as we
assume, run down the slope, the mains must run at the foot of the slope
and across it.

It will be found in practice, that all the circumstances alluded to,
will combine to locate the mains across the foot of regular slopes; and
whether in straight or curved lines, along through the natural valleys
of the field.

In locating the mains, regard must always be had to the quantity of
water and to the fall. Where a field is of regular slope, and the
descent very slight, it will be necessary, in order to gain for the main
the requisite fall, to run it diagonally across the bottom of the slope,
thus taking into it a portion of the fall of the slope. If the fall
requires to be still more increased, often the main may be deepened
towards the outlet, so as to gain fall sufficient, even on level ground.

If the fall is very slight, the size of the main may be made to
compensate in part for want of fall, for it will not be forgotten, that
the capacity of a pipe to convey water depends much on the velocity of
the current, and the velocity increases in proportion to the fall. If
the fall and consequent velocity be small, the water will require a
larger drain to carry it freely along. The size of the mains should be
sufficient to convey, with such fall as is attainable, the greatest
quantity of water that may ever be expected to reach them. Beyond this,
an increase of size is rather a disadvantage than otherwise, because a
small flow of water runs with more velocity when compressed in a narrow
channel, than when broadly spread, and so has more power to force its
way, and carry before it obstructing substances.

We have seen, in considering the size of tiles, that in laying the minor
drains, their capacity to carry all the water that may reach them is not
the only limit of their size. A one-inch tile might in many cases be
sufficient to conduct the water; but the best drainers, after much
controversy on the point, now all agree that this is a size too small
for prudent use, because so small an opening is liable to be obstructed
by a very slight deposit from the water, or by a slight displacement,
and because the joints furnish small space for the admission of water.

Mains, however, being designed merely to carry off such water as they
may receive from other drains, may in general be limited to the size
sufficient to convey such water, at the greatest flow. It might seem a
natural course, to proportion the capacity of the m