CHEMISTRY AND PHYSICS. 63 
crop or a remarkable product, this is now rightly considered as a happy result, 
but an entirely secondary one. 
Chemical precipitation has been tried with all sorts of substances as precipi- 
tants, but securing only a partial purification, which amounts to saying that the 
effluent is still impure, and, therefore, unsafe. Where such a degree of treat- 
ment is wanted as will insure a stream from becoming offensive, and where this 
is not used for a source of supply, the method may find an application. An 
electrical treatment results chiefly in the formation of an antiseptic, which is 
supposed to act directly on germ life, and also of a precipitant set free by the 
current. It fails of complete action. 
Broad irrigation results in a water-logged soil that neither benefits the crop or 
ereates a good effluent. 
Intermittent irrigation, rightly managed, and where a sufficient area of suita- 
ble open, sandy soil can be obtained, solves the problem quite well. It is, how- 
ever, of limited application, because of inability to find the right kind of land in 
necessary quantity within easy reach of many towns. The method has been in 
use for some years, but the soil’s action was not understood, or very imperfectly 
so, until the state board of health of Massachusetts, through its experiment sta- 
tion, at Lawrence, and the high skill and great ability of its experts there at 
work, discovered the matter and gave their results to the world. They first 
brought the slow sand-filtration process into line with exact science, whether ap- 
plied to sewage treatment or the purification of water. 
In this process the water or sewage is slowly filtered through a bed of sand 
about four feet thick. The filtering material may be of almost any average size 
of grain, from fine sand to coarse gravel, the rate of filtration and the size of the 
applied dose being regulated and properly adapted to each grade. In the upper 
few inches of this bed all of the chemical changes take place by means of the 
bacteria that there become established. These are very largely of the aerobic 
type, so that the filter has to be worked intermittently; that is, allowed to drain 
itself for a time between applications of doses, so as to draw into the sand the air 
that is needed by the bacteria for their life work. After these bacteria have be- 
come well established, which takes some time, the results are excellent. The 
experimental filters gave effluents in which nearly all of the nitrogen was changed 
to nitrates, and from which practically all of the organisms that were in the ap- 
plied water or sewage were excluded; even such bacteria as did appear probably 
coming from the colonies that had established themselves in the underdrains, 
and not directly derived from the filter. 
It is to be observed that these filters call into action only the aerobic organ- 
isms, and that, therefore, in order to attain a maximum efficiency, the first stage 
of the nitrogen change, due to the action of anaerobic forms, must be brought 
about before the application of the liquid to the filter; that is, the solids must 
have been broken up and the organic matters rendered soluble. Asa matter of 
fact, both forms exist in these filters, or, rather, there are forms present that can 
act in both ways, as anaerobic when the filter is saturated, and as aerobic when 
working in the presence of oxygen. In case the filters are treating a sewage that 
comes from a long line of sewers, the first stage has been passed to a large de- 
gree by the time the filters are reached, the anaerobics doing their work in the 
sewers, leaving only a small part of the organic nitrogen to the action of the 
forms that can adapt themselves. Also in the case of the filtration of river-water, 
much of the work of this first stage has been already accomplished. Hence, 
good results can be expected. But if a fresh, raw sewage is to be treated, these 
filters will give trouble, through the deposit of unbroken-down solids that quickly 
