The stoppage was nevertheless repeated, and was then traced to an elm tree far more distant than those which had been sacrificed. Early in the autumn of 1850 I completed the drainage of the upper part of a boggy valley, lying, with ramifications, at the foot of marly banks. The main drains converge to a common outlet, to which are brought one 3-inch pipe and three of 4 inches each. They lie side by side, and water flows perennially through each of them. Near to this outlet did grow a red willow. In February, 1852, I found the water breaking out to the surface of the ground about 10 yards above the outlet, and was at no loss for the cause, as the roots of the red willow showed themselves at the orifice of the 3-inch and of two of the 4-inch pipes. On examination I found that a root had entered a joint between two 3-inch pipes, and had traveled 5 yards to the mouth of the drain, and 9 yards up the stream, forming a continuous length of 14 yards. The root which first entered had attained about the size of a lady’s little finger; and its ramifications consisted of very fine and almost silky fibres, and would have cut up into half a dozen comfortable boas. The drain was completely stopped. The pipes were not in any degree displaced. Roots from the same willow had passed over the 3-inch pipes, and had entered and entirely stopped the first 4-inch drain, and had partially stopped the second. At a distance of about 50 yards a black Italian poplar, which stood on a bank over a 4-inch drain, had completely stopped it with a bunch of roots. The whole of this had been the work of less than 18 months, including the depth of two winters. A 3-inch branch of the same system runs through a little group of black poplars. This drain conveys a full stream in plashes of wet, and some water generally through the winter months, but has not a perennial flow. I have perceived no indication that roots have interfered with this drain. I draw no general conclusions from these few facts, but they may assist those who have more extensive experience in drawing some, which may be of use to drainers." Having considered some of the principles on which our work should be based, let us now return to the map of the field, and apply those principles in planning the work to be done to make it dry. *The Outlet* should evidently be placed at the present point of exit of the brook which runs from the springs, collects the water of the open ditches, and spreads over the flat in the southwest corner of the tract, converting it into a swamp. Suppose that, by going some distance into the next field, we can secure an outlet of 3 feet and 9 inches (3.75) below the level of the swamp, and that we decide to allow 3 inches drop between the bottom of the tile at that point, and the reduced level of the brook to secure the drain against the accumulation of sand, which might result from back water in time of heavy rain. This fixes the depth of drain at the outlet at 3-1/2 (3.50) feet.