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 the same diameter, exactitude of form was of more importance than smoothness of surface; that glass pipes, which had a wavy surface, discharged less water, at the same inclinations, than Staffordshire stone-ware clay pipes, which were of perfectly exact construction. By passing pipes of the same clay--the common red clay--under a second pressure, obtained by a machine at an extra expense of about eighteen pence per thousand, whilst the pipe was half dry, very superior exactitude of form was obtained, and by means of this exactitude, and with nearly the same diameters, an increased discharge of water of one-fourth was effected within the same time." So all sudden turns or angles increase friction and retard velocity, and thus lessen the capacity of the drain--a topic which may be more properly considered under the head of the junction of drains. "On a large scale, it was found that when equal quantities of water were running direct, at a rate of 90 seconds, with a turn at right-angles, the discharge was only effected in 140 seconds; whilst, with a turn or junction with a gentle curve, the discharge was effected in 100 seconds." We are indebted to Messrs. Shedd & Edson for the following valuable tables showing the capacity of water-pipes, with the accompanying suggestions: "DISCHARGE OF WATER THROUGH PIPES.