CHEMISTRY AND PHYSICS. 65 
to these stages, but Mr. Scott-Moncrieff has devised a system that more nearly 
causes a separation of the different bacteria than could be hoped for from these 
suggested schemes. 
In this the sewage is passed upward through a coarse filter from a shallow 
chamber underneath, thus forming right conditions for the anaerobics, which 
seem to be able to deal with all of the solid matter and prevent its accumulation. 
The overflow from this bed is discharged automatically and continuously onto a 
series of shallow trays, arranged each over the other, and each receiving its sup- 
ply from the discharge from the one next above. The automatic device works 
periodically, so that the materials of the trays are aerated regularly, the sewage 
trickling down through them and carrying air along with it. There is a rapid 
formation of nitrites in the two upper trays, but this decreases rapidly in those 
lower down, and runs out entirely in the eighth and ninth, the last of the series. 
The lower a tray the more active are the bacteria that produce nitrates, and in 
the lowest they seem to be the only organisms at work. The entire passage 
through the nine trays takes only about eight or ten minutes, and the rate of 
filtration is about one million gallons per acre per twenty-four hours. Mr. 
Rideal, an English expert, claims that this process gives better nitrate results 
than any other with which he is acquainted. 
In the accounts of this and other English work which I have seen the results 
are stated in percentages of nitrogen that have been removed, without any state- 
ments of the bacterial purity of the effluents secured. This can be inferred to 
some extent from the results of chemical analyses, and yet must be considered a 
serious lack to our full understanding of the matters presented. 
In attempting to use the Massachusetts methods for the purification of the 
waters of the region west of the Alleghany mountains, it has been found that they 
are not applicable. The fine, clayey matters that are carried in suspension in all 
western rivers— matter that is absent from the New England streams which flow 
from granite and sandstone hills—clog the pores of sand filters and prevent their 
use as bacteria beds. Some waters lack organic matter in the form needed to coat 
sand grains with a gelatinous envelope. Resort must be had to other methods. 
So there is now going on an active study of what is called the American system, 
or rapid mechanical filtration. The typhoid death-rate of Louisville, Cincin- 
nati, and Pittsburg, as well as other inland towns, has driven them to elaborate 
scientific investigations as to the best methods for treating their water-supplies 
in order to remove the fine silt and render them safe. 
The process is not biological, but a combination of chemical and mechanical 
methods. As worked out at Cincinnati, the river-water was first allowed to settle 
for some hours in an open reservoir, then pumped into tanks, where it was mixed 
with a precipitating agent, sulphate of alumina or crude alum, and thence into 
tanks containing about four feet of sand, through which it passed intoa clear well 
and into the city mains. The precipitate formed by the alum is flocculent, and 
entangles much of the organic matter from the very start. In the filter the alum 
is caught by the sand surface, where it forms a gelatinous coating that strains out 
both the fine silt and all suspended matter, including even such minute things as 
the bacteria. The bacterial efficiency rarely dropped below ninety-five per cent. 
and averaged about ninety-nine per cent.; and it was further found that when 
this was good the character of the effluent was also satisfactory as to clearness 
and the removal of organic matter. 
When the alum layer becomes too thick and offers too much resistance to the 
flow, the filter is washed by its own filtered water until the collected matter is 
—d 
