these pipes was carefully trapped it was supposed that no return of foul air could take place, and so every one slept in peace. The important condition of the problem, which had so far been overlooked in making the arrangement in this way, was the fact that the pipes were not constantly full of water but of air. From its lightness, the air necessarily fills all the spaces from trap to trap, whilst the water flows down to the lowest level which it can find. The air, of course, becomes polluted by the frequent passage of discharges, and by constant contact with the dirty pipe. “Evil communications corrupt good manners.” Whatever disease-germs may be present will thus be carried about, not only by foul water, but by the air which it has contaminated. It soon became evident to some who studied the subject that when the pipes were thus charged, if any water was poured into one of them, it would, owing to its superior gravity, fall through the air, and either cause the immediate displacement of an equal bulk of it, or produce a temporary increase of pressure in the pipes. This pressure would be gradually relieved by the escape of air and water until the normal condition was resumed. Now, the all-important question was, “In what direction will the immediate displacement or gradual escape of this air take place?” If it went downwards with the discharged water, and passed through the several traps into the drain or cesspit, all might be comparatively well. But obviously this could not take place. In such a race the heaviest fluid must necessarily win the race, and, in its descent, must force upwards most of the air which was stationary in the pipes. At the best, the water could only force before it, or carry along with itself, a small portion of the air, leaving the greater part to reascend the pipes and force its way, until the normal condition of pressure was restored, through one or other of the traps. Nor is this forcing of its way through the water in a trap, which some from want of experience may say is impossible, its only way of escape into the house. This airtight system of pipes involves other dangers, one of which is that the water may be siphoned out of any trap by the sudden discharge of a large quantity of water, either through itself, or down some other pipe of the same system. If such a body of water rushes impetuously down a nearly vertical pipe, temporarily forcing onward some of the air, it must produce a partial vacuum behind itself. The inrush of air from the house to fill this vacuum will often be strong enough to carry with it so much of the water which was lying in the trap as to leave a free passage for the subsequent ascent of the foul air. When this ascent takes place, if it happens that the waste-plug or valve has not been returned to its place immediately after the discharge has passed away, the foul air will slowly

float upwards into the house and become a source of danger to health. If it becomes at the same time a cause of offence to the nostrils, we may congratulate ourselves, because we shall be warned of what has taken place. Another danger is that mere evaporation from the trap of some discharge-pipe which is seldom used may leave an unobstructed passage for foul air to enter the house.

When this fuller knowledge of the conditions of the problem was properly realised it was soon seen that the airtight system must be given up. This abandonment must not, however, be understood to indicate an indifference to the presence of leakage in the pipes or drains. So far from this, freedom from leakage is now more carefully guarded against than formerly, and, in order to secure it, better materials and more perfect jointing are demanded. In considering how to provide a vent for the air displaced by every discharge of water into the pipes, it became necessary to guard against the possibility of its entering the house. It was seen that the vent for it must be carried above the highest window, above the eaves, and even above the joints of the slates. A simple method of doing this was to carry the soil-pipe, or a smaller continuation of it, upwards to the ridge of the roof, or to a stack of chimneys, instead of stopping it at the highest water-closet. Another step in advance was to form a free inlet for air at the bottom of the soil-pipe. By these two improvements a current of fresh air was made to pass constantly through the whole length of the pipe when not being used, and this would render any accumulation of injurious air impossible. When a soil-pipe so fitted came to be used, the discharged water would encounter nothing to retard its exit. The air below the descending water would be forced forward through the lower opening, and that above it would freely yield to the downward suction. This would prevent the risk of the demand for air causing the water to be sucked out of the traps. It will be seen also that each of these openings acts as an inlet or an outlet, as occasion demands. Plate XLIX., fig. 1, shows a soil-pipe ventilated in the way described.

