re is no crack nor crevice through which water can trickle, and
then fill this hogshead to the top with earth, of the same character with
that used in the other case. These hogsheads should stand where the water
of a small roof, (as that of a hog-pen,) may be led into them, by an
arrangement which shall give an equal quantity to each;--this will give
them rather more than the simple rain-fall, but will leave them exposed to
the usual climatic changes of the season. A vessel, of a capacity of a
quart or more, should be connected with each outlet, and covered from the
dust,-- these will act as silt-basins. During the first few storms the
water will flow off much more freely from the first barrel; but, little by
little, the second one, as the water finds its way through the clay, and
as the occasional drying, and repeated filtration make it more porous,
will increase in its flow until it will, by the end of the season, or, at
latest, by the end of the second season, drain as well as the first, if,
indeed, that be not by this time somewhat obstructed with silt. The amount
of accumulation in the vessels at the outlet will show which process has
best kept back the silt, and the character of the deposit will show which
would most probably be carried off by the gentle flow of water in a nearly
level drain.

It is no argument against this experiment that its results cannot be
determined even in a year, for it is not pretended that drains laid in
compact clay will dry land so completely during the first month as those
which give more free access to the water; only that they will do so in a
comparatively short time; and that, as drainage is a work for all time,
(practically as lasting as the farm itself,) the importance of permanence
and good working for long years to come, is out of all proportion to that
of the temporary good results of one or two seasons, accompanied with
doubtful durability.

It has been argued that _surface water_ will be more readily removed by
drains having porous filling. Even if this were true to any important
degree,--which it is not,--it would be an argument against the plan, for the
remedy would be worse than the disease. If the water flow from the surface
down into the drain, it will not fail to carry dirt with it, and instead
of the clear water, which alone should rise into the tiles from below, we
should have a trickling flow from above, muddy with wasted manure and
silty earth.

_The remaining fi