PURIFICATION OF WATER FOR DRINKING PURPOSES 159' 



entire mass, although the actual discoloration of the 

 sand is only apparent in the first, or at most second, cm. 



In a small experimental filter constructed by Piefke 

 with sterilised waterworks sand, and set to work in the 

 usual way, it was found that the power of retaining 

 microbes was nil^ that in fact more organisms were 

 often found in the filtrate than in the unfiltered water, 

 multiplication having taken place during passage through 

 the filter. These experiments with sterilised sand thus 

 clearly demonstrate that it is the slime deposit on the 

 sand which constitutes the real filtering material in the 

 waterworks filter. 



It used to be a doctrine very generally enunciated 

 by the German school of investigators, that these ordi- 

 nary sand filters were endowed with the almost miracu- 

 lous power of arresting all micro-organisms in the 

 water passing through them, the presence of the microbes 

 actually found in the filtered water being attributed 

 entirely to post-filtration sources. Now, although in 

 some individual filter beds, when the rate of filtration 

 is well chosen and carefully regulated, the number of 

 microbes is actually reduced almost to nil^ as seen from 

 the investigations conducted by one of us on the indi- 

 vidual London water companies' filter beds ; yet we 

 have from the very first pointed out that these sand 

 filters can only lay claim to relative and not to absolute 

 efficiency, and the experiments we have been consider- 

 ing abundantly prove the correctness of our conclu- 

 sions and deductions, which are now shared by our 

 neighbours on the Continent,-^ 



Purification of Sewage by Sand-Filtration. — We 

 have so far only considered the purification of water by 



^ ' The Hygienic Value of the Bacteriological Examination of Water/ 

 Percy Frankland. Transactions of the International Congress of Hy- 

 giene and Demography^ vol. ii. p. 291. London, 1891. 



