which has possibly been raised to some height by steam or horse power, gives at once a motive force at the cheapest rate; while, as channels must necessarily be made for the conveyance away of the waste and dirty water, which has been used for domestic purposes, they can be used with a little alteration for excreta also. It would be a waste of economy to allow this water to pass off without applying the force which has been accumulated in it for another purpose.

But if this is obvious, it is no less so that certain conditions of success must be present, without which this plan, so good in principle, may utterly fail. These conditions are, that there shall be a good supply of water, good sewers, ventilation, a proper outfall, and means of disposing of the sewer water. If these conditions cannot be united, we ought not to disguise the fact that sewers, improperly arranged, may give rise to no inconsiderable dangers. They are underground tubes, connecting houses and allowing possibly, not merely accumulation of excreta, but a ready transference of gases and organic molecules from house to house, and occasionally also causing, by bursting, contamination of the ground, and poisoning of the water supply. And all these dangers are the greater from being concealed. It is probably correct, as has been pointed out, that in deeplaid sewers the pressure inward of the water of the surrounding soil is so great as frequently to cause an inflow into the sewer, and so prevent the exit of the contents; but in other cases, the damage to the sewer may be too great to be neutralized in this way, and in the instance of superficially laid and choked-up pipes, the pressure outward of the contents must be considerable. The dangers of sewers have now been greatly reduced, by having good material, better construction, good ventilation, sufficient water supply, and means of disposal of the sewage water.

Amount of Water for Sewers intended for Excreta.

Engineers are by no means agreed on the necessary amount. We have already named 25 gallons per head per diem, on the authority of Mr. Brunel, as the amount required to keep common sewers clear, and even with this amount there should be some additional quantity for flushing. But in some cases, a good fall and well-laid sewers may require less, and in other cases, bad gradients or curves or workmanship may require more. It is a question whether rain water should be allowed to pass into sewers; it washes the sewers thoroughly sometimes, but it also carries débris and gravel from the roads, which may clog; while in other cases storm waters may burst the sewers, or force back the sewage.'

Construction of Sewers.

Sewers are differently constructed according to the purposes they are to serve, i.e., whether simply to carry off house and trade water, or the solid excreta in addition, or one or both, with the rainfall.

In following out the subject, it will be convenient to trace the sewers from the houses to the outfall.

1 Storm overflows require to be provided; for a description of them see Bailey Denton, op. cit., sections lxii. and lxxxv.

House Pipes and Drains.

It will be convenient to call the conduits inside the house, which run from sinks and closets, "house pipes," and to give the term "drain pipes to the conduits which receive the house pipes, and carry the house water into tanks or main sewers. The house pipes may be divided into sink and water-closet or soil pipes; they are made of metal (lead, iron, or zinc, or two of these) or of earthenware. The drain pipes are usually made of wellburnt, hard, smooth, glazed earthenware. All bricks, porous earthenware, or substances of the kind, should be considered inadmissible for drain pipes. Iron pipes are not much used in this country, but are common and in some places compulsory in America, when pipes have to be carried under houses. When made of heavy cast-iron, jointed and well caulked with lead or Spence's metal, they are the best in many circumstances. Inside they may be enamelled, or coated with Dr. Angus Smith's composition, or treated by Barff's process. The pipes and drains vary in size from 4 to 16 inches diameter, but the usual size of stoneware pipes is 4 to 9 inches; they are round or oval in shape.'

4

Connection of House Pipes with the Drains.-It is customary to commence the drains at the basement of the house, and the sink and closet pipes pass down inside the house and join on, a water-trap being placed at the junction. As the aspiratory power of the warm house is then constantly tending to draw air through the water-trap, and as the trap is liable to get out of order, it is most desirable to alter this plan. The drains should end outside the house, and as far as possible every house pipe should pass outside and not inside or between walls to meet the drain. The object of this is that any imperfection in the pipe should not allow the pipe air to pass into the houses. At the junction of the house pipe and drain, there should not only be a good water-trap, but also complete ventilation and connection with the outside air at the point of junction. The

'Mr. Baldwin Latham cautions us to see that the socket of the drain pipe is made with and is a component part of the pipe, and not merely joined on.