The next step in the direction of improvement is to disconnect all water-wastes from the soil-pipe. It must be obvious that the connection of these with a soil-pipe connected with a drain will bring many further dangers of contamination to the air of the house, because it means that openings are provided into other rooms than the water-closet—into rooms, indeed, where bathing, washing-up, and cooking are being done. Such connections, therefore, not only bring additional dangers, but greater ones, because of the longer time during which the air of these rooms is breathed by the occupants of the house. The

cleansing which water-wastes may effect in the soil-pipe by their discharges is very small, and this is the less necessary after the pipe has been opened at both ends, and especially as thereby its own discharges will better than formerly effect the purpose now that they get away more rapidly. There being, then, no valid reason for continuing such a plan, it was seen to be much better to carry each water-waste, or several combined, through the wall of a house, and to let the end project over, and discharge in the open air upon, an ordinary grated sink connected by a good trap to the drain. By adopting this arrangement we get (1) a free inlet for air at the bottom of the waste-pipe, and (2) an entire severance of it from the drain. All such wastes should be well trapped at the top, because the pipes necessarily get foul, and will consequently emit tainted air. In the event of two or more wastes running into the same main pipe an upper air-inlet also must be provided, lest the discharge rushing down the main pipe should suck the water out of one of the traps. In the case of any discharge which may be a heavy one, as from a bath or slop-sink, such an inlet should also be provided to prevent it from untrapping itself. In all such cases there will be the additional advantage of a free current of air through, as described in the improved soil-pipe. There is not the same urgency for this, but where first cost does not stand in the way it is very desirable. A good plan is to carry the waste-pipe itself up, and out through the wall a little above the highest vessel to be discharged.

The rainwater pipes should never be connected to the drains, but should either discharge upon such a grating as that just described, or else into an open channel communicating with one. Where they are connected to a drain the foul air will flow up them in dry weather, and may enter a window or find its way in at the eaves or under the slates. On one occasion when visiting at a friend's house in London, in the summer-time, I left the bedroom-window open, and woke in the night to find the room filled with the vilest smell. In the morning I looked out of the window to discover the probable cause. Underneath was the top of a bay-window, with a small and innocent-looking pipe in one corner to carry off the rain. My host got a plumber at once, and found that this pipe was carried direct, without any trap, into the drain which led to the main sewer in the road. The small pipe was made of zinc, and was perfectly blackened inside by the foul gas for which it may have been for years the principal outlet. The danger in such cases, however, is not confined to houses connected with public sewers. In a house to the sanitation of which I had devoted much care, and which had its own cesspit, I noticed a bad smell on the top of the porch. The London experience

occurred to me, and I went at once to the head of the rain-water pipe. There was a most offensive odour rising from it. The tops of two bay-windows were fitted in the same way, and above each there was a bedroom-window. I found the rainwater pipes connected directly to the drain, and had them at once cut off above ground, so that they should discharge into the open air. The reasons which make foul air so ready to enter a house will be dealt with presently under the heading of “Ventilation.” Plate XLIX., fig. 2, shows a system of water-wastes arranged in the way described. As each principal part is named, it will probably explain itself pretty fully.

The next point with which I will deal is the connection of the soil-pipe with the drain. This should be done in such a manner as to insure the carrying-out of the same principle as that adopted in connecting the water-closet and soil-pipe. This principle is the prevention of the escape of foul air from the pipe into the house when each discharge is made. In the case now under notice we must similarly prevent the passage of foul air from the drain into the soil-pipe. The trap which forms the connection must not therefore be of the old air-tight description, but must contain a vent open to the air from the house-side of the water-seal of the trap. If such an opening is provided it will prevent any foul air from passing into the soil-pipe, because it supplies an easier outlet. This opening will at the same time answer for the fresh-air inlet required at the foot of the soil-pipe. Care must of course be taken that this vent shall not be directly under, or near to, any of the windows of the house. Plate XLIX., fig. 1, shows the position of this intercepting trap, and Plate XLVIII., fig. 1, gives a detail of it.

The final point of importance is the ventilation, where practicable, of the drain connecting the house with the main sewer. This will not be possible in a public street, but there are many cases where a house stands well back from the road, and others where there is a private cesspit in the grounds. To insure real ventilation, here as in other cases, an outlet merely is not sufficient, but an inlet must also be provided. Both should be kept well away from windows or much-frequented paths, and the outlet should be carried up as high as possible. An exhaust ventilator on the outlet, and an inlet ventilator, such as a cowl arranged to keep its mouth instead of its back to the wind, will greatly add to the efficiency of the ventilation. Of the two arrangements possible, the better one would be to have the inlet at the end near to the house and the outlet at the distant one.

