In speaking of the forms of tiles, the superiority of rounded openings over those with flat bottom has been shown. The greater head of water in a round pipe, gives it force to drive before it all obstructions, and so tends to keep the drain clear. _Obstructions at the Outlet._ The water from deep drains is usually very clear, and cattle find the outlet a convenient place to drink at, and constantly tread up the soft ground there, and obstruct the flow of water. All earthy matter, and chemical solutions of iron, and the like, tend to accumulate by deposit at the outlet. Frogs and mice, and insects of many kinds, collect about such places, and creep into the drains. The action of frost in cold regions displaces the earth, and even masonry, if not well laid; and back-water, by flowing into the drains, hinders the free passage of water. All these causes tend to obstruct drains at the outlet. If once stopped there, the whole pipe becomes filled with stagnant water, which deposits all its earthy matter, and soon becomes obstructed at other points, and so becomes useless. The outlet must be rendered secure from all these dangers, at all seasons, by some such means as are suggested in the chapter on the Arrangement of Drains. _Obstruction by roots._ On the author's farm in Exeter, a wooden drain, to carry off waste water from a watering place, was laid, with a triangular opening of about four inches. This was found to be obstructed the second year after it was laid; and upon taking it up, it proved to be entirely filled for several feet, with willow roots, which grew like long, fine grass, thickly matted together, so as entirely to close the drain. There was a row of large willows about thirty feet distant, and as the drain was but about two feet deep, they found their way easily to it, and entering between the rough joints of the boards, not very carefully fitted, fattened on the spring water till they outgrew their new house. A neighbor says, he never wants a tree within ten rods of any land he desires to plow; and it would be unsafe to undertake to set limits to the extent of the roots of trees. "No crevice, however small," says a writer, "is proof against the entrance of the roots of water-loving trees." The behavior of roots is, however, very capricious in this matter; for, while occasional instances occur of drains being obstructed by them, it is a very common thing for drains to operate perfectly for indefinite periods, where they run through forests and orchards for long distances. They, however, who lay drains near to willows and ashes, and the like cold-water drinkers, must do it at the peril of which they are warned. Laying the tiles deep and with collars will afford the best security from all danger of this kind. Thos. Gisborne, Esq., in a note to the edition of his Essay on Drainage published in 1852, says: