MODERN METHODS OF WATER PURIFICATION. 329 



There is one very important experiment recorded in their 

 research, and that is, that when testing- the bacterial efficiency 

 of the filter they tried two lines of enquiry to note the effect 

 of (i) varying" the coagulant amount and (2) increasing the 

 speed of filtering". They found that, provided there was 

 adequate coagulant added, the speed could be pressed and the 

 filter still give good results, but that there was a point below 

 which, if the amount of coagulant was allowed to fall, a bad 

 effluent resulted, even if the filter was put at a lower rate of 

 work, so it was proved that the alum can not be allowed to fall 

 below a certain minimum without immediate deterioration of 

 quality. But still their recommended rate of filtration was 

 about I gallon per square foot per minute, namely, half the 

 amount of the American " normal." 



The control, after the Jewell filters were fully installed, was 

 at first constantly supervised by Dr. Gottschlich, and daily 

 bacteriological and weekly chemical tests made. When I was 

 there the control was in the hands of an employee of the 

 \\"aterworks Company, and consisted in the hourly examina- 

 tion for turbidity measured by Hazen's Platinum Wire method. 

 This wire is viewed through water in a long glass tube, and a 

 standard or scale of turbidity is made from the vanishing" point 

 of the wire fixed at right angles on the end of a glass rod. 

 In other words, as the bacteria and turbidity were found to be 

 removed pari passu, so turbidity was regarded as the measure, 

 and the turbidity kept at the practically imperceptible level 

 at which it was known no dangerous bacterial pollution could 

 be present. Careful records are made of the turbidity, coagu- 

 lant required, and loss of head on gauges. 



There seemed to be perfect confidence in this method of 

 control by turbidity estimation, and while satisfactory only a 

 weekly bacteriological examination is made by Professor 

 Gottschlich, on behalf of the Town Council of Alexandria, 

 whose Medical Officer of Health he is. Before adopting" the 

 Jewell filters the method tried was precipitation by perman- 

 ganate, followed by a slow sand bed filtration. I expected to 

 have found more frequent bacterial examinations, after the 

 manner of the daily (at times twice daily) bacterial examina- 

 tion of each separate sand filter efifluent and general efifluent 

 carried out at Berlin and Hamburg waterworks, which I had 

 visited. 



At Little Falls, X'.J.. the observations kept as detailed by 

 Geo. W. Fuller are samples of river water and clear water 

 basin collected three times daily, and examined for numbers of 

 bacteria, turbidity, amorphous matter, colour, alkalinity, etc. 

 Two-hourly examinations of the water for turbidity are also 

 made, and at least once each day a sample from the effluent 

 of each filter is examined bactcrially, and once a week as a 

 rule chemical analyses of raw and filtered water are made. At 

 frequent intervals careful tests are made regarding presence 

 of coagulant in filtered water. I had an opportunity of in- 

 specting the records kept, containing the amount of coagulant 

 applied, varying" sometimes from hour to hour: also records 

 of each filter as to an"iount filtered, period between washing 



