"In the Stanford joint tightness is obtained by easting upon the spigot and in the socket of each pipe, by means of moulds prepared for the purpose, rings of a cheap and durable material, which, when pat together, fit mechanically into each other, and by making these rings of a spherical form, a certain amount of movement or settlement may take place without destroying the accuracy of the joint. In laying these pipes, therefore, all that is necessary is to insert the spigot of one fairly and firmly into the socket of another previously laid, and the joint is complete and perfectly watertight. A smearing of some kind of grease is frequently found to be of advantage."

Half socket or access-pipes are sometimes useful, where it becomes necessary often to inspect the house drain. They should be located close to angles, bends, junction branches, running traps, etc. They are not much used in this country, owing, probably, to the fact that, should the main drain run over one-half full, sewage may leak out through the access-pipes into the soil.

Care should be taken to lay the pipes on a firm bed of sand or gravel, and if this is not available, a concrete base should be provided in the trench. The pipes should be laid in straight lines, all changes of direction should be effected by curves of as large a radius as possible, formed of bent pipes. All branches should join the main under an acute angle, by special Y pieces, for a right-angled junction (by a T branch) tends to form eddies and consequently deposits in the main drain.

In laying drains, care should be taken to avoid, as much as possible, trees. The roots of these are frequently found to penetrate and often choke the pipes, and are certainly a dangerous obstruc-tion to the flow in the drain. If the line of the drain must necessarily pass near trees, the use of iron pipes is recommended. The coating of the pipes with coal tar on their outside, the use of asphaltum for joints, and sometimes the surrounding of the drain with a strong layer of concrete are said to be effectual protections against roots of trees.

I now must speak of the grade of the drain, as this is a matter of prime importance. "Upon the inclination of a pipe depends the velocity of the water flowing through it. If this velocity should be insufficient, deposits will occur, and the drain will in time become choked. Pipes of 4 inches diameter should have a velocity of flow of from 3 to 4 1/2 ft. per second; those of 6 and 9 inches diameter should have a velocity of not less than 2 1/2 to 3 ft. A velocity of 2 ft. per second should be considered the minimum allowable in house drains. As a general rule the inclination of a house drain should be as great as attainable, and must be, wherever local conditions will permit, continuous. It is not unfrequently found by uncovering old drains that, in order to save digging, they are laid very flat, often perfectly level, from the point where they leave the house to nearly their junction with the sewer, at which place they are turned with a steep pitch downwards, and often enter the sewer at its crown.

By distributing the whole available fall over the total length of the drain a much better grade would have been secured.

In order to lay a drain with a true grade, especially where the fall is little, a level should be used. The elevation of bottom of pipe, where it leaves the house - at a depth of not less than 3 feet in the New England States, as a protection against frost - should be ascertained, as well as the elevation of the junction with the sewer (or else inlet to cesspool or flush tank). A profile of the ground along the line of the drain should also be determined by levelling. Thus, the proper available fall can be determined, with a little additional trouble, it is true, which, however, will be well repaid by securing a much better quality of the work.

A fall of from 1 in 40 to 1 in GO is desirable for pipes of 4 or 6 inches diameter, but this cannot always be had. I would consider a grade of 1 in 100 as the least to be given to house drains, in order to keep them self-cleansing. When laid with such fall and running full or half-full, a six-inch drain has a velocity of 3 1/2 feet, a four-inch drain-a velocity of nearly 3 feet, which is sufficient to carry along such suspended matters as only ought to enter a house drain. Where the available fall is less than 1 in 100, special flushing apparatus, such as Field's flush tank, McFarland's tilting tank, or Shone's hydraulic syphon ejector should be used.

I have thus fully explained the right method of laying drain pipes, because, even with the best plumbing inside of the house, it is of the greatest importance to have the outside drains of good quality, properly laid, and properly jointed.

The next question to be considered is: What is the proper size for house drains?

This will, of course, depend to some extent upon the grade of the drain, the size of the house and number of its occupants, the amount of water used per head per day, and finally, unless the rain stream flowing through it diminishes. The diagram shows that the velocity is the same for drains running full or half full; it also shows that the maximum velocity of flow occurs not when the sewer is running full, but when the depth of flow is about .813 of its diameter. The maximum velocity is about 11 per cent. greater that that of a pipe running full or half full. The maximum discharge, however, does not coincide with the maximum velocity. The discharge is a maximum when the depth of flow is about .95 of the diameter. At a depth of flow of one fourth of the diameter the velocity is only about 77 per cent. of that when running full or half full, and for lesser depths of flow it diminishes rapidly.

For an ordinary city dwelling a drain four inches in diameter is ample, even including all the rain-fall. For a larger lot and residence a six-inch drain is all that is needed, even if the fall should be only 1 in 100. As a general rule, house drains have been constructed of too large a diameter, and one often meets with the objection that a four-inch pipe will clog up with grease in a short time, or will be obstructed by solid substances. To this, I answer, that in regard to grease the only safe way, where it is allowed to waste, or in case of large boarding-houses and hotels, is to keep it altogether out of the drain (which can be easily accomplished by a suitable grease trap). Grease congealing in a drain is sure to clog it, no matter how large it is made. The stoppage would be only a question of time, and nothing could be gained by postponing this inevitable result. In regard to obstructions by solid matters, I may assert that nothing which passes through the strainer of a sink or from the water-closet bowl can possibly obstruct the drain. What may enter through carelessness of servants, or of the householder, such as "sand, shavings, sticks, coal, bones, garbage, bottles, spoons, knives, forks, apples, potatoes, hay, shirts, towels, stockings, floorcloths, broken crockery, etc.," to quote from Mr. J. Herbert Shedd's Report on the Sewerage of Providence, cannot rightfully be expected to be carried away in a drain. To guard against such obstructions, the drain should be made accessible, especially near bends, junctions and the main trap.

The following useful table, calculated by Robt. Moore, Esq., C.E., from Weis-bach's formula for flow of water through open culverts, gives the size and velocity in house drains, laid at different inclinations, and for various sizes of lots, the rain-fall being 2 inches per hour, and the pipes running 3/4 full. It should be said that the smallest sizes of the table (below 3 or 4 inches diameter) are given only for the sake of completeness, and not as sizes to be recommended for actual use.

Take, for example, an ordinary city lot of 25 x150 ft. = .0861 acres. The rainfall to be provided for may be 2 inches per hour. Though such storms are not frequent, provision should be made for them in the calculation of the size of house drains, as the rain falling on roofs and on paved yards reaches the drain very soon after having fallen. A. rainfall of 1 inch per hour per acre very nearly yields 1 cubic foot per second, therefore 2 inches per hour give 2 cub. ft. per sec. per acre. The number of cubic feet of rain from the above lot is therefore .0861 X 2 = .1722 cub. ft. per second or 60 X .1722 = 10.332 cub. ft. per minute.

We further assume 6 persons to the house, and 75 gallons per head per diem, which is a very liberal allowance. The waste water of the house is therefore 6 X 75 = 450 gallons per day. If one-half of this amount is estimated to run off in 8 hours, the maximum per hour would be about 28 gallons or .0624 cub. ft. per minute. This quantity is so insignificant compared with the rainfall that we may safely neglect it.

Should the drain be allowed to run three-quarters full, and have a fall of 1 in 100, a diameter of 3 3/4 inches would suffice, according to above table.

As a second example, I shall take a large lot, say 80 X 150 ft. = .2755 acres.