DOMESTIC FILTRATION OF WATER 



71 



(composed of porcelain formed ty a mixture of kaolin and other 

 clays) were the only filters out of the substances named above which 

 were reliable and protective against bacteria. They tested over three 

 dozen of the Pasteur filters, and " in every case these gave a sterile 

 filtrate." Pure cholera bacillus in suspension (5000 bacilli to every 

 c.c.) and typhoid bacillus in suspension (8000 per c.c.) were passed 

 through these filters, and not a single bacillus was detectable in the 

 filtrate. The Berkefeld filter (siliceous earth) came second on the 

 list as an effective filter, and had but the one fault of not being a 

 "continuous" steriHser. A certain Parisian filter ("Porcelaine 

 d'Amiante"), made of unglazed porcelain, ren- 

 dered water absolutely free from bacteria. Its 

 action was, however, very slow. Setting aside 

 these three efficient filters, we are face to face 

 with the fact that most filters do not produce 

 germ -free filtrates, even though they are nomin- 

 ally guaranteed to do so. It is professed for 

 animal charcoal, which is widely used, that it 

 absorbs oxygen, and so fully oxidises whatever 

 passes through it. This may be so at first, but 

 after a little use it does more harm than good. 

 It appears to add nitrogen and phosphates to 

 water, which are both nutritive substances on 

 which bacteria grow, and it readily absorbs im- 

 purities from the air. As a matter of experiment 

 and practice, it has been found by Franldand, 

 Woodhead, and others, that charcoal actually 

 adds to the number of germs after it has been 

 in use for some time. 



Subsequent experiments were made in this 

 country by Lunt and Horrocks. Lunt working 

 in 1897 investigated the power of the Berkefeld 

 filter to intercept pathogenic bacteria, especially 

 the typhoid bacillus. He concluded as a result 

 of his inquiry that on the first day an efficiently sterilised Berkefeld 

 filter gave an absolutely sterile filtrate, but that on the second or 

 third day of using some water bacteria passed through. For thirty- 

 nine days the S. coli did not pass through, though the organism could 

 be detected on the outside of the filtering candle. Lunt found that 

 the action of the filter depended very much upon its method of 

 use : forcing or intermittently pumping water through the filter 

 resulted in a filtrate containing bacteria, whilst if the same filter was 

 steadily used a germ-free filtrate was obtained. In short, the result 

 of Lunt's work was to show the necessity for a frequent sterilisa- 

 tion of the filter, for though it allows ordinary water bacteria to 



Fig. 10. — Pastedr- 

 Chameebland Filter. 



Attached to Water 

 Supply. 