2

Pipes are made up to 36 inches, usually round up to 16 or 18 inches, and oval above that. Engineers are now desirous of restricting the term "drain" to a pipe that merely draws off moisture from land, using the term "sewer" for a pipe carrying sewage or liquid refuse of any kind. This distinction, however, has not been made in the Public Health Act of 1875, in which "drain" is used for the pipe that receives the "house pipes," and "sewer" for the main pipe of a system. (See Bailey Denton's Sanitary Engineering, p. 16.)

See Mr. William Eassie's Healthy Houses (2d edition) for much information on this and kindred subjects. Some of the drawings given here have been copied from Mr. Eassie's work, by his permission; reference may also be made to Sanitary Arrangements for Dwellings, by the same author.

4 Builders are always anxious to conceal tubes, and thus carry them inside the walls, or in the case of hollow walls, between the two. The consequence is that any escape of air must be into the house. The leakage of a closet pipe carried down in a hollow wall often constantly contaminates the air of the house. It would be infinitely better to run the pipes at once through the wall to the outside. Few persons have any idea of the carelessness of plumbers' work-of the bad junctions, and of the rapidity with which pipes get out of order and decay. When a leaden pipe carrying water is led into a water-closet discharge pipe, it is frequently simply puttied in, and very soon the dried putty breaks away, and there is a complete leakage of gas into the house. Even if well-joined, the lead pipe will, it is said, contract and expand, and thus openings are at last formed. Dr. Fergus, of Glasgow, and Dr. N. Carmichael, have directed particular attention to this, in the case of lead closet pipes, which become easily perforated, and which have only a limited duration of wear.

rule, in fact, should be, that the union of any house pipe whatever with the outside drain should be broken both by water and by ventilation. In addition, it should be a strict rule, that no drain pipe of any kind should pass under a house; if there must be a pipe passing from front to back, or the reverse, it is much better to take it above the basement floor than underneath, and to have it exposed throughout its course. In such a case it ought to be of cast-iron, às already mentioned. It is hardly possible to insist too much on the importance of this rule of disconnection between house pipes and outside drains. Late events have shown what a risk the richer classes in this country now run, who not only bring the sewers into the houses, but multiply water-closets, and often put them close to bedrooms. The simple plan of disconnection, if properly done, would insure

them against the otherwise certain danger of sewer air entering the house. Houses which have for years been a nuisance from persistent smells have been purified and become healthy by this means.

Cleansing of Pipes and Drains.-Pipes are cleaned by flexible bamboo or jointed rods with screws and rollers to loosen sediment. The safest plan of cleaning drains is from man-holes, the drains being laid in straight lines from man-hole to man-hole. By this means obstructions are easily detected and removed. The use of movable caps runs the risk of leakage, it being difficult to make the drain water-tight again after removing the cap,

but with care such caps (see Figs. 75 to 77) are useful with small pipes, where man-holes cannot be employed. Drain pipes should also be cleared out by regular flushing, carried out not less often than once a month. This is best done by means of an automatic apparatus such as Field's flush tank (Fig. 79). By regulating the flow of water, it may be made to empty itself as often as necessary.

Laying of Drains.-They should be laid very carefully on concrete in all soils. Sometimes, in very loose soils, even piling for the depth of a foot must be used besides the concrete. When pipes are not laid on a good foundation, leakage is sure to occursooner or later, and the final expense is far more than the first outlay would have been. The greatest care must be taken in laying and joining the pipes, and in testing them afterward to make sure they are water-tight. In a wet soil, a good plan is to have a firm basis, or invert block, which

is itself perforated to carry off subsoil water, and to put the drain over this, as in the plan of Messrs. Brooks & Son, of Huddersfield (see Fig. 92).

وم مرم مرام هرم

The "junction" of pipes is accomplished by special pipes, known by the names of single and double squares, curved or oblique junctions, according to the angle at which one pipe runs into the other. The square junctions are undesirable, as blockage will always occur, and the oblique junctions should be insisted upon. When one pipe opens into another, a taper pipe is often used; the calibre being contracted before it enters the receiving pipe. All jointing must be in good cement, unless special patent joints (such as Stanford's) are used. Clay jointing is wholly inadmissible. Fall of Drain Pipes.-1 in 30 for 4-inch drains, and one in 40 for 6inch; or roughly, for small drains 1 inch per yard.

FIG. 80.-Junctions.

