TABLE.

The following table I first computed from tables given by B. Latham, in his work on sewers, in 1877. His table shows discharge in cubic feet per minute, when full. So far as I know, it is the only table which gives the acres drained by a given-sized tile on a given grade. The rainfall to be removed is considered equal to one-half inch in depth each day of twenty-four hours. The table has been used quite extensively by drainage engineers in this State, and has been found to give good results. I have revised it, corrected some errors and indicated, by omission of figures, what should be the maximum and minimum grade for each size of tile.

[NOTE.-Tile should not be laid to grades when numbers are replaced by a dash.]

CONSTRUCTION OF A DITCH.

In almost every instance this work will have to be begun at the outlet of the drain, in order that water may not interfere with the work. If the soil will stand for a long time—that is, until the excavation for the whole ditch is finished-it is advisable to commence to lay tile at the upper end of the ditch and work towards the outlet. This method permits only clear water to enter the drain. I have, however, never seen a case in which it was practicable, but I have no doubt there are localities in which the soil will stand until the whole ditch can be excavated. Good working under-drains, even in very flat land, could be constructed without leveling under such circumstances; for, after the ditch is excavated for its whole length, irregularities in the bottom could be detected by pouring in or allowing water to run through it.

In nearly every case, however, in order to prevent caving, tile must be laid as fast as the ditch is excavated, and in such a case even the test of running water is worthless. Suppose, for instance, you are to build a drain 40 rods long, in which the maximum fall is only two feet. The profile of almost all these drains will show a rapid rise close to the outlet, and near the upper end little or no rise; frequently the drain must cut through a bank or dam. Now, in working by a water level, there are no means of ascertaining the amount of rise per rod, and a ditcher is almost certain to rise too fast at first, and, in order

to maintain his depth at the upper end, must have his bottom incline the wrong way. Even in that case it would work when the water rose above the highest point in the tile; but there would be no current above that part, and the tiles would rapidly silt up.

In excavating it is important to dig as narrow a ditch as possible. In the first place it means less earth to heave out, and in the second place a narrow

ditch forms a snugger bed for the tilean important consideration. Where a plow can be used the cutting of one or two furrows will aid the ditchers amazingly; but for some reason they frequently will not admit that it helps at all; they do not like the rough work left by the plow. For digging the common spade is the best implement for most purposes. The last spading, however, had better be done with a spade of the general form of the ordinary one, but as narrow as can be worked and admit the tile. The long narrrow, tapering, tiling spades are worth. less for most soils. They are good when the soil will admit of clean work, but upper end prevents the use of this spade

The tile scoop, shown in figure 8, which any blacksmith can make, for cleaning the bottom, should be made of a sheet of half-round steel with the handle fastened opposite its center, so that it may be used as a push or pull scoop. The handle should rise at an angle of about 30 or 45 degrees. The scoop usually on sale is of thick, heavy wrought iron, with the handle fastened at the back end-a more awkward instrument could hardly be devised. In the first place, the earth is loosened at a disadvantage as compared with the work of the scoop with a handle fastened at the center; and, again, it is only of use as a push scoop. For two, three, or four-inch tile, one with a diameter of four inches will do nicely. For larger tile-six to eight inch-the diameter may be increased to six inches, although the smaller scoop may be made to do the work. The length should be 12 to 15 inches.

In quicksand, the excavation cannot be carried ahead of the tile, and much difficulty is experienced in keeping the tile open while constructing the drain. If the quicksand is deep-say, three or four feet-a curbing will have to be used, and it is best put in as follows: Take 12 pieces of two-by-four scantling, each about six feet long; drive these down in pairs of two on each side of the ditch, within a distance of 12 feet; put between each pair horizontal, inch boards 12 feet long. By braces between the inner uprights, prevent the earth from pushing in the sides. In excavating, first remove the sand from below the boards on each side; shove them down, throw out the center dirt last; so continue, and the quicksand cannot possibly get into the ditch. The curbing can be sunk down as deep as necessary. The boards can all be removed by taking out the bottom one first, filling with earth, and then the second one, and so on. This is best worked in three lengths at one time, taking up one length and putting one down at the same time.

When the quicksand is not over two feet in depth, it can be held back by a U shaped box of iron without bottom or top, as seen in Fig. 9.

Wood is

rather thick to sink in the sand, but it could, doubtless, be successfully used. Fig. 9 is a view of a box, built after a design of my own of 1-16 boiler plate iron; its length is five feet; depth, 15 inches; width, one foot. Two handles, B B, are put on for the purpose of moving; an edge of angle-iron is riveted on the top. This was made for laying eight-inch tile. For smaller tile, the pieces might be much nearer together. The back handle, B, is made of three-quarter inch round iron to prevent the collapse of the back end, owing to pressure from the sand. The arrangement has worked well.

KIND OF UNDER-DRAINS.

Curbing Box. Fig. 9.

Under-drains are best constructed of tile, and the shape best suited for every place is the plain cylindrical form. The tile are now usually laid with the ends abutting against each other, leaving as smooth a channel as possible for the water. In soil that is not easily washed, no protection at the joints is needed; in sandy or peaty soils the tops should be covered, so that the water will be forced to enter at the bottom of the joint; it will then be likely to be free from sediment. The best covering for joints of tile is, in my opinion, a piece of closely mown turf, laid grass down. This is especially good, as it seems also to prevent lateral displacement of the tile. Small pieces of tarred paper, of old tiles, of tin, etc., answer a good purpose. Tarred paper is always cheap and convenient to use. Turf is sometimes objected to because of the silt that may be worked into the tile. The water gets into the tile principally at the joints, though the tile is in itself porous. This was strongly illustrated by a construction that was carried out by the Lansing Wheelbarrow Works, owing to a misunderstanding of some of my directions. In laying the tile the joints were completely enveloped with tarred paper. The result was that no water entered the tile, and the paper had to be removed before the tile would work.