In endeavouring to make sanitary arrangements, do not be deceived by the supposed efficacy of deodorisers when applied to the various openings. It is not the mere smell of the foul

quite straight may be sufficient. All the pipes should be solidly bedded, and a small trench made under each socket to give free access for filling the lower side of the joint. The second point is the proper filling, without leaving any of the material to project inside the pipe to form an obstruction. To prevent this, a small gaskin of hemp, dipped in liquid cement, should be fastened round the pipe, and pushed home to the bottom of the socket. When this is done the socket should be well filled all round with Portland cement, nicely finished off, taking as much care with the under part as with the upper. It is further desirable to insert a well-padded and close-fitting plug, called a mouse, into the first pipe, and, by means of a strong cord, to draw it forward past each joint after it is made. This will wipe off any cement which may have oozed past the gaskin. Clay is often used for filling, but there are several peculiarities which render it unfit for a jointing material. First, it shrinks in drying; second, if it is not very hard the weight of the pipes may cause them to compress it on the under side, and thus form a vent on the upper; third, both these defects, by enabling the water to commence a run, lead to the clay-filling being constantly reduced, with the frequent result of undermining other parts which were perfect, and ultimately causing the collapse and stop page of the drain; fourth, and worst of all, it is possible for fibrous roots to find their way through the clay and choke the drain.

There are several reasons which make it desirable that drains should be tight: (1.) They are often under a house, and always come close to its walls. (2.) If they leak the liquid will impregnate the ground and cause an unhealthy condition. (3.) It may contaminate the water of a well, and produce typhoid fever. (4.) Foul gas may arise and be attracted into the house by the warmer air. (5.) The escape of the liquid reduces its flushing-power, and therefore increases the deposition of sediment, so that the risk of the drain becoming stopped and the certainty of its becoming fouler than otherwise are rendered greater. In cases where it is necessary to lay a drain under a house it should be entirely bedded in and covered by cement concrete so as more effectually to guard against leakage or the cracking of a pipe as the result of settlement of the ground or foundations.

Wherever a branch has to be connected to a drain it should be done through a socketed junction-pipe, set at the proper angle, and entering in the direction of the fall of the drain. The insertion of a branch into a hole chipped out of the drain should never be tolerated. It is difficult to make tight; the broken chips enter the drain, and the end of the branch is very likely to form a permanent obstruction. A proper junc-

perfectly, and are at the same time more expensive than the simple earthenware form. No valve closet is convenient for pouring a large pailful of slops down, because the valve presents an obstruction, even if held up, which is not an easy operation to perform while pouring. There are two principal varieties of the closet recommended, which are generally distinguished as the “wash-down” and the “wash-out” closet. On Plate XLVIII., fig. 2 shows a section of the former, and fig. 3 of the latter. For the former it is contended that the force of the flush goes directly downwards into the trap and carries all before it more effectually than it can do in the latter, in which it first discharges into the pan, and then flows over the rim into the trap. I am inclined, notwithstanding this, to prefer the latter, because, owing to there being a larger surface of water immediately under the centre of the pan, its sides are less likely to become soiled. It will be seen that both of these closets are of the simplest possible construction, and that as long as they remain unbroken, and are used frequently enough to prevent the water in the trap from evaporating, it is absolutely impossible for any return of foul air from the pipe to take place. Even if the water-seal should be destroyed by evaporation such air from the drain would find a readier exit upwards through the ventilating-pipe rather than by descending between the lips of the trap into the pan. There is also the minimum of surface, and that of the smoothest, for any foul matter to collect upon, and there is no mechanism to get out of order.

The closet should always have a flushing-pipe not less than 1.¼in. diameter, which should come from a waste-preventer cistern, separate from all others, holding not less than two gallons of water, and placed at least 6ft. above the closet. The use of a separate cistern will prevent the possibility of contaminating any water used for drinking or washing. It should be placed in the room, where it is always accessible, and not above the ceiling.

To insure the most wholesome closet arrangement the wooden casing round it should be entirely dispensed with, and only a hinged flap should be provided as a seat. By adopting this plan it will, when the lid is raised, serve the purpose of a slop-sink and urinal as efficiently as anything that could be contrived for these uses, and will at the same time save the space and extra cost of providing and fitting them. If expense does not stand in the way a very nice addition is to cover the floor and surrounding walls for a short space with glazed tiles set in cement. You have then the most perfect arrangement, in my humble opinion as a householder, which has as yet been suggested by the experts.

The detailed arrangements which have so far been de-

revolving from west to east, and, as the atmosphere is not carried round at the same speed, there is a tendency to produce a current in the opposite direction—i.e., from the east. These two movements combined cause the air-currents to take a middle course, and to form in the Northern Hemisphere the “north-east trades” and in the Southern the “south-east trades.

The only part of this illustration which will help our present investigation is the first part—viz., that which shows (1) that warm air will always rise, and (2) that cold air will readily flow in to take its place. The law which governs this action is simply that which causes water to seek its own level. The reason of the latter tendency is that, when water and air together are drawn downwards upon the earth's surface by the attraction of gravitation, the water, having a greater specific gravity than air, is drawn to the lowest parts, and displaces the air from them. But, if quicksilver was also in the race, it would as speedily displace the water. This law, then, is universal in its operation, and causes all fluids to range themselves in horizontal layers according to their specific gravities—the lighter at the top and the heavier at the bottom. Now, warm air is of less specific gravity than cool air, and therefore floats above the latter, and, for all purposes with which ventilation has to do, acts just as if it was a distinct fluid.

Having realised that the law governing our problem is simply that of gravitation, let us apply it to the conditions which we know exist in connection with a house. We are all conscious that the average temperature of the air inside a house is greater than of that which is outside. It must follow that there is a constant effort being made by the inner air to float out at the higher openings, and by the outer air to flow in at every opening, whether high or low. A simple illustration will make this action clear: Supposing it to be possible to lift a house bodily, and to immerse it suddenly in a lake of water, the result would be that at all the joints of doors, windows, and slates, under the eaves, and down the chimneys water would commence to flow in. The air would at the same time begin to escape and to bubble up to the surface. A little consideration will show, however, that it would not come out of the house at every opening at which the water was entering. As soon as the water rose on the inside of the house above any opening, that opening would cease to emit air, because it would always be impelled upwards and never downwards. If there were no outlets from the rooms above the tops of widows or doors a quantity of air would be imprisoned in them as in a diving-bell. To make this illustration apply to ventilation we have only to substitute outer (or cool) air

for water, and inner (or warm) air for air, and to remember also that the air which flows into a house is being continually warmed, and caused in its turn to float out, to see that this is the constant operation which gravitation tries to carry on in a house.

To pass from this illustration, another fact muat be borne in mind—viz., that the various openings into a house are constantly acted upon by the varying currents of air which are moving outside. The effort of the cool air to get in at some openings from which warm air is trying to escape, coupled with the baffling influence of outside currents, constitute the principal difficulties with which we have to deal. These efforts which Nature makes to prevent a stagnation of air in any part of her domain do, within certain limits, ventilate our houses. Why, then, should we not be content with what Nature does? The answer must be that her efforts are too rough-and-ready, and therefore need regulation in order to adapt them to our requirements.

Before attempting to regulate them we must decide—firstly, what the nature of our requirements is; secondly, what is the measure of the necessary modification; and, thirdly, under what restrictions must this modification be carried out. The nature of our requirements is the maintenance of a certain degree of coolness and purity in the air contained in our rooms so that it may suit the purposes of breathing. The measure of the necessary modification of nature's work will be the amount by which it falls short of providing the cool and pure air required in any given instance. The restriotions to be observed are threefold—(1) That from the exits provided for the warm air to escape at no downdraughts must proceed; (2) that in admitting fresh air no draught must be produced such as may endanger our health or comfort; and (3) that the air must not be changed so rapidly as to lower the temperature unduly.

Before suggesting various contrivances for effecting these objects, let me say that, as I am treating only of dwelling-houses of moderate dimensions and not of large buildings for public assemblies, it will be unnecessary to speak of steam-fans, of furnaces, of enclosed gas-jets, or even of water-sprays as agents for the removal of warm air. The original cost and current expense of two at least of these methods entirely prevent their use in ordinary dwellings. Of the two others—viz., the enclosed gas-jet and the water-spray—I have no experience, having been inclined to view them too much in the light of an extravagance to make a trial of them. I am therefore left to make the best I can of gravitation and of the wind, two agents which cost nothing, but which be guided in such a way as to render valuable service

without adding much to the original cost of building a house.

