in the zvorking, but because of the essential 

 and inherent inability of the slow sand filter to deal with that 

 particular class of water, if nothing" more than natural sub- 

 sidence prepares the water for filtration. The Xew Orleans 

 experiments carried out by R. S. Weston, Fuller and Herring" 

 and others, as also a great amount of the work in different parts of 

 the West and South of the United States of America, are avail- 

 able as demonstration. Also the Cincinnati records demon- 

 strate the facts, but particularly New Orleans when a choice 

 was asked of the experts between (i) the English filter, i.e., 

 the subsidence and slow sand filtration; (2) Modified English, 

 i.e.. chemically precipitated water applied to slow sand-beds, 

 but filtered through at double the rate of the first method ; (3) 

 the American, what we usually call the Jewell method of 

 coagulation by aluminic sulphate and rapid mechanical filter 

 treatment, passing the water through this special filter at the 

 rate of some 30 to 40 times the rate of the first method. In 

 New Orleans the last was adopted, the choice being made, on 

 collective advice of the distinguished experts mentioned above, 

 on the consideration of efticiency and economy. We use the 

 second method, and the summing" up, after consideration of 

 the many points at issue on the subject, is: no "roughing" 

 filters" and no "preliminary filtration" or "double filtra- 

 tion " is capable of taking the place of the precipitation. We 

 must coagulate. It is not a question of '' zcith or without" 

 chemical sedimentation, that is the indispensable, we must keep 

 that clearly in mind as the essential in our present state of 

 knowledge. vSome method may be discovered whereby the 

 agency of electricity, or oxygen, or ozone made available from 

 the air in such a form and amount as to purify the water will 

 oust everything" else; but at present, for such water as ours, 

 the chemical treatment is our sheet-anchor in the preparation 

 of the water for the filter, and I trust that will not be regarded 

 as interchangeable with any scheme of "roughing" filters, or 

 increased settling pool accommodation, if such is only to be 

 used as natural subsidence basins. Professor Smith says 

 where water is possibly contaminated by disease germs, he 

 believes that sand-beds alone, or sand-beds with a preliminary 

 "roughing" filter, represent this the only safe method for 

 adoption. He says nothing of the preliminary chemical pre- 

 cipitation, and thus is shown English experience limitation, 

 which does not yet generally recognise that there is something" 

 else needed in certain turbid waters, or the finest sand filters, 

 and the best roughing filters might fail to give w'hat is wanted. 

 Our present system is giving very satisfactory results, but if 



