yfr fit *~ / pt~4 BY THE SAME AUTHOR. THIRD EEVISED AND ENLARGED EDITION. "THE OPEN FIREPLACE IN ALL AGES." Quarto, with over 3OO Illustrations. PRICE, IN CLOTH, $4.00. " This is undoubtedly the most extensive and critical work on this subject in the language. The author takes up the subject of heat- ing and ventilation and treats it in extenso, showing existing de. fects, giving a full historical resume" of the theme, and ends by criticism and practical suggestions." Chicago Journal. " A very beautiful book, artistically and practically." Louisville Courier-Journal. " It is brimful of good suggestions." N. T. Graphic. "The book is excellent from cover to cover, and a real contribu- tion to useful literature." Boston Advertiser. IN PRESS. SANITARY PLUMBING. A Practical Treatise on Modern Sanitary Plumbing. 12mo, with over 400 Illustrations. Republished from the American Architect and Building News. " These investigations have been set forth quite fully in illus- trated communications to the American Architect, which papers certainly mark a very important step forward in Sanitary Litera- ture." COL. GEO. E. WAKING, JR., in the Century Magazine for December, 1884. ["LECTURES ON THE PRINCIPLES OF HOUSE DRAINAGE, DELIVERED BEFORE THE SUFFOLK DISTRICT MEDICAL SOCIETY (Section for Clinical Htdicine, Pathology, and Hygiene), AND THE BOSTON SOCIETY OF ARCHITECTS, AT THE MASS. INSTITUTE OF TECHNOLOGY. \ J. PICKERING PUTNAM, Part I reprinted from the Boston Medical and Surgical Journal. Nos. 17, 18, and 19, 1885. From the Editor of the Sanitary Jtecord, London. I5y permitting me to reproduce these " admirable articles in the Sanitary Record, you will greatly oblige me and serve the cause of Sanitary Science in England." " THE EDITOR. BOSTON : TICKNOR AND COMPANY. 1RS6 PRESS OF STANLEY AND USHER, BOSTON. THE PRINCIPLES OF HOUSE DRAINAGE. PART I. [Reprinted from the Boston Medical ami Surgical Journal.] As announced in the invitation cards to this lec- ture, a number of the appliances used for illustra- tion are inventions of my own, and some of them are patented. These devices have been clearly marked on the drawings with their name, " Sanitas," and in referring to them they will always be so designated. The object of this is that every one may know when anything of my own is referred to, and be able to assure himself that the subject is treated without bias. My intention is to make no statement which is not founded on facts recognized by all or easily demonstrated, nor to follow any course of reasoning which is not perfectly clear and logical. In order, moreover, to make our meeting as satisfactory and fruitful as possible, it is hoped that every one will make a note of any point which may not be perfectly clear to him, or of any deductions with which he may not be fully in accord, and mention them in the discussion follow- ing the lecture, so that the reasoning leading to these deductions may be reexamined or more clearly stated. It is proper to add that the devices are the out- growth of a careful practical study of plumbing made from the unprejudiced standpoint of the architect working for the interest of his client. 4 THE PRINCIPLES OF They are the result, and not the cause, of the investigations. As plumbing is now practised, the architect or the sanitary engineer is, from the nature of his work, the one upon whom we must depend chiefly for its improvement. Evidently the most important part of an architect's work is that which concerns the health and comfort of his client. The arrangement of the plumbing pipes and fixtures influences the entire plan from foundation to roof. Some of the rooms, such as the laundry and toilet rooms, are designed exclusively for the plumb- ing, and all are more or less dependent upon its arrangement. The walls and beams must be slot- ted and framed for its reception, and differently for each different kind of fixture or system of piping and for their lighting and ventilating apparatus. Hence, the architect must be familiar with all the details of the work, and upon him lies the responsi- bility, not only for the healthfulness, convenience, and cost of the particular work over which he has immediate charge, but also, in a great measure, for the general status of the art of plumbing through- out the countrv. The plumber, like the architect, may see defects in the methods of plumbing now in vogue, but he has comparatively little interest in promoting reform because the authority lies with the architect.* The plumber has contracted to do a certain amount of work for a certain amount of money and it is not easy for him to alter the con- tract. If he is directed to put into a house a few hundred dollars' worth of piping more than * Some architectural linns now employ regularly a Sanitary En- gineer, eithei directly or indirectly connected with their offices, to take charge, in cooperation with the architect, of the department of heating, ventilation, and plumbing. The custom has much to recom- mend in it. HOUSE DRAINAGE. 5 is necessary or desirable, or set complicated or ill-devised fixtures, it is not his duty to protest. Competition has reduced his profits to so small a figure that the plumber cannot afford to be an active reformer. The physician seldom interferes in the details of plumbing work, and the public are, as a rule, pro- foundly ignorant of them. The architect, however, acting directly asthe agent of the owner, is bound, when he discovers defects affecting the health as well as the pocket of his client, to use all the influence he possesses to remove them. Every sanitarian recognizes the serious defects of our present plumbing methods and apparatus. Our common soil and waste pipes are neither gas nor water tight on account of their defective form and jointing, and they are expensive to lay. Our traps either are incapable of retaining their water-seal against the adverse influences affecting them in common use, or they are bulky, unscientific, and expensive. Our lavatories are slow emptying, incon- venient, and complicated. Most of the water- closets in use are full of defects and flushed on incorrect principles, and, in short, there is scarcely a single point in which our plumbing is not evidently susceptible of great improvement. Let us examine these defects and study the prin- ciples which should guide us in effecting their cure : Simplicity. The tendency at present is toward undue complication. The plumbing work is becom- ing each year more elaborate and costly, more difficult to set correctly, and more difficult to com- prehend and repair when correctly set, so that the public are becoming alarmed and confused. They 6 THE PRINCIPLES OV despair of being able to understand the intricate system of piping and machinery for the supply and waste of fixtures. The result is a general feeling of insecurity and a tendency to forego the conven- ience of plumbing fixtures wherever their presence is not an absolute necessity. Our byword should be "simplicity." Rather than reduce the number of our fixtures, let us reduce the amount of machinery connected with them, pro- vided we can do so without diminishing the security they are intended to afford. If we find that our process of simplification actually increases the security of the work, all will be gainers the public as well as the plumber. For what is best for the public by increasing their confidence is also best for the plumbers, though they appear often to lose sight of this fact. Accessibility. Another leading principle is that all plumbing work in a house should be everywhere, without exception, accessible and as far as possible visible. Pipes should never run behind plaster when it is possible to expose them on walls and ceilings. The pipes both waste and supply as well as the bathtub traps of a bathroom should be placed, if possible, on the ceiling underneath the plaster of the bath- room or china-closet below never between the floor- joists. There is nothing in a neatly arranged line of lead, brass, and iron piping that one needs to be ashamed of. On the contrary when skilfully placed and neatly jointed in a workmanlike manner, as would be the case when the plumber knew they were to be forever exposed to view, these bright metal pipes become quite ornamental when mounted with lead or brass clamps on strips of finished woodwork varnished and symmetricallv arranged in corners or where good taste and judgment direct. HOUSE DRAINAGE. 7 In the darker ages of architectural art, chimneys were despised and hidden from view. Now they become the most prominent features of a design, suggesting hospitable comfort and healthful ventila- tion within. So should it be with the piping. A knowledge on the part of the house-owner that all the pipes which provide him with the comforts of pure water and safely carry off the foul are in full view and in a sound condition will afford him much solid satisfaction. Avoidance of Mechanical Obstructions. A third principle is to avoid all mechanical obstructions, such as valves, balls, gates, and all other impediments to the water-way, and in a system of water-carriage to do all trapping by means of a water-seal alone. Mechanical devices form no reliable security against the passage of sewer gas. These valves and balls cannot be made to fit their seats with such accuracy as to exclude liquids and gases, or micro- scopic disease-germs, even when new. They soon become more or less fouled with dirt and corrosion and then their inefficiency becomes evident even to the eye. A sound water-seal, however, properly protected, is found to be entirely re- liable in excluding noxious matters of all kinds. Fig. 1 represents a trap having the undesirable mechanical seal in the form of a ball. Moreover we are obliged to rely upon a simple water-seal whether we desire to or not, because our water-closet j Thejen . traps or their overflows are , and must nings trap, with be, constructed without mechanical mec1mnical 8eaL obstructions. Evidently if the water-seal is in- efficient we must either construct our water-closet 8 THE PRINCIPLES OF traps and their overflows on a different principle or else give up the system of water-carriage alto- gether. It is useless to apply mechanical closures to our smaller traps if we leave the large water- closet traps without them. It has been shown by Dr. Carmichael and others that if a water-seal be properly maintained against evaporation and siphonage, or destruction from any cause, the amount of sewer gas that can pass through in twenty-four hours, even under the worst conditions, but with a ventilated soil-pipe, is infini- tesimal and absolutely harmless, and that disease- germs cannot pass at all through water at rest at normal temperatures. Dr. Carmichael also experi- mented with an unventilated soil-pipe, and found here that the quantity of carbonic-acid gas, the largest component of sewer gas, given off from the traps in twenty-four hours was less than that obtained " when a bottle of lemonade was opened," and less than that which is exhaled by a man in five minutes. As for the ammonia, sulphuretted hvdrogen, and other gases or vapors which accompany the carbonic acid, their combined amount, even under the un- favorable conditions of a foul sewer and unventila- ted soil-pipe, was hardly equal to the one thou- sandth part of that of carbonic acid, and this amount diffused in twenty-four hours through the atmosphere of a house is evidently absolutely insig- nificant and harmless. With a ventilated soil- pipe the quantity which can pass through the water- seal was found to be about four times less, proba- bly far less than what would come into our city houses through the doors and windows from the ventilating openings in the streets of the public sewers. HOUSE DRAINAGE. Drs. Carmichael, Pumpelly and Smith, Naegeli, Wernich, Miquel, and others have shown that disease-germs and bacteria generally have the same " mechanical affinity " for water which we observe in all solid particles, particularlv of organic nature. They cannot rise spontaneously from the surface of water at rest, and at the normal temperature of our houses. It is only when the surface is violently agitated, or when gaseous bubbles rise to the top and burst, that these particles are released and dissipated in the atmosphere. With a ventilated soil-pipe no such effervescence in the water of a trap can take place, and the agitation of its surface caused by properly arranged flushing does not throw water out of the trap nor allow of the escape of any germs of disease, for any water which may be washed up on the sides of the trap above the normal water-line is quietly carried down again by the upper flushing stream and swept into the sewer. The experiments of Dr. Carmichael resemble absolute demonstrations and may be accepted as conclusive. He concludes his report as follows : " Water-traps are, therefore, for the purpose for which they are employed, that is, for the exclusion from houses of injurious substances contained in the soil-pipe, perfectly trustworthy. They exclude the soil-pipe atmosphere to such an extent that what escapes through the water is so little in amount and so purified by infiltration as to be perfectly harmless, and they exclude entirely all germs and particles, including, without doubt, the specific germs or contagia of disease, which we have already seen are, so far as known, distinctly particulate." Tightness of Joints. A fourth principle is that all joints should be permanently tight, and to secure this evident desideratum no material should be used 10 THE PRINCIPLES OF in jointing which is injuriously affected by any of the substances brought in contact with them or by movement produced by changes of temperature, concussion, or shrinkage. Soundness of Material. A fifth principle is that all the material used be sound, and all pipes of even thickness and capable of resisting a suitable pressure-test. Ventilation. A sixth principle is that all the main lines of soil and drain pipes be thoroughly ventilated from end to end. Flushing. A seventh principle is that all parts of the waste receptacles and pipes be thoroughly flushed with water from end to end in such a man- ner as to remove all foul matter instantly from the house as soon as it is generated. Automatic Operation. An eighth principle is that the working of all parts of the plumbing system should be as far as possible automatic. Noiselessness. A ninth principle is that the operation of all parts of the work should be noise- less. Economy and Prevention of Water-waste. Finally, all parts of the work should be economical in con- struction and designed in such a manner as to avoid the chances of waste of water through leakage. These ten broad principles are not only accepted by all the leading sanitarians, but are self-evident and may be at once adopted as axioms without dis- cussion. In the manner of applying them in prac- tice, however, we do not find the same universal harmony. Where all are in accord I shall make no reference to authorities. But where there is a dif- ference of opinion among experts I shall call atten- tion to the fact, so that each one present may form an independent judgment of his own. HOUSE DRAINAGE. 