20 in. at top, and about 18 in. deep, which is a good size for the minor or sub-drains. Covered drains may be formed in grass-land 6 in. wide in the bed, 18 at top, and about 16 in. deep, at 4d. per rood, by cutting out the upper turf, the whole width across the cut, with the spade, casting out the lower portions subsequently, and then carefully replacing the turf, thus leaving an open space below, equal to the quantity cast out. The permanence of this drain is, however, very insecure; and, if cattle are admitted on the surface, they will certainly tread the turf to the bed of the drain, and thus destroy it. The first cost will, moreover, nearly equal that of a pipe drain which needs a much narrower cut, and will remain permanently efficient.

116. In land to be planted with forest trees, open drains are always to be recommended, as covered ones are certain of destruction by the natural tendency of the roots of the trees to choke them in their search for moisture during dry seasons. The main drains should be laid along the hollows in the surface, and made at least 3 ft. deep, with a flat bed 1 ft. wide, and the banks inclined at the rate of 1 base to 1 perpendicular, except in firm clay soils, in which the banks may be formed much steeper. The minor drains should be for clays not less than 20 in. deep, and light soils 14 in., with a bed in both cases 9 in. in width, and the inclination of the banks regulated as for the mains. The cost of the former will be about 1d. per rood, and the latter 3d. per rood. The cost of the mains will be in nearly the same proportion, according to the quantity of soil removed. The best distance at which to lay the minor drains from each other will vary in extreme cases from 5 to 40 yards, according to

the levels of the site and character of the soil, the retentive clays requiring the drains closer than the lighter soils.

117. A general and most important principle as to the capacity of drains of all kinds whatsoever is, that it should exceed rather than be deficient of the dimensions ordinarily required to discharge the quantity of water for which provision is to be made. The principal use of a drain being to attract water towards it through the soil, besides passing the water thus collected away, its dimensions cannot be adequately estimated by simply considering the quantity to be conveyed within any given time. These dimensions should, therefore, be such as to present large surfaces of the soil intersected, and, other circumstances being the same, the efficiency of the drain will be in proportion to the extent of the surfaces, that is, to the depth of the drain. But, on the other hand, if the greater depth of the drain causes it to intersect porous strata overcharged with water from higher land, it will become injurious rather than beneficial, and this evil will be much aggravated if the greater depth be admitted as a reason for the proportionate infrequency of the drains. There can be no doubt that, in tenacious soils, shallow drains laid closely are, within certain limits, more useful than deep drains laid wide apart; but, if contiguity can be observed, the deeper they are made the better, in ordinary cases.

118. Some reasons to guide the depth of drains may be derived from a consideration of the action of the soil upon the water which reaches it, as produced by its mechanical structure. Thus, in light and porous soils, the force of gravity is active in carrying the water to the bottom of the stratum; whereas, in the dense clays and soils, a certain capillary action is exercised upon the

water introduced to them, which tends to raise it from the bed, and sustain it in general diffusion throughout the mass. Therefore, while porous soils evince little or no water on the surface, the lower part of the layer will be kept in a state of excessive wetness if it lies upon a clay bed; and, if its thickness be such that the roots of the vegetation reach the wet, the depth of the drains should at least equal that of the porous soil, so that the entire body may be relieved of the water. On soils of this nature shallow drains are utterly useless, unless they happen to reach an impervious subsoil, and conduct the water into mains of greater depth.

119. In arable land, the minimum depth for covered drains may be estimated upon the depth to which the plough penetrates, and making such an allowance below this depth as will secure the materials of the drain from disturbance under any circumstances. Mr. Stephens calculates the depth of a furrow-slice with a two-horse plough at 7 in.; but, in cross ploughing, 9 in. If four horses be used, the depth of the furrow will be 12 in.; and, if the four-horse plough follow the common one, the depth will be increased to 16 in. Subsoil ploughing will penetrate 16 in. below the common furrow of 7 in. Allowing 3 in. between the lowest disturbed part of the soil and the surface of the materials in the drain, and restricting the effectiveness of the drain to that portion of it which is below the ploughed surface of 7 in. in depth, the minimum depth of drains should be such as to allow 19 in. below the furrow slice, or 26 in. below the surface and above the constructed portion of the drain, and so much more than this if subsoil ploughing be practised. Allowing 6 in. for the depth of the drain occupied by the pipes or tiles, Mr. Stephens estimates 33 in. as the minimum depth of drains in porous subsoils, and 50 in.

in clay subsoils, with an additional 6 in. in each case if stones are employed as filling materials in the drain.

120. The size for the water-passage or duct of a drain should be determined by reference to a variety of circumstances, the combined influence of which may generally be estimated in practice, although not reducible to any very exact rules. Thus, the quantity of rain which falls upon the surface has to be considered, not as an annual or season quantity, but as a maximum per diem. Then, the nature of the soil and the state of the atmosphere, as affecting the ratio of evaporation, require attention. Beyond these considerations, the general level of the district in relation to the surrounding country, by which the tract to be drained may be made the recipient of foreign waters, on the one hand, or kept in a dry condition by the action of gravity, on the other, must be noticed. Again, the structure of the soil affects the quantity of the water which passes through it, and, also, the rapidity of its passage; and the amount of water to be met with will be modified by the part of the stratum at which the drain is situated. Thus, in porous materials, smaller ducts will suffice in the top of the layer than are required below; and the dimensions must be increased in proportion to the depth of soil above. As an auxiliary fact in enabling us to determine the capacity of the ducts of drains, the frequency of them upon the plan of the district will be greatly influential.

121. The best evidence on these points, viz., the dimensions and distance of drains, is to be gathered from the records of extended practice. In the wealdclay of Kent, which is commonly of a very tenacious character on the surface, but milder below, the body of the water naturally passes downwards until arrested by

a more retentive stratum, and, therefore, the deeper the drain the more efficiently they will act. In other parts of the weald the soil is compounded of the supersoil or cultivated earth and of a strong clay, upon which it lies. This soil admits of percolation; but the tenacious clay beneath does not, and, if this clay be at a considerable depth from the surface, there will be little utility in carrying the drains into it. In these strong clays, not subject to springs, drains 2 ft. deep have been found more efficacious than those made 4 ft. deep. In the heavy lands of Norfolk, the drains which answer best are 2 ft. deep, and laid at the distance of 22 ft. apart. When they are made deeper, in clay in which flint and chalk boulders are found dispersed about, the expense of taking out the lower bed of 16 or 18 in. is very expensive, costing in that county from 6 to 8 pence per rod of 5 yards. In the clay lands of Hampshire, the drains made from 30 to 36 in. in depth, and 18 to 24 ft. apart, have been found most successful. We can

readily understand that, as vegetation requires a certain amount of moisture, it is possible to drain land so effectually that sufficient moisture is. not left to fulfil the purposes of cultivation, and the clay soils, which are so reluctant both to receive and to discharge water, will yet suffer a slow and sure deprivation through the agency of deep drains, which will be injurious to the health of vegetation; while drains of less depth would have left the lower part of the stratum in a damp condition, and capable, by the capillary action of the soil itself, of supplying the entire mass with a genial moisture. In Lincolnshire it is a known fact, that if the water in the ditches is reduced to a level below 3 ft. in depth from the surface, the grass land is, in dry summers, most decidedly injured. In the neighbour