Under this limitation let us consider what we have to do in order to maintain a reasonably pure and cool air in any room. Our aim must be twofold—viz., to remove the warm and vitiated and to introduce cool and pure air. It is very little use to provide for either one of these operations without the other, while the intelligent carrying-out of both tends to render the working of each more effective. Thus, if the warm air is removed freely it is reasonable to expect that cool air will flow in more readily, and, if the entrance of cool air is facilitated by easy and suitable passages, then the warm air is likely to be withdrawn more readily.

First, then, as to the removal of warm air. The restriction imposed upon us in doing this is that we must avoid any arrangement which will allow of a downdraught. Naturally, the removal must be effected at the upper part of the room—the warmest place. Above the top of windows and doors in an unventilated room, especially in one which is artificially lighted, there is a close reservoir of hot and impure air which cannot escape. It is prevented from going upwards because it finds no vent, and it cannot descend because it is lighter than the air below it. No person could long endure it without fainting; its effects are seen upon the bindings of books on high shelves; and I know of nothing for which it is good except, perhaps, for keeping cigars dry. In designing a new house I should endeavour to prevent this accumulation by providing an opening through the ceiling above a perforated centreflower, because this is at the very highest point and because it is central, and if over a gas-pendant it would prevent the vitiated air from the latter from spreading over the ceiling. After entering the centre-flower the warm air should be conducted by an air-tight passage containing as few bends as possible up to an exhaust ventilator on the roof. The type of head which I prefer to fit on the top of the pipe is a fixed exhaust ventilator, such as the Torpedo, or one of Boyle's make. Any form of revolving ventilator is less certain than these except in high winds, while a cowl to turn from the wind may fail to do so in a calm, and will sometimes stick fast. It is of no use to discharge the warm air direct into the roof or into the space between two joists, however well provided the roof may be with means of exhaust, or the joist-space with gratings through the walls. In either case you are sure at times to be troubled with a serious downdraught, and are not likely, even occasionally, to find the warm air carried off. In old houses it might in many instances be costly to fit up a tubular arrangement. Where the cost stands in the way the best substitute will be one or two non-return ventilators, such

as the Arnot or Sherringham, fixed into the chimney just below the cornice. These should not be put through the outer wall, because in such a position they are exposed to direct currents of air. I am unable to suggest any other plan that is worth trying, but wish to emphasize the statement already made that, unless some provision has been made for carrying off the warm air, there is very little good in spending money for the purpose of admitting cool air.

Next as to the admission of pure and cool air: The restriction under which this must be done is that no draught must be produced. The most urgent demand for air is that made by a fire: it must have a sufficient supply or it will burn dead and emit smoke into the room: if it gets no supply it will go out as soon as it has exhausted that which was contained in the room. As every operation is conducted by nature in the easiest available way, the demand of a fire for air, being made near the floor, is most easily supplied from the space under the doors or French casements. The result is that a draught is created round the feet and ankles of those who are sitting round the fire in the hope of obtaining warmth. A very simple arrangement will prevent this. Through the floor, just inside the rim of a fender which has no bottom plate, let a number of holes be bored into the space between two joists. Nail some perforated zinc or copper over these holes to prevent sparks from entering. Break through the wall between the two joists and fix a large grating in the hole. You will then get a copious supply of fresh air to the fire from openings so placed that the current from them will not pass your feet. But there is another advantage which is not so obvious. It is this: that a fire supplied in this way is more effective in warming the air of a room. Most of the heat derived from an ordinary fireplace is radiated heat—heat which shines out into the room and thus warms the persons, the furniture, and the air which it contains. But I have shown that the air which ordinarily supplies a fire is drawn across the room from points at some distance. It is therefore obvious that the air in front of the fire, which has already received some of the radiated heat, is constantly travelling towards the fire, and passing up the chimney, whilst its place is being supplied by another lot of cold air from the doors, &c. This process acts continuously to curtail the zone of the fire's warming influence. Now look at the contrast presented by the improved method. Under it the fire draws its supply of air from the outside through apertures close at hand and at the floor-level. It does not draw any general current across the room. At the same time, the rays of the fire penetrate and warm, as before, a quantity of air in front of itself, but which is now stationary instead of moving towards it. This air, instead of being sent

to waste up the chimney, rises as soon as it is sufficiently warm, and circulates about the room, to be replaced by cooler air from parts beyond the direct influence of the fire. By this method, then, we have supplied the fire with air without creating a draught, and have at the same time increased its effectiveness as a dispenser of warmth. We may next proceed to cut off its old and objectionable sources of supply by fitting carpet-slips on the floor close up to the underside of the doors or casements. Plate XLVIII., fig. 6, shows a section of the openings for admitting air to the fire.

