PURIFICATION OF WATER FOR DRINKING PURPOSES 151 



examined by one of us, and the results are of interest 

 as exhibiting the manner in which defective filtration 

 is detected by the bacteriological examination. The 

 unfiltered water, which is of moorland origin, is derived 

 from two different reservoirs (Eowbank and Camp Hill) 

 of large size, and is submitted to filtration at an unusually 

 high velocity, viz. about five gallons per square foot 

 per hour. On the occasion in question all the unfiltered 

 water was coming from the Camp Hill reservoir. The 

 samples of filtered water were in every case taken 

 directly from the filter wells of the individual beds, and 

 the results obtained were as follows : 



Camp Hill Water (unfiltered} 

 Filter No. 1 j 



2 ,- Johnstone Works 



6 L Stanley Works 



7 

 8 



Number of Micro-organisms 

 in 1. c.c. 



. 3,414 

 . 181 

 . 282 

 . 197 

 . 74 

 . 324 

 . 566 

 . 112 

 42 



It is particularly worthy of note that in the case of 

 Filters Nos. 5, 6, and 7, the unfiltered water was visibly 

 leaking through the retaining walls of the filter beds 

 into the filter wells, and two of these, it will be seen, 

 exhibited a particularly large number of bacteria. 



If the above results be compared with those already 

 recorded for the London and Zurich waterworks, it will 

 be seen how greatly inferior is the efficiency of the 

 process of sand-filtration as practised at Paisley and 

 Konigsberg. 



These investigations clearly show that sand-filtration, 

 when carefully pursued, offers a most remarkable and 

 obstinate barrier to the passage of microbes ; and there is, 

 as already pointed out, every justification in presuming 



