urbed to a sufficient depth to prevent the
possibility of seepage from the surface. It is, of course, also quite
necessary that the well be of sufficient depth--the lower we go the more
likely are we to secure a perfectly pure water. In regions where the
water rises to within eight or ten feet, or less, of the surface, the
possibility of the well being contaminated during the rainy season by
seepage is considerably increased, and the waters of such wells should be
used only after analyses have shown that they are pure; where this cannot
be done, the water should be boiled before being drunk. Of course, the
possibilities of contamination are greatly increased if the locality be
thickly inhabited.

As has been before remarked, cisterns are more liable to contamination
from the air than are wells, chiefly owing to the fact that they are
supplied by water that is conducted into them by gutters from the tops of
houses. There is no question that during the dry seasons dust containing
many kinds of bacteria is deposited all over the tops of houses and
remains there until washed away by the rains. While it is true that the
sunlight quickly kills most germs that produce disease a certain number
of them would inevitably escape, and having gained entrance to a
cistern, would be likely to multiply and later cause trouble. It is thus
seen that however pure the rain-water may originally have been--and it is
among the purest of all waters--it is likely to become contaminated in
the process of collection, and may ultimately in this way become the
source of disease. Where any doubt exists as to the purity of such water
it should be boiled before use.

Surface-streams also occasionally supply drinking-water in rural
districts, and while the use of such waters may not always be attended by
danger, their contamination by disease-producing germs is much more to be
feared than when they are derived from wells or springs; where streams
arise from and keep their course through uninhabited districts the
probabilities are strong that their waters are pure and fit for use, but
where they run through cultivated fields, and particularly where they
pass in the neighborhood of houses, their waters should never be looked
upon as being drinkable,--except after being boiled or properly filtered.
Inasmuch as adequate filtration is exceedingly difficult to carry out,
and requires a somewhat extensive and costly plant, this is, as a rule,
not feasible for