Having provided for the needs of the fire, we have only to consider those of the gas and of the occupants. If we open a window at the top we get a direct draught across the room or down upon the heads of those sitting; if we open one at the bottom we get a draught across the body or neck. These methods are therefore not admissible. We must arrange to bring air in—(1) at a level at least higher than our heads when we are sitting; and (2) in such a direction as will insure its passing upwards, and not downwards or across the room. If we can do this we shall obtain a pure-air supply with the minimum of risk to health or comfort. There are two ways in which we may do it. The simpler plan in a new house is to make the lower bar of each window-sash 1in. deeper, and then to build up the sill inside 2in. higher than usual. This provides 3in. more than the customary overlap. It will then be possible to lift the lower sash 3in. without leaving a less than ordinary overlap, and therefore without making any opening through which a direct draught could pass across the room. But by this lift we have separated the dividingbars of the two sashes about 1.½in., and in such a way that the air entering at this point must flow upwards, and cannot create a draught downwards or across the room. Plate XLVIII., fig. 7, shows the section of a window-sash and sill so constructed.

In an old house a rough method may be adopted to effect this object. Fit a piece of wood 3in. deep accurately to the lower side of the window-sash and into the recess of the sill. When ventilation is desired lift the sash, place the piece of wood, and shut the sash down upon it. The inconveniences of this plan are that the window cannot be fastened without removing the filling-piece, and that the latter has to be stowed somewhere out of the way.

The other way to admit air satisfactorily is to fit Tobin tubes against the wall, and to supply them with air from the space between two joists. Where it is possible to do so a space should be selected which extends right through the house, and each end should be provided with a grating. The reason for this through passage is that if there is only one

opening into the space a wind blowing directly upon it will produce a strong rush of air up the tube, and very likely a draught in the room, whilst an opposite wind might sometimes check the proper action of gravity. The principle of the action of the Tobin tube is that of a fountain. It provides an entrance for heavy air into a reservoir of light air. The tube, being placed vertically, directs the air, which enters with a slight force, straight upwards, until its greater weight prevails and brings it gently downwards over the room in the form of the spray from a fountain. It descends so gently as not to cause a draught in the room. Plate XLVIII., fig. 8, will explain the general arrangement of one of these tubes.

It now remains to consider the measure of the various appliances for removing the warm air and supplying cool. It is not easy to lay down rules which are of general application; but I think it may be safely stated that, where ventilation fails, it is more often owing to the apertures being too small than too large. All that has to be guarded against is the production of draughts and the too rapid lowering of the tem perature by admitting the air too fast. This can be readily prevented by closing some of the openings. The provision must depend upon the number of persons likely to occupy a room at one time; upon the consecutive number of hours that it may be used without an opportunity being afforded for a thorough blow-through; upon the amount of gas consumed; and in some cases upon the position of the room in relation to its surroundings. Some writers give rules which it may be useful to apply; but it is beyond the scope of a short paper and beyond the range of my experience to attempt to criticize these or to suggest others. It may, however, afford a basis for discussion if I give a few particulars of appliances which I have recently had fitted up in a new house. I will only give these for the dining-room, which will also be largely used as a general sitting-room by eight to twelve persons. This room measures 20ft. by 16ft., and is 12ft. high. It will probably be fairly illuminated with four gas-burners, and will rarely have six lighted. It has a fireplace, two windows, one door from a passage, and two from adjoining rooms. Within the marble fender there are openings through the floor into the joist-space equal to 54 square inches in area; but this is reduced considerably by the perforated-copper gratings. On the outer wall a 12in. by 6in. grating is placed over the opening into the joist-space. There are two Tobin tubes, measuring 10in. by 2.½in. clear inside, with a fine-wire grating at the top, making an inlet of about 50 square inches. They communicate with a joist-space extending right through the house, with a grating on the outer wall at each end, 12in. by 6in. The lower sashes of both windows are fitted so that they can be raised 2.½in.,