148 THE AMERICAN MONTHLY [ August, 
that by agitating sterilized water with air no purification resulted, but if the 
same water be mixed with ordinary water containing microbes it became 
purified by contact with air. He concludes that self-purification of water is 
impossible without microbes, and that direct oxidation is not possible. 
The reader is already acquainted with this subject through the valuable 
article published last year in these columns, written by Dr. T. Smith.* 
In looking over the literature at hand we have found several valuable con- 
tributions in Zhe Chemical News of the year 1885. The methods of obser- 
vation are practically the same everywhere. Dr. Koch’s method of plate 
culture on sterilized gelatin being universally adopted, sometimes supple- 
mented by potato cultures and growths on other media. Among the most 
important contributions is that of Percy F. Frankland, Ph.D., F. C. S., etc., 
entitled The Removal of Micro-organisms from Water.f In this article the 
effect of several methods of purification by filtration, subsidence, and pre- 
cipitation, is shown by plate cultures. 
The substances selected for filters were green sand, silver sand, powdered 
glass, brick dust, coke, animal charcoal, and spongy iron, all being passed 
through a sieve of forty meshes to the inch. The filters were tubes of glass, 
one inch in diameter, drawn out at the bottom to a small aperture. A small 
piece of asbestos was first placed in the tube and above it the filtering ma- 
terial to a depth of six inches. 
The results of the experiments are of interest, but they can only be briefly 
summarized here. Green sand at first completely removed the organisms 
from stale urine water. After thirteen days, during which 7.1 litres had 
passed through, the filtered water contained. 1,071 centres of growth, the un- 
filtered. 8,193, thus showing some efficiency, and even after an entire month 
it continued to remove a large proportion of the organisms. 
Animal charcoal removed all the organisms from a very bad water after 
twelve days of continuous action, 4.2 litres having passed through. At the 
end of one month 14.6 litres had passed, but not only had the filter lost its 
efficiency, but it was found to contaminate the water and greatly increase the 
number of organisms in it. The unfiltered water contained 1,281 and the 
filtered water 6,958 centres of growth per cubic centimétre. 
Spongy iron removed the organisms entirely for twelve days (3.6 litres), 
and after one month, when g litres had been filtered, only two centres of 
growth were observed in the filtrate, against 1,280 in the original water. 
Brick dust is not efiicient, even when fresh. 
Coke is next to spongy iron in efficiency, and a more rapid filter. After 
five weeks 86 centres of growth were observed in the filtrate against 5,932 in 
the unfiltered water. 
Sand, freed from iron by washing with hydrochloric acid, filtered very rap- 
idly ; but it does not, even at first, remove all the organisms. It diminishes 
their number, however, 11,232 being reduced to 1,012 per cubic centimetre. 
Powdered glass acts very much like sand. 
Another series of experiments showed that by merely shaking up the water 
with the filtering materials above mentioned, and allowing them to subside, 
a decided reduction in the number of organisms will sometimes be effected. 
Thus, with spongy iron, 609 centres were reduced to 28; with chalk, 8,325 
to 274; animal charcoal, 8,325 to 60; while coke reduced them from a num- 
ber too great to be counted to none at all. Clay and brick dust have no no- 
ticeable eftect. 
It has been presumed that subsidence would lead to the separation of living 
* Notes on the Biological Examination of Drinking Water, with a few Statistics of Potomac Drinking Water. 
This Yournad, vii (1886), 61. 
t Chem. News. Lii (1885), 27, 40. 