House Traps. As the traps are usually the only safeguard against the warm house drawing sewer air into it, the utmost attention is necessary to insure their efficiency. There is almost an infinite diversity, but they can be conveniently divided into the siphon, the midfeather, the flap-trap, and the ball-trap.

The siphon is a deeply curved tube, the whole of the curve being always full of water. It is a useful trap, and efficient if the curve is deep enough, so that there is a certain depth of water (not less than inch) standing above the highest level of the water in the curve, and if the water is never sucked out of it, and if the pipe is not too small, so that the water is carried away, when it runs full, by the siphon action of the pipe beyond. If two siphons succeed each other in the same pipe, without an air opening between, the one will suck the other empty.

The midfeather is in principle a siphon; it is merely a round or square box, with the entry at one side at the top, and the discharge pipe at a corresponding height on the opposite side, and between them a partition reaching below the lower margin of both pipes. Water, of course, stands in the box or receptacle to the height of the discharge, and therefore the partition is always to some extent under water. The extent should not be less than of an inch. Heavy substances may subside and collect in the box, from which they can be removed from time to time; but as ordinarily made, it is not a good kind of trap, as it favors the collection of deposit, and is not self-cleaning. The common ball-trap, with its modifications, is a variety of the midfeathertrap, but it is so inefficient that it ought to be given up. The best kind of sink trap is the simple siphon, with a screw cap by which to clean it (Fig. 81).

The flap is used only for some drains, and is merely a hinged valve which allows water to pass in one direction, but which is so hung as to close afterward by its own weight. It is intended to prevent the reflux of water into the secondary drains, and is supposed to prevent the passage of sewer gas. But it is probably a very imperfect block.

FIG. 81.-Siphon Sink Trap, with movable Screw for cleaning.

The ball-trap is used in some special cases only; a ball is lifted up as the water rises, until it impinges on and closes an orifice. It is not a very desirable kind.

However various may be the form and details of the water-trap, they can be referred to one or other of these patterns.

Efficiency of Traps.-Water should stand in a trap at least of an inch above openings, and it should pass through sufficiently often and with sufficient force to clear it. An essential condition of the efficiency of all traps is that they should be self-cleansing. Many traps are so constructed that no amount of velocity of water can clear them. Such traps are the common mason's or dip-trap (Fig. 82), and the notori

田

FIG. 82.-Common Mason's or

Dip-trap. Bad form of Trap.

FIG. 83.-D Trap. Bad
form of Trap.

ous D trap (Fig. 83), both of which are simply cess-pools, and could never be cleaned without being opened up. Such traps ought to be unhesitatingly condemned. Traps are often ineffective,-1st, From bad laying, which is a very common fault. 2d, From the water getting thoroughly impregnated with sewer effluvia, so that there is escape of effluvia from the water on the house side. 3d, From the water passing too seldom along the pipe, so that the trap is either dry or clogged. 4th, From the pipe being too small (2 or 3 inches only), and "running full," which will sometimes suck the water out of the trap; it usually occurs in this way, as frequently seen in sink traps: the pipe beyond the trap has perhaps a very great and sudden fall, and when it is full of water it acts like a siphon, and sucks all the water out of the trap; to avoid this, the pipe should be large enough to prevent its running full, or the trap should be of larger calibre than the rest of the pipe. This, however, will not always prevent it, as even 6-inch pipes have sometimes sucked a siphon dry. The question has lately been very carefully investigated in America, by Messrs. Philbrick and Bowditch,' whose report has shown the danger of unsiphoning which small pipes are exposed to. The remedy appears to be to introduce an air-vent at the crown of the trap (see Fig. 84), and not to have too small a pipe, especially when several pipes unite in one general waste. The experiments also showed how unsiphoning might take place from the pressure of descending water from upper floors, so that air might be forcibly driven

FIG. 84.-Siphon Closet Basins with ventilating pipes. A, Soil pipe passing up above the eaves, with open top. B, Subsidiary Ventilating Pipe (also passing up above eaves, with open top) to prevent sucking of the siphon.

1 The Sanitary Engineer, vol. vi., p. 264, 1882 (New York). The Siphonage and Ventilation of Traps, Report to the National Board of Health, by E. W. Bowditch and E. S. Philbrick, C.E.