332 MODERN' METHODS OF WATER PURIFICATION. 



favour of sand filtration, and gave all the well-known data of 

 the German standards of maxima and minima of sand and 

 gravel and stones, and head of water, and amount of filtration 

 rate, and quality and size of sand, and also recommended 

 double filtration if found necessary. All these data I need not 

 repeat. Nothing new has been added to the subject since that 

 report as far as normal English slow sand filtration goes, i.e., 

 without coagulation. Coagulation has changed everything. 

 It is the sine qua non for us now. I believe if we on that 

 commission had known in 1902 what 1 know now, we would 

 have brought in a recommendation, in view of the turbidity 

 of our water, to try the American filtration. Purity of water 

 and bacterial efficiency is first, economical working' of a com- 

 pact modern up-to-date apparatus at cheaper first cost and 

 cheaper upkeep is only second, but there is evidence accumu- 

 lating to show that properly run and carefully tended there is 

 prospect of equal results. The Hyatt patent, which underlay 

 all the American mechanical filters, only expired in 1901, when 

 the field was opened for improvements, and the prog"ress has 

 been great and rapid since then. 



Before shortly describing a few of the systems I have in- 

 vestigated I will quote the opinion of some experienced 

 authorities on the all-important question of bacterial efficiency. 

 Hazen says tlmt the efficiency of the American mechanical filter 

 (in the bacterial efficiency respect) depends absolutely upon 

 the faithful and continuous application of the coagulant, and 

 the securing at all times the application of the coagulant to 

 the water in the right amount. Comparing sand filter efficiency 

 under suitable conditions, he says, generally speaking, sand 

 filters are adapted to treating only such waters as are capable 

 of being purified in that way without any preliminary treat- 

 ment. If the water ordinarily requires coagulation, then, as a 

 rule, it will be better to make the coagulation thoroug'h and use 

 mechanical filters for the final treatment. Fuller says from 

 the analytical standpoint there is no very decided difference 

 in efficiency between the two types of filters (rapid mechanical 

 and slow sand) when both are well constructed and well 

 operated. Gage said that the Lawrence sand filter cleared the 

 town of typhoid, and was for New England waters the better 

 adapted; for the muddy waters of the south coagulation was 

 required, as sand filters could not deal with their turbidity, and 

 coagulation would be accompanied by adoption of the rapid 

 mechanical filters. He considered the advantage of the Eng- 

 lish sand filter over the American mechanical was that it was 

 fool-proof. W. R. Smith states that Bitter and Gottschlich, 

 who carefully examined the action of the American filters at 

 Alexandria, found that a reduction of i : 2,000 or more of the 

 bacteria was regularly obtained, but that it took 30 minutes of 

 working the filters, after washing, before these results were 

 obtainable. 



The American mechanical filters include the Jewell and the 

 Warren, but. as noted above, the expiry of the Hyatt patent 

 opened the road for many new developments and new devices 

 on the part of constructors of filters, so in describing the 