11 The first subject we shall consider will be the trap. Its form depends upon the nature of the work it is called upon to do ; the form which is suitable for a water-closet being quite unsuitable for other fixtures. The agencies which tend to destroy the water- seal and efficiency of traps are : siphonage, evapo- ration, back pressure, capillary attraction, self- siphonage, leakage, and the accumulation of sedi- ment. These agencies must therefore all be con- sidered in the design of our trap. What form shall we give it to enable it successfully to with- stand them ? We find that if we adopt the simplest possible form, that of the S trap, which con- sists merely of a bend in the pipe deep enough to make a seal, we obtain a device which, with proper flushing, is suffi- ciently self-cleansing and fur- nishes the easiest outlet for the water. But it is unable to do any more without ex- ternal aid, and quickly loses its seal under the slightest dis- Fi s- 2 - Ordinary s trap, turbance of atmospheric pressure produced by a sudden flow of water through the pipes with which it is connected. Three methods have been employed with a view to preventing the destruction of the seal by siphon- age. One is to ventilate each trap by connecting it with a special ventilating pipe constructed for the pur- pose. 12 THE PRINCIPLES OF A second is to increase the size of the upcast limb of the trap until it becomes a " pot," or " reservoir," trap large enough to accomplish the same result without external aid. A third method is to construct the trap in such a manner as to render it both autisiphonic and self- cleansing at the same time. The first method adds greatly to the cost and complication of the work. It has given rise to the so-called "trap-vent" law, which rigidly requires every trap, under all circumstances, to be venti- lated. In regard to the practical working of trap venti- lation two things have been found : First, that it is not always efficient in preventing siphonage. Second, that it is always more or less active in destroying the seal through evaporation. Nevertheless, this method still has a few advocates of recognized ability. But they now adhere to it chiefly, if not entirely, on account of an alleged indirect advantage produced by the air-current in partially oxidizing foulness in the waste-pipes. The second method is both inexpensive and simple and is much more efficient in resisting siphonic action than the first. It has, however, the serious disadvantage of involving the use of cess- pools or filth-retainers in the house, and such re- tention is in violation of a leading principle of sani- tary drainage which calls for complete removal of foul matters from the premises the instant they are generated. This method has however a very large number of advocates who consider the retention of a limited quantity of filth in the trap less of an evil than the dangers of difficulties coming from trap-venting. HOUSE DRAINAGE. 13 They claim that a guard which is only sometimes reliable is worse than none at all as giving a false sense of security, and that the purification of the branch waste-pipes can be effectually accomplished by powerful water-flushing, making the induction of the air-current for this purpose quite superfluous. They find, moreover, that abundant aeration goes on without the aid of the vent-pipe both from diffusion of the air in the ventilated soil-pipe, and from the powerful influx of air induced with, and after, the water-flushing at each usage of the fix- ture. The third method is the simplest and least ex- pensive of all. It is more reliable than either of the others in resisting siphonic action, and does away with the serious objection of the second method : that of filth retention. It has already the advocacy of many of the lead- ing sanitarians gf the country and promises to be universally adopted as soon as it becomes generally known. Let us now examine these three methods carefully in detail, since the question is not only one of the most important and interesting ones in the whole domain of sanitary plumbing, but its investigation will throw light on every other part of the subject. Trap-ventilating. Until very lately it was supposed that trap ventilation afforded a reliable cure for siphonage, and under that supposition the trap-vent law was made. This law has been in operation but a very few months and in a few large cities, yet it has been in force long enough to show in the first place that it is by no means able to do what it pretends to even when the vent is newly and skilfully applied, and in the second place that it gives rise to new evils as great or greater than 14 THE PRINCIPLES OF those it was intended to obviate, and in the third place that the vent-pipe itself tends to become foul in usage, and that in some cases the accumulation of foulness goes on to such an extent, especially at its point of connection with the trap, as to com- pletely close the air-passage and destroy its oper- ation. We will first test the efficiency of the trap-vent when it is new and clean and afterward consider the question of its partial or complete closure by filth accumulation. Tests on Traps Newly Ventilated. In making these tests two points have been very carefully followed, and these must be distinctly understood. In the first place the apparatus and arrangement used is precisely the same in character as is found most commonly in the best ordinary practice. A large number of return bends and a very long stack of piping has been put together in order to permit a variety of different tests to be made with a single compact apparatus. But as openings have been made in the pipe at various points, we are able to cut off one or all of the bends and any part of the length of pipe at will. Hence, the apparatus may be made to correspond with that in any form of house we desire to imitate. In the second place, though our tests will be very severe, they will, nevertheless, be no more so than is often encountered in practice. Our object is not to show what usually takes place but what at any time may take place. We do not of course pretend to say that a new vent-pipe can never protect a trap, but that it can- not always be relied upon, and that this being the case, it affords a false sense of security, and is therefore worse than nothing, for we can never tell at what moment it will fail in its duty. HOUSE DRAINAGE. 