Water should enter at the bottom of the tiles, and not at the top. It is a well-known fact in draining, that the deepest drain flows first and longest. A familiar illustration will make this point evident. If a cask or deep box be filled with sand, with one hole near the bottom and another half way to the top, these holes will represent the tiles in a drain. If water be poured into the sand, it will pass downward to the bottom of the vessel, and will not flow out of either hole till the sand be saturated up to the lower hole, and then it will flow out there. If, now, water be poured in faster than the lower hole can discharge it, the vessel will be filled higher, till it will run out at both holes. It is manifest, however, that it will first cease to flow from the upper orifice. There is in the soil a line of water, called the "water-line," or "water-table;" and this, in drained land, is at about the level of the bottom of the tiles. As the rain falls it descends, as in the vessel; and as the water rises, it enters the tiles at the bottom, and never at the top, unless there is more than can pass out of the soil by the lower openings (the crevices and pores) into the tiles. It is well always to interrupt the direct descent of water by percolation from the surface to the top of the tiles, because, in passing so short a distance in the soil, the water is not sufficiently filtered, especially in soil so recently disturbed, but is likely to carry with it not only valuable elements of fertility, but also particles of sand, which may obstruct the drain. This is prevented by placing above the tiles (after they are covered a few inches with gravel, sand, or other porous soil) compact clay, if convenient. If not, a furrow each side of the drain, or a heaping-up of the soil over the drain, when finished, will turn aside the surface-water, and prevent such injury. In the estimates as to the area of the openings between pipes, it should be considered that the spaces between the pipes are not, in fact, clean openings of one-tenth of an inch, but are partially closed by earthy particles, and that water enters them by no means as rapidly as it would enter the clean pipes before they are covered. Although the rain-fall in England is much less in quantity and much more regular than in this country, yet it is believed that the use of two-inch pipes will be found abundantly sufficient for the admission and conveyance of any quantity of water that it may be necessary to carry off by drainage in common soils. In extraordinary cases, as where the land drained is a swamp, or reservoir for water which falls on the hills around, larger pipes must be used.