5. _The manure washes off and is lost._ The three or four feet of water which the clouds annually give us in rain and snow, must either go off by evaporation, or by filtration, or run off upon the surface. Under the title of Rain and Evaporation, it will be seen that not much more than half this quantity goes off by evaporation, leaving a vast quantity to pass off through or upon the soil. If lands are ridged up, the manure and finer portions of the soil are, to a great extent, washed away into the open ditches and lost. Of the water which filters downwards, a large portion enters open ditches near the surface, before the fertilizing elements have been strained out; whereas, in covered drains of proper depth, the water is filtered through a mass of soil sufficiently deep to take from it the fertilizing substances, and discharge it, comparatively pure, from the field. In a paper by Prof. Way (11th Jour. Roy. Ag. Soc.), on "The Power of Soils to retain Manure," will be found interesting illustrations of the filtering qualities of different kinds of soil. In addition to the above reasons for preferring covered drains, it has been asserted by one of the most skillful drainers in the world (Mr. Parkes), "that a proper covered drain of the same depth as an open ditch, will drain a greater breadth of land than the ditch can effect. The sides of the ditch," he says, "become dried and plastered, and covered with vegetation; and even while they are free from vegetation, their absorptive power is inferior to the covered drain." Of the depth, direction, and distance of drains, our views will be found under the appropriate heads. They apply alike to open and covered drains. BRUSH DRAINS. Having a farm destitute of stones, before tiles were known among us, we made several experiments with covered drains filled with brush. Some of those drains operated well for eight or ten years; others caved in and became useless in three or four years, according to the condition of the soil. In a wet swamp a brush drain endures much longer than in sandy land, which is dry a part of the year, because the brush decays in dry land, but will prove nearly imperishable in land constantly wet. In a peat or muck swamp, we should expect that such drains, if carefully constructed, might last twenty years, but that in a sandy loam they would be quite unreliable for a single year. Our failure on upland with brush drains, has resulted, not from the decay of the wood, but from the entrance of sand, which obstructed the channel. Moles and field-mice find these drains the very day they are laid, and occupy them as permanent homes ever after.