■464 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



The disadvantage of this form of nitration is that the cost of con- 

 struction on a large scale is practically prohibitive, and it only quite 

 recently occurred to me that, although this is true, nevertheless an 

 apparatus of quite moderate dimensions is sufficient for the purposes 

 of horticulture, and is capable of supplying an almost perfect plant food — 

 say, up to 150 gallons each twenty-four hours at an altogether nominal 

 working expenditure. 



It is on the strength of these facts that I ventured to offer a paper for 

 the acceptance of this Society, and I shall now give a more detailed 

 description of the apparatus and say something of the chemical trans- 

 formations which it is capable of producing. 



It will be well to say something in the first place about the general 

 principles that are involved in a solution of the problem of converting 

 the organic substances of sewage into their mineral equivalents by the 

 aid of the bacterial action of a percolating filter. It is now many years 

 since I laid down the proposition that, excluding temperature, which in 

 practice is a question of climate, there are four, and only four, factors to be 

 dealt with. These are (1) the rate of flow, or, in other words, the quantity 

 per unit of time that should be discharged on the filter ; (2) the depth 

 of filter required to give any desired standard of purification ; (3) the 

 period of rest between each discharge ; and (4) the amount of air necessary 

 to enable the nitrifying organisms to carry out their work to the greatest 

 advantage. So far as I know this statement has never been contra- 

 dicted, and my contention is that in every case there should be an exact 

 estimation of these factors in order to know how the best results can 

 be obtained. Of course in comparing the results of identical conditions 

 upon different liquids, the filtering materials must also be identical. 



Dealing with the first of the four factors, in order to discover the best 

 rate of flow, it is necessary to make estimations by error ; but in practice 

 it has been found that, in dealing with ordinary domestic sewage, with 

 a well- constructed filter, the rate of flow may be taken at an average of 

 about 1,000,000 gallons to each acre in every twenty-four hours, or at the 

 rate of one gallon to each foot each hour. It has also been shown by the 

 installation of these bacterial trays at Caterham that the rate of flow for 

 any given standard of purification must be inversely as the amount of 

 organic nitrogen contained in the sewage itself — that is to say, for a 

 sewage of double the strength the rate of flow would have to be reduced 

 by ;"30 per cent., of three times the strength by 66 per cent., and so on 

 proportionately in every case. 



With regard to the second factor — that of filtering depth required for 

 any given surface area and rate of flow — it is obvious that in the case of 

 the trays this speaks for itself, because, if the filter is made deep enough 

 to begin with in the testing experiment, it is only necessary to select 

 the tray that gives the required standard of nitrification. To obtain the 

 highest possible the flow must be reduced until the highest is produced. 



The third factor — that of the best period of rest between each 

 discharge — is a very important one. When it is realized that time is 

 a necessary element of the process it is obvious that, if the organic 

 matter is delivered to the organisms more quickly than they can deal 

 with it, or under unfavourable conditions, then there must be an 



