that is quite a different matter from draining the
whole adjoining region, and requires a different mode of operation. If
your field is in the middle, or at the foot, of an undrained slope, from
which the water runs on the surface over your land, or soaks through it
toward some stream or swamp below, provision must be made not only for
drainage of your own field, but also for partial drainage of your
neighbor's above, or at least for defence against his surplus of water.

The first, and leading idea to be kept in mind, as governing this
question of the direction of drains, is the simple fact that _water runs
down hill_; or, to express the fact more scientifically, water
constantly seeks a lower level by the force of gravitation, and the
whole object of drains is to open lower and still lower passages, into
which the water may fall lower and lower until it is discharged from our
field at a safe depth.

Water goes down, then, by its own weight, unless there is something
through which it cannot readily pass, to bring it out at the surface. It
will go into the drains, only because they are lower than the land
drained. It will never go _upward_ to find a drain, and it will go
toward a drain the more readily, in proportion as the descent is more
steep toward it.

To decide properly what direction a drain should have, it is necessary,
then, to have a definite and a correct idea as to what office the drain
is to perform, what water is to fall into it, what land it is to drain.

Suppose the general plan to be, to lay drains forty feet apart, and four
feet deep over the field, and the question now to be determined, as to
the _direction_, whether across, or up and down the slope, there being
fall enough to render either course practicable. The first point of
inquiry is, what is expected of each drain? How much and what land
should it drain? The general answer must be, forty feet breadth, either
up and down the slope, or across it; according to the direction. But we
must be more definite in our inquiry than even this. From _what_ forty
feet of land will the water fall into the drain? Obviously, from some
land in which the water is higher than the bottom of the drain.

If, then, the drain run directly _across_ the slope, most of the water
that can fall into it, must come from the forty feet breadth of land
between the drain in question, and the drain next above it. If the water
were falling on an impervious surface, it would al