View report page

236
day after it was started, 479 microbes per c.c. ; and eleven days
later it still contained 434 microbes or their spores per c.c."
" Thus it would appear that the thorough washing of this filter
had seriously impaired its quality for effecting bacterial
purification." Dr. Frankland gives particulars, on the
same lines, with regard to the other companies' waters, in
which, however, we are not locally interested. These
bacteriological investigations have been going on for twenty
months, and they are being supplemented by an extensive
series of experiments upon the conditions necessary for
effective microbial filtration. Whilst refraining for the present
from drawing any definite conclusions from the results contained
in the tables which supplement the remarks on each of
the companies waters, Dr. Frankland states that "efficient
filtration does not appear to depend, essentially, upon the
thickness of sand upon the filter beds .... A regular rate of
filtration throughout the 24 hours is probably of much more
importance than thickness of any filtering medium." " The
maximum rate to secure bacterial purity should not exceed 21
gallons per hour per square foot of filter surface." It is stated
that " the bacterial efficiency of sand filtration depends also, to
a very large extent, upon the formation of a thin film upon the
surface of the sand, derived from subsidences of suspended
particles in the raw water supplied to the filters. This film
forms a nidus for, and becomes thickly populated with,
microbes, which, in all probability, aid materially in the purification
of the water passing through the filter. This film
should therefore be removed as seldom as possible, consistent
with the passage of sufficient water through the filter." This
statement represents the newest, and a very novel view on the
subject, and will, at first, appear as surprising as the further
statement that " new filters and thoroughly washed filters are
very ineffective for a considerable time." It is suggested that
it would be well " to ascertain by direct experiments the
length of time required to bring a scraped or washed filter