There are cases, undoubtedly, where shallow drains should be joined with, if not substituted for, deep drains. On very plastic clays, dig a hole into the soil, and permeate its bottom with small holes, and arrange so that water shall gradually weep through its sides. At first it will rapidly disappear through the bottom, but slowly small particles of clay will wash down, and form a film or puddle over the bottom, after which no water will escape. In all clays which will so puddle, you will observe that much of the water must be removed by surface drains and shallow drains, as the deeper the drain the more chance there would seem to be for puddling. This is not, however, true, in its full application, because another feature in clays may be improved in connection with deep drains, namely, cracking, We know that a clay field will contract under the sun's heat, so as, in many cases, to make during summer really fearful fissures.

As the rains return in the autumn, the clay absorbs the water, swells, and refills the cracks. But if the field was thoroughly and deeply drained, this rain water will be conveyed into the drains all the more rapidly for the fissures, and that which was its curse, becomes its blessing; the clay will close more slowly and perhaps not at all, giving the desired opportunity for ploughing, subsoiling, and sanding, by which the clay may be made exceedingly fertile. When shallow drains shall be used for deep is a very nice question, and can only be settled by a professional and experienced man. The depth of drains seems to be settled, at any rate approximately. Let us now consider their distance apart.

Water can only get into drains by gravity; and just as certainly as it seeks its own level it also seeks the lowest point to which it has access and for which it may run. Let A be a section of a field to be drained. Imagine the

drains, bb, to be made at no matter

what distance apart, but 4 feet deep. The water will at once escape into the drain, being pressed into it by the weight of the mass

overhead. This will be followed by the drops overhead up to the top of its water table, which directly over it or adjacent may be

reduced to 4 feet; but at the same time the water will tend to these two drains from both sides, as is shown by dotted lines. Experiment has shown that in ordinary soils the water needs a slope of about one inch to the linear yard to overcome friction; in very porous soils less will do; in very retentive, more will be necessary. To lower the water table in the centre of our land at a to the desired depth of 4 feet, we must have our drains distant from the centre 18 feet, or the drains 36 feet apart. I have said very light soils will allow the drains to be farther apart. The distance will be affected also by the slope of the land, which may so slope that the water will move rather in the line of steepest descent through the earth, gradually approaching the surface, than laterally into the drain. B is a hillside; the arrow shows the line of steepest de

scent, and the inclination is 1 foot in 20 feet, or 1 inch in 20 inches. Now to overcome friction, we must allow 1 inch to the yard, or 3 inches in 20 feet, for the drains. But the natural slope here is 12 inches in 20 feet, or four times as much, and the inclination is therefore four times stronger to run to the surface at a lower part of the hill than to the drain. If the drains are made diagonally to the hill this is somewhat counteracted. I only mention these cases to show the need there is for judgment and caution. But some one may object that land thus drained at every 40 feet with drains 4 to 5 feet deep will be too dry, and plants will perish. The following cut shows some Wheat plants whose roots were carefully traced when growing freely in open mould; they penetrated 5 feet. Indian Corn has been traced 7 feet; Parsnips 5 and 6 feet. Thus will be seen the enormous advantage

of that treatment of soil which gives the plant 5 feet to feed in, rather than 6 inches, and also that where the plant has room it will push down through the warm soil to the water table, and there get water much warmer than were it standing one foot above.

No one ever saw crops dry up in deep-drained and well-tilled land, whilst all of us have seen it in land where hard pan is within six inches or a foot of the surface, - as is shown

by the wood cuts,-in July. These roots must find all their food and moisture within this shallow soil,

which is soon warmed by the sun and dried out.

Experiments have shown that all land not decidedly sandy is benefited by underground drains; some have found that even sandy and dry land was improved by them, as they seemed to serve as conduits to admit moist air to the roots of the plants, and allow the moisture of the subsoil to rise more freely to the surface. It is therefore claimed that the principle is of universal application. I will not press it to that extent now, but simply assert that all springy, rocky, deep, wet, and low lands should be drained. The rule laid down by English farmers is, a drain every 40 or 60 feet; but this distance is to be varied according to the judgment of the cultivator.

The surfaces to be drained will be either side-hills, undulating, high and level surfaces, or low and level surfaces. In the first and last cases, the adjoining land being higher, experience teaches that the surplus water, to some extent, runs into the place we wish to drain from this higher land near by. We must therefore find an outlet for our drains. The best outlet is a natural brook whose high-water level is lower than the bottom of the drains we are to make. If the high-water level is higher than this, we must still be content, for brooks fall rapidly from their high-water mark; and if at the highest flood our drains do not

work well, we may rest assured that as the flood lowers they will draw the water from the land more rapidly than it could flow off if we had no drains and were compelled to rely on the natural drainage. If no such brook can be found, look for the nearest pond which is ordinarily lower than the bottom of our drains; and if no pond can be found, ascertain the lowest point which is under our control on the farm, and there make a pond, first digging out all the loam, etc., and then, if the bottom is hard pan, cutting through it to some open gravel or sand stratum below. The water discharged into this will soak away; or if we do not open upon a sand stratum we may make an artificial pond.

Having found a point of discharge, dig a main drain as wide as is requisite for the rapid removal of the water, and as deep as the surface will allow at the end most remote from the outlet; if the distance is considerable, the drain will be very much deeper at the outlet, as there should be about 3 inches fall in 100 feet of drain to insure the water's overcoming any slight obstacles to its flow. If the distance to be traversed by the main drain is very long, so that by starting at 3 feet depth and dropping 3 inches to the 100 feet it would sink lower than the outlet, the upper end must be made more shallow, or there must be less drop in 100 feet.

To render this description more intelligible, consult the accompanying diagrams.

inches deep; sink 3 inches to the 100 feet, which gives at the lower end 54 inches in depth. Dig the other main, b' from the brook back, the 150 feet; at the commencement, 49 inches deep; at the outlet, 54 inches. Now dig at the foot of the hill, and following its curve as shown by the diagram, a drain c c c c, nowhere as deep as the main drain by at least 2 inches for every 100 feet that the main drain is distant from it. This drain should be as a general rule not more than 30 feet from the base of the hill. At various intervals cut cross-drains d, from c c c to the main, B, which, starting from the bottom level of c c c, shall sink gradually so as to enter the main above its bottom level. The distance between these cross-drains will vary according to the character of the surface, its unevenness, wetness, and level. As has been said, 20 to 60 feet is common practice, and they may be farther or nearer as necessity directs.

When you can conveniently, enter the side into the main drains at an acute angle, for the flow of the water should be facilitated as much as possible, and every rectangular turn obstructs it more than an acute angle. Also, enter them above the bottom if possible, that the water may reach the bottom of the main after a slight fall. But if our surface were a side-hill it should be treated as in diagram No. 2. Here it is necessary to retard the flow of water, rather

B