60 Dr. G. Eeid. [May 7, 



was placed in the position it had previously occupied in the depth of the 

 filter, and seven days were allowed to elapse for recovery before collecting the 

 first samples. 



As regards analytical methods, the free ammonia was determined in the 

 usual way by distillation and Nesslerising, and the albuminoid ammonia by 

 Wanklyn's process, the quantities taken respectively being — sewage, detritus 

 tank, and septic tank, 50 c.c, filter effluents, 500 c.c, the former being 

 made up to 500 c.c. by ammonia-free water. The oxygen absorbed was 

 determined by Tidy's process, the nitrous nitrogen by Griess' method, and the 

 nitric nitrogen by Sprengel's method. 



In order to discount possible subsequent changes from the lower to the 

 higher oxide, and vice versa, the tests for nitrous and nitric nitrogen were 

 made at the works immediately on the collection of the samples, and again in 

 the laboratory the following day, when the complete analyses were made. 



The following figures represent the mean results of a series of analyses, the 

 individual figures of which are given in the detailed tables attached. The 

 samples were collected at intervals extending over a period of about six 

 months, the filters being steadily worked at the previous rate, namely, 

 200 gallons per superficial yard per 24 hours, and the delivery being uniform 

 and continuous day and night. 



In the first place, before commenting generally on the conclusions which 

 may be drawn from the results as a whole, I propose to refer shortly to 

 each analytical record separately : — 



1. With regard to the suspended solids it will be noticed that a reduction 

 of 73 per cent, is effected in the detritus tanks, and that a further reduction 

 of 15 per cent, takes place in the septic tank, making a total reduction of 

 88 per cent., and resulting in an effluent being passed on to the filter 

 containing 7*6 parts per 100,000, exactly one-half of which is mineral 

 matter. This suspended matter, it will be seen, is practically all retained 

 in the top layer of the filter, where the organic portion is liquefied, in all 

 probability, by aerobic organisms. The mineral matter, however, must 

 remain in the filter, and in time, no doubt, it will be found necessary to 

 remove the filtering medium to a depth of a few inches for the purpose of 

 washing it, but so far, after over three years' constant working, no such 

 necessity has arisen. As a matter of fact, if the total mineral suspended 

 solids passing on to the filter during the three years were deposited in a 

 uniform layer over the whole surface, the depth of the coating would be less 

 than 1J inch. 



2. The reduction of the free ammonia at a depth of 1 foot is remarkable, 

 especially considering the fact that the change has been effected in about 



