let, by the necessity for a fall and the difficulty which often exists in securing a sufficiently low outlet. In such case, the only rule is to make it as deep as possible. When the fall is sufficient, it should be placed at such depth as will allow the laterals and sub-mains which discharge into it to enter at its top, and discharge above the level of the water which flows through it.

3. Subsidiary mains, or sub-mains, connecting with the main drains, should be run up the minor valleys of the land, skirting the bases of the hills. Where the valley is a flat one, with rising ground at each side, there should be a sub-main, to receive the laterals from each hill side. As a general rule, it may be stated, that the collecting drain at the foot of a slope should be placed on the line which is first reached by the water flowing directly commences its lateral should, if possible, be

Fig. 10.-STONE PIT TO CONNECT SPRING

WITH DRAIN.

down over its surface, before it movement down the valley; and it so arranged that it shall have a uniform descent for its whole distance. The proper arrangement of these collecting drains requires more skill and experience than any other branch of the work, for on their disposition. depends, in a great measure, the economy and success of the undertaking.

4. Where springs exist, there should be some provision made for collecting their water in pits filled with loose

stone, gravel, brush or other rubbish, or furnished with. several lengths of tile set on end, one above the other, or with a barrel or other vessel; and a line of tile of proper size should be run directly

to a main, or sub-main drain. The manner of doing this by means of a pit filled with stone is shown in Fig. 10. The collection of spring water in a vertical tile basin is shown in Fig. 11.

5. Where a ledge of shelving rock, of considerable size, occurs on land to be drained, it is best to make some provision for collecting, at its base, the water flowing over its surface, and taking it at once into the drains, so that it

SPRING WITH DRAIN.

effect this, a ditch should be dug along the base of the rock, and quite down to it, considerably deeper than the level of the proposed drainage; and this should be filled with small stones to that level, with a line of tile laid on top of the stones, a uniform bottom for the tile to rest upon being formed of cheap strips of board. The tile and stone should then be covered with inverted sods, with wood shavings, or with other suitable material, which will prevent the entrance of earth, (from the covering of the drain,) to choke them. The water, following down the surface of the rock, will rise through the stone work and, entering the tile, will flow off. This method may be used for springy hill sides. 6. The points previously considered relate only to the

collection of unusual quantities of water, (from springs and from rock surfaces,) and to the removal from the land of what is thus collected, and of that which flows from the minor or lateral drains.

The lateral drains themselves constitute the real drainage of the field, for, although main lines take water from the land on each side, their action in this regard is not usually considered, in determining either their depth or their location, and they play an exceedingly small part in the more simple form of drainage, that in which a large tract of land, of perfectly uniform slope, is drained by parallel lines of equal length, all discharging into a single main, running across the foot of the field. The land would be equally well drained, if the parallel lines were continued to an open ditch beyond its boundary,-the main tile drain is only adopted for greater convenience and security. It will simplify the question if, in treating the theory of lateral drains, it be assumed that our field is of this uniform inclination, and admits of the use of long lines of parallel drains. In fact, it is best in practice to approximate as nearly as possible to this arrangement, because deviations. from it, though always necessary in broken land, are always more expensive, and present more complicated engineering problems. If all the land to be drained had a uniform fall, in a single direction, there would be but little need of engineering skill, beyond that which is required to establish the depth, fall, and distance apart, at which the drains should be laid. It is chiefly when the land pitches in different directions, and with varying inclination, that only a person skilled in the arrangement of drains, or one who will give much consideration to the subject, can effect the greatest economy by avoiding unnecessary complication, and secure the greatest efficiency by adjusting the drains to the requirements of the land.

Assuming the land to have an unbroken inclination, so as to require only parallel drains, it becomes important to

know how these parallel drains, (corresponding to the lateral drains of an irregular system,) should be made.

The history of land draining is a history of the gradual progress of an improvement, from the accomplishment of a single purpose, to the accomplishment of several purposes, and most of the instruction which modern agricultural writers have given concerning it, has shown too great dependence upon the teachings of their predecessors, who considered well the single object which they sought to attain, but who had no conception that draining was to be so generally valuable as it has become. The effort, (probably an unconscious one,) to make the theories of modern thorough-draining conform to those advanced by the early practitioners, seems to have diverted attention from some more recently developed principles, which are of much importance. For example, about a hundred years ago, Joseph Elkington, of Warwickshire, discovered that, where land is made too wet by under-ground springs, a skillful tapping of these,-drawing off their water through suitable conduits, would greatly relieve the land, and for many years the Elkington System of drainage, being a great improvement on every thing theretofore practiced, naturally occupied the attention of the agricultural world, and the Board of Agriculture appointed a Mr. Johnstone to study the process, and write a treatise on the subject.

Catch-water drains, made so as to intercept a flow of surface water, have been in use from immemorial time, and are described by the earliest writers. Before the advent of the Draining Tile, covered drains were furnished with stones, boards, brush, weeds, and various other rubbish, and their good effect, very properly, claimed the attention of all improvers of wet land. When the tile first made its appearance in general practice, it was of what is called the "horse-shoe " form, and,-imperfect though it was,— it was better than anything that had preceded it, and was received with high approval, wherever it became known.

The general use of all these materials for making drains was confined to a system of partial drainage, until the publication of a pamphlet, in 1833, by Mr. Smith, of Deanston, who advocated the drainage of the whole field, without reference to springs. From this plan, but with important modifications in matters of detail, the modern system of tile draining has grown. Many able men have aided its progress, and have helped to disseminate a knowledge of its processes and its effects, yet there are few books on draining, even the most modern ones, which do not devote much attention to Elkington's discovery; to the various sorts of stone and brush drains; and to the manufacture and use of horse-shoe tile;-not treating them as matters of antiquarian interest, but repeating the instructions for their application, and allowing the reasoning on which their early use was based, to influence, often to a damaging extent, their general consideration of the modern practice of tile draining.

These processes are all of occasional use, even at this day, but they are based on no fixed rules, and are so much a matter of traditional knowledge, with all farmers, that instruction concerning them is not needed. The kind of draining which is now under consideration, has for its object the complete removal of all of the surplus water that reaches the soil, from whatever source, and the assimilation of all wet soils to a somewhat uniform condition, as to the ease with which water passes through them.

There are instances, as has been shown, where a large spring, overflowing a considerable area, or supplying the water of an annoying brook, ought to be directly connected with the under-ground drainage, and its flow neatly carried away; and, in other cases, the surface flow over large masses of rock should be given easy entrance into the tile; but, in all ordinary lands, whether swamps, springy hill sides, heavy clays, or light soils lying on retentive subsoil, all ground, in fact, which needs under.