WATPJR FILTRATION. 



XXXIX. 



" During the month of January lower quantities of coagu- 

 lant were used intentionally with the view of seeing the 

 efficiency of different ratios, and, with the uniform rate of 

 application of coagulant, the bacteria results obtained 

 during that month are of most value for this particular 

 purpose. 



Removal of colour. — " It has been found as a result of 

 repeated examinations, especially in noting the appearance 

 of the filtered water in a large porcelain-lined bath-tub, 

 that it is not possible for a consumer to recognise any 

 colour when the filtered water contains 10 parts per million 

 or less. A colour of 20 parts is also practically unnotice- 

 able to the consumer. These works have been operated 

 with the view of keeping the colour below 10 parts. It 

 has been found, however, that the quantity of coagulant 

 necessary to reduce it to that point is practically as great 

 as that required to reduce it to 5 parts." 



The following table shows the monthly average results 

 of the applied coagulant, and the colour in the raw and 

 filtered water : — 



Month 



Sulphate of Alumina 



Colour 



Grains per Gallon 



River Water 



Filtered Water 



1902 









September... 



074 



31 



11 



October 



1-59 



52 



7 



November ... 



1-63 



45 



7 



December ... 



1-70 



44 



5 



1903 









January 



0-84 



31 5 



Removal of turbidity. — "The turbidity of the filtered 

 water usually ranges from nothing to 2 parts. Now and 

 then it becomes as high as 3 parts. All these amounts 

 may be properly regarded as traces, and it may be said 

 that for all practical purposes the filtered water has no 

 turbidity. 



Removal of organic matter. — " The following table shows 

 the average removal of organic matter: — 



