36 BACTERIOLOGICAL EXAMINATION OF WATER. 



suddenly increasing it had no effect except on new coarse filters. 

 With gradually increasing age and consequent thickening of the 

 gelatinous layer in the filter, the rate of filtration, without a 

 dangerous head, became so low that cleaning of the surface 

 became inevitable. The depth of sand removed from the 

 surface, and the mode in which the filter was again put into 

 operation after cleaning, were found to be the principal factors 

 influencing bacterial efficiency. 



In Germany the gelatinous layer on the surface appears to be 

 regarded as essential, and the main body of the filter is simply 

 looked upon as an under-drain and support for the surface film, 

 having a steadying influence on the process of filtration during 

 times of abnormal stress. The Lawrence experiments also 

 showed that the surface film removed more bacteria than any 

 other layer in the filter. Still, in many cases, filters were found 

 to remove bacteria in the absence of any surface coating. Also 

 0*1 to 0*3 inch could be scraped off a continuous filter without 

 affecting its action, and coarse filters did not give perfect results 

 when the coating was so thick as to almost completely clog the 

 filter. 



From a practical point of view the influence of the removal of 

 clogging on bacterial efficiency depends on (a) the depth of the 

 material which is removed from the surface by scraping, and on 

 (&) the mechanical disturbance of the main body of the sand 

 which may be caused by the scraping and the process of refil- 

 ling. Until the scraping reaches about one inch the influence of 

 the removal is slight, but when more than one inch is removed 

 the change in the bacterial contents of the effluent becomes 

 marked. As the depth of material removed increases, not only 

 do the bacteria in the effluent increase in number, but also the 

 period of lessened efficiency of the filter becomes longer. The 

 most efficient method of putting filters into operation after 

 treatment to relieve clogging, has been carefully studied at 

 Lawrence. Filling from the top did not appear to affect small 

 experimental filters ; but on the large scale, in order to get the 

 upper layers of the filter dry enough to walk on, it was found 

 necessary to lower the water to a greater depth below the 

 surface, thereby furnishing an opportunity for mechanical 

 disturbance of the main body of the sand when its pores are 



