be _right up and down the slope of
the land_, in the line of steepest descent. For a long time after the
general adoption of thorough-draining, there was much discussion of this
subject, and much variation in practice. The influence of the old rules
for making surface or "catch-water" drains lasted for a long time, and
there was a general tendency to make tile drains follow the same
directions. An important requirement of these was that they should not
take so steep an inclination as to have their bottoms cut out and their
banks undermined by the rapid flow of water, and that they should arrest
and carry away the water flowing down over the surface of hill sides. The
arguments for the line of steepest descent were, however, so clear, and
drains laid on that line were so universally successful in practice, that
it was long ago adopted by all,--save those novices who preferred to gain
their education in draining in the expensive school of their own
experience.

The more important reasons why this direction is the best are the
following: First, it is the quickest way to get the water off. Its natural
tendency is to run straight down the hill, and nothing is gained by
diverting it from this course. Second, if the drain runs obliquely down
the hill, the water will be likely to run out at the joints of the tile
and wet the ground below it; even if it do not, mainly, run past the drain
from above into the land below, instead of being forced into the tile.
Third, a drain lying obliquely across a hillside will not be able to draw
the water from below up the hill toward it, and the water of nearly the
whole interval will have to seek its outlet through the drain below it.
Fourth, drains running directly down the hill will tap any porous water
bearing strata, which may crop out, at regular intervals, and will thus
prevent the spewing out of the water at the surface, as it might do if
only oblique drains ran for a long distance just above or just below them.
Very steep, and very springy hill sides, sometimes require very frequent
drains to catch the water which has a tendency to flow to the surface;
this, however, rarely occurs.

In laying out a plan for draining land of a broken surface, which inclines
in different directions, it is impossible to make the drains follow the
line of steepest descent, and at the same time to have them all parallel,
and at uniform distances. In all such cases a compromise must be made
between the tw