15 If we are to be forced by the law to put our clients to the great expense and danger of ventilat- Fijf. 3. Apparatus used for trap testing. ing every trap, we have a right to demand : first, that the means employed shall actually afford us the se- curity it pretends to, and not fail at the first critical 16 THE PRINCIPLES OF moment, and second, that no other simpler and better means exist for securing the desired results. Our apparatus consists of a stack of four-inch soil-pipe with two ordinary plumbing fixtures ten feet above the wastes of the traps to be tested. On the left is a Jennings closet and on the right a Zane : kinds which have been perhaps until lately the most popular in this country. Fig. 3 represents the apparatus used. The dis- tance from the floor to the upper platform which supports the water-closets is fourteen and a half feet. From the floor to the ceiling is seventeen feet. It is hardly necessary to explain to the present audience that the smooth bends and returns we have used add but very slightly to the friction. Smooth bends of a radius equal to, or greater than, the diameter of the pipe are found to have very little effect in retarding the flow of fluids. To show just the effect such bends produce in the present case we have provided openings at different points in the length of the piping and after experi- menting with all the bends, we will make other tests without them or with only a portion of them, and compare the results. We have used the ordinary two-inch iron pipe for back ventilation. Just above the floor of this room we have pro- vided two Y branches for the trap waste-pipes to be tested, using ordinary one and a half -inch bath or basin lead waste-pipe of the average length. We find the length of these branches within reasonable limits does not appreciably affect the siphoning action. Experiments on a one and a half-inch S trap. The first test we will make on an ordinary one and a half- inch S trap unventilated. The soil-pipe we leave full length. The seal is two inches deep. HOUSE DRAINAGE. 17 (Discharge of Z and J together.) We see that a single discharge of the two closets has completely destroyed the seal in a second, leaving scarcely a drop of water in the trap. If we shorten the soil-pipe one half by removing the plug at forty-five feet length, we find substan- tially the same result. (Discharge of Z and J together.) The seal is again instantly destroyed. This is the simplest possible illustration of the phenomena of trap siphon age, and so far the result is probably familiar to most of the audience here to-night. All are aware that an unventilated lava- tory S trap with even an unusually deep seal pos- sesses scarcely any power to resist siphonage. When the falling water in the soil-pipe produces the partial vacuum behind it as it descends, if the soil-pipe extension above it is short and closed at the top, the action is at its maximum because there is very little air to expand. If the pipe is short and open it is at its minimum. If it is long and closed still the action is powerful, but if it is long and open above, a medium effect is produced, and this is the condition we have to-night in our apparatus. Let us next see what the effect of a discharge of a single closet will be, leaving the soil-pipe forty-five feet long. This cuts off three bends, leaving it four. (Discharge of Z alone.) The Zane closet alone has siphoned out the trap in a single discharge. Let us try the Jennings alone. (Discharge of J alone.) The Jennings alone has also instantly unsealed the trap. Thus we see that either a Zane or a Jennings plunger-closet is easily able to destroy the seal of an ordinary S trap under the simplest conditions of 18 THE PRINCIPLES OF plumbing. Any other form of plunger-closet or any valve or properly constructed hopper-closet would probably do the 'same. Let us now ventilate our trap with a vent-pipe the full size of the bore of the trap. Leaving the soil and vent pipes the full length we will discharge the closets as before. (Discharge of Z and J.) The first discharge has reduced the seal from two inches to one and a half inch. It will be observed that our vent-pipe is actually considerably larger than the bore of the pipe which is contracted at the bends, so that the protection afforded by this is greater than it would be in ordinary practice. Another dischai'ge has lowered the seal to one fourth of an inch. A third discharge has completely de- stroyed the seal, leaving an open passageway into the house for sewer gas. Thus we see