ACTION AND UTILITY OF FILTERS. 277 



expressed as parts per 100,000 : Total solids, 44 parts ; volatile 

 solids, 16 parts ; chlorine, 6'2 parts ; free ammonia, 0'526 part ; 

 albuminoid ammonia, 0'60 part ; oxygen absorbed in four hours, 

 2'2 parts; nitrites, absent; nitrates (NO 3 ), 1'005 part. The 

 water in the mantle was inoculated with a loopful of a twenty-four 

 hours agar culture of B. typhosus, and filtration carried on daily 

 as in the previous experiments. The broth in the flask remained 

 perfectly clear for eight days, but on the ninth day it was found 

 to be turbid. A loopful of the turbid broth, when examined as 

 before, was found to contain a pure culture of B. typhosus. In 

 this experiment the nutriment supplied was greater than in 

 Experiment III., but not so great as in Experiments I and II. 

 Corresponding apparently to the nutriment supplied, the time 

 taken by the typhoid bacillus to grow through the candle was 

 eight days as compared with four days in Experiments I. and 

 II., and nine days in Experiment III. 



Experiment V. In this experiment the mantle was filled with 

 a specimen of sterile barrack sewage, which contained a large 

 quantity of organic matter. The results of the analysis expressed 

 as parts per 100,000 were as follows : Total solids, 55 parts ; 

 volatile solids, 15 parts ; chlorine, 9*5 parts ; free ammonia, 

 8'28 parts ; albuminoid ammonia, 5*07 parts ; oxygen absorbed 

 in four hours, 6'4 parts ; nitrites, absent ; nitrates, absent. The 

 sewage was inoculated with a twenty-four hours agar culture of 

 B. typhosus, and filtration carried on in the same manner as in 

 the previous experiments. The broth in the flask remained 

 quite clear for four days, but was found turbid on the fifth day. 

 The turbidity was caused by B. typhosus. In this experiment 

 the available organic matter was much greater than in Experi- 

 ments III. and IV., and consequently the bacillus only required 

 four days to grow through the walls of the candle. 



Experiment VI. The same form of apparatus was used, a 

 new No. 12 Berkefeld candle being fitted into the mantle. A 

 specimen of water collected from a reservoir at Netley was 

 sterilised and then analysed. The following results, expressed as 

 parts per 100,000, were obtained : Total solids, 18 parts ; volatile 

 solids, 5 parts ; chlorine, 3'3 parts ; free ammonia, 0'00816 

 part; albuminoid ammonia, 0*0158 part; oxygen absorbed in 

 four hours, 0'5 part ; nitrites, absent ; nitrates (NO 3 ), 0*8 part. 



