Sewage Disposal and Rivers Pollution. 
351 
obtained when the filtering material was 5 feet of gravel and stones as 
large as beans. “ For nine months sewage pumped directly from the 
city sewer was applied nine times a day for six days in the week, in 
quantity equivalent to 81,400 gallons per acre per day. Here we 
find that 98 - 6 per cent, of the organic matter is removed by being 
burned and converted into nitrates, and more than 99 per cent, of 
the bacteria that were in the sewage were killed.” Even the rate 
of 126,000 gallons per acre per day was successfully maintained 
for three months, and the stones remained as clean as they were a 
year before. Only after two years’ working did the effluent become 
foul, when it was found that the air in the interstices had become 
impure from the amount of organic matter which had been retained. 
A layer of sand on the surface would perhaps have prevented this. 
The upper surface only of the sand filters became clogged after four 
years’ working. 
Now sewage is a difficult material to filter, and some descriptions 
will not filter at all. That such results as are recorded above have 
been obtained with crude or slightly strained sewage is therefore 
very encouraging. When it comes to filtering an effluent from a 
precipitation tank, of course vastly greater rates of filtration can be 
secured. We are told, for example, that a daily flow of 2| million 
gallons of sewage is treated at Mortlake by precipitation, and the 
effluent filtered through 1^ acre only of filter beds composed of 
layers of gravel, sand, and carbon, the surface being covered with a 
thin layer of earth sown with grass (=T6 million gallons per acre 
per day). This is about the ordinary rate of filtration of the London 
Water Companies’ filter beds. The rate claimed for some patent 
filtering materials is even twice as great as this, when used upon an 
effluent resulting from the use of a patent precipitant. 
The amount of purification by destruction of dissolved organic 
matter effected by filtration through porous materials was measured 
in the case of the Massachusetts experiments by the quantity of 
organic nitrogen that had disappeared by conversion into nitrate. 
Now, although this process of nitrification may not be an exact 
measure of the whole process of purification, it is doubtless a very 
important part of it. We know about it that it is effected by the 
agency of living organisms, that these are found in surface soil 
only, or, as in the case of the Massachusetts filters, they doubtless 
attach themselves to the surfaces of all the filtering particles which 
have free access to air during the intervals of filtration. Hence the 
supreme importance of porosity, and even coarseness, in the filter, 
and of intermittence in filtration. This seems to us to have a direct 
bearing on a remark of Dr. Dupre, who considers that turning a 
sewage effluent into a river is something like filtering it through 
land, or, to quote more exactly : — 
Here he might point out a very common mistake made by advocates of 
sewage-farming, namely, the comparison of the effluent from the farm with 
the effluent from the chemical precipitation works. The river into which 
the latter effluent was discharged must, in a measure, be looked upon as 
taking the place of the sewage farm, and he maintained that if the river was 
a a 2 
