92 THE AMERICAN MONTHLY [March 



Under tliese conditions it was determined how long 

 after the commencement of the run the filtering ability 

 was a;t a maxmium and also the capacity of the filtering 

 media to remove organisms and also the possibilty of re- 

 moving organisms foreign to river water and simulating 

 pathogenic bacteria in their life history. In this last 

 experiment the Cruikshank bacillus and bacillus prodigi- 

 osus were used, since from their pathogenic properties 

 they could be readily distinguished from the water 

 bacteria. 



For an understanding of the proportion of bacteria 

 found in the applied water and the number to be found 

 in the filter water, table No. 3 of the report is here ap- 

 pended. 



As a result of the whole series of experiments the totals 

 shown in table No. 9 will give an idea of the averages. 

 In consideration of this table, it must be remembered 

 that the introduction of only one result, which may be 

 far below the average, will readily reduce what would 

 otherwise be a most favorable average, to a lower point. 

 This one result might occur from a temporary contamin- 

 ation of the eflluent pipes at the time of collecting the 

 sample, and which might not represent the exact results 

 of filtration. 



During the application of the cultures of bacillus pro- 

 digiosus in large quantities suspended in the nutrient 

 media, the numbers of the common water bacteria mater- 

 ially increased in the effluent, the particles of quartz be- 

 coming covered with a slimy brownish deposit. Unsuc- 

 cessful efforts were made to cleanse the quartz of this 

 growth by steaming and boiling the quartz for one hour. 

 Finally on the application of a solution of one pint of 

 caustic soda to twenty-four parts of water and steaming, 

 the normal white color of the quartz returned. The effi- 

 ciency of the filter was raised by this process of cleans- 



