e
diameter, lying nearly level in the ground, in ten hours.

In ordinary aqueducts, for supplying water, and not for drainage, it is
desirable to have a high pressure upon the pipes to ensure a rapid flow;
but in drainage, a careful distinction must be made between velocity
induced by gravitation, and velocity induced by pressure. If induced by
the former merely, the pipe through which the water is swiftly running,
if not quite full, may still receive water at every joint, while, if the
velocity be induced by pressure, the pipe must be already full. It can
then receive no more, and must lose water at the joints, and wet the
land through which it passes, instead of draining it.

So that although we should find that the mains might carry a vast
quantity of water admitted by minor drains from high elevations, yet we
should bear in mind, that drains when full can perform no ordinary
office of drainage. If there is more than the pressure of four feet head
of water behind; the pipes, if they passed through a pond of water, at
four feet deep, must lose and not receive water at the joints.

The capacity of a pipe to convey water depends, then, not only on its
size, but on its inclination or fall--a pipe running down a considerable
descent having much greater capacity than one of the same size lying
nearly level. This fact should be borne in mind even in laying single
drains; for it is obvious that if the drain lie along a sandy plain,
for instance, extending down a springy hill-side, and then, as is
usually the case, along a lower plain again, to its outlet at some
stream, it may collect as much water as will fill it before it reaches
the lower level. Its stream rushes swiftly down the descent, and when it
reaches the plain, there is not sufficient fall to carry it away by its
natural gravitation. It will still rush onward to its outlet, urged by
the pressure from behind; but, with such pressure, it will, as we have
seen, instead of draining the land, suffuse it with water.

FRICTION,

as has already been suggested, is an element that much interferes with
exact calculations as to the relative capacity of water-pipes of various
dimensions, and this depends upon several circumstances, such as
smoothness, and exactness of form, and directness. The smoother, the
more regular in form, and the straighter the drain, the more water will
it convey. Thus, in some recent English experiments,

"it was found that, with pipes of