TILE LAYING TO GRADE.

As already noticed in a previous place, this requires much care. It is usually necessary to test the bottom of the ditch in advance of laying the tile. And also every tile after it is laid. As this work will be a permanent job, if done well, no pains should be spared in laying the tile, which is the least laborious part of the work.

My method is-and this method has only to be tried to be adopted by allto stretch a strong cord about three feet above the ground, and get this parallel to the required grade, as shown in Fig. 10. Now by supporting this string at intervals of 30 feet (see Fig. 10), the variation from a straight line will be insignificant, and by measuring down the required distance one can determine the correct bottom of the ditch with exceeding accuracy and with great celerity. This method of preserving a true line for the tile I wish to call particular attention to; for it may be used not only to preserve the grade line, but to secure it, and hence be a substitute for a line of levels. Fig. 10 shows, I think, quite clearly how the string is carried, the stakes or posts are driven on opposite sides of the ditch, and then connected with a cross-bar. This cross-bar should be arranged so that either end can be raised so as to be made truly level. For this purpose three forms of clamps are shown in detail at A,

B, and C. These, I think, will be understood without explanation. The clamp B is formed by merely sawing a slot in the cross-bar, in which slip the upright stake; two bolts are put through, one to prevent splitting; the other to tighten by means of a wrench or a thumb-screw. The common quilting frame clamps would answer very nicely. If a survey has been previously made, and the depth at each point is given, the cross-bars are set by measuring up from the surface so as to be a uniform distance from the bottom. They should be seven or eight feet above bottom of ditch, where they will not interfere with tile laying. Over the center of the ditch on the cross-bars pull the line with a stick whose length is equal the hight of the line above the bottom, say seven or eight feet as the case may be, the bottom can be readily tested.

If no survey has been previously made, we can use these to determine the fall and the grade line. For instance, at 1 in Fig. 10, we set our first crossbar and stakes, say so as to give a depth of four feet.

At the farthest distance that can be seen we set another cross-bar and stakes,

so as to give us the required depth at that point, say three feet. To find the fall between these two, we have only to sight over a level of any construction, from over first cross-bar to our last, and measure the distance of the intersection of this sight line to our cross-bar; this last method is a rough one, and will only serve to show whether we have any fall or not. After we find there is sufficient fall between the two cross-bars, any number of intermediate ones can be set by sighting, and accurately too. Leave these in until the tiles. are laid under any given one, and you will find your work done with all the accuracy needed, and in very much less time than if you had attempted any method which depended on leveling the bottom itself. This method will substitute certainty for uncertainty, and give you success, where by laying tile by the eye or running water you would be likely to fail.

Raval New Yorken

Fig. 11.

LAYING THE TILE.

It

It is not a good plan, unless the bottom is unusually hard, to walk in the ditch after it has been prepared for the tile. If tiles are laid by hand, the better way is to keep about one foot of earth filled over the tiles, and, standing on that, reach over and lay the tiles. This gives an excellent position for using the push-and-pull tile scoop, and in sandy or mucky soils when the tiles laid need frequent cleaning by hand, it is certainly the best method. The bottom is tested in advance by measuring from the overhead string, and each tile is tested after being laid. In peaty or mucky soils great care is necessary that the ditch be not excavated too deep. No random filling will give as uniform a bottom as the original undisturbed muck. might be a good plan to excavate several inches below the grade line and fill to the grade with gravel; but this is hardly necessary. In quick-sand, the side pressure is more likely to cause the tiles to rise, and the ditch should be filled at once to hold the tiles down. No board should be used for bottom in either case. Tile will settle uniformly if the soil is of the same character throughout. In quick-sand no alarm need be felt if the tile fills one-half or three-fourths full of sand when the work is progressing. If there is any current whatever, sand may be washed out clean by a flush of water. It is a good thing in such soils to keep a long iron rod in the tile, to be worked ahead occasionally to prevent the sand from settling in the tile. The joints of tile in quick-sand and muck should be made as close as possible; for this purpose a wrapping of tarred paper is excellent. Collars might in such soils serve a good purpose. Where the soils will stand well, the tiles can be laid from the surface with good results. A pole with an iron pin long enough to hold one tile is commonly used.

Fig. 11 is a view of the instrument usually used for this purpose. A much better arrangement for small tile is shown in use in Fig 12. The rod on which the

pipes are strung is sufficiently long to hold six tiles, and be inserted six or eight inches in those previously laid. They can be laid and the joints wrapped, if necessary, while on the pin. After the tiles are laid they are at once covered sufficiently to hold them firmly in place, and the pin is pulled out with the rope attached to it. An old scythe handle is the best form for the handle to

Fig. 13.

the long pin. The diameter of the pin should be such as to slip into the tile easily, but not much smaller. In filling, care must be taken when the dirt is first thrown in, not to move the tiles laterally. After they are covered to a depth of about a foot, if the ground is sufficiently solid, the filling can be