t adhere to the pipes in the same way as that found
near the surface. Arrangements should also be made for examining the
drains by means of wells, and for flushing them by holding back the
water until the drains are filled, and then letting it suddenly off, or,
by occasionally admitting a stream of water at the upper end, when
practicable, and thus washing out the pipes. Mr. Denton says: "It is
found that the use of this contrivance for flushing, will get rid of
the per-oxide of iron, about which so much complaint is made."

_Obstruction by Filling at the Joints._ One would suppose that tiles
might frequently be prevented from receiving water, by the filling up of
the crevices between them. If water poured on to tiles in a stream, it
would be likely to carry into these openings enough earthy matter to
fill them; but the whole theory of thorough-drainage rests upon the idea
of slow percolation--of the passage of water in the form of fine dew, as
it were--through the motionless particles which compose the soil; and,
if drains are properly laid, there can be no motion of particles of
earth, either into or towards the tiles. The water should soak through
the ground precisely as it does through a wet cloth.

In an article in the Journal of the Society of Arts, published in 1855,
Mr. Thomas Arkell states that in 1846 he had drained a few acres with
1-1/4 inch pipes, about three feet deep, and 21 to 25 feet apart. The
drains acted well, and the land was tolerably dry and healthy for the
first few years; but afterwards, in wet seasons, it was very wet, and
appeared full of water, like undrained land, although at the time all
the drains were running, but very slowly. His conclusion was that mud
had entered the crevices, and stopped the water out. He says he has
known other persons, who had used small pipes, who had suffered in the
same way. There are many persons still in England, who are so
apprehensive on this point, that they continue to use horse-shoe tiles,
or, as they are sometimes called, "tops and bottoms," which admit water
more freely along the joints.

The most skillful engineers, however, decidedly prefer round pipes, but
recommend that none smaller than one-and-a-half-inch be used, and prefer
two-inch to any smaller size. The circumference of a two-inch pipe is
not far from nine inches, while that of a one-inch pipe, of common
thickness, is about half that, so that the opening is twice as extensive
in the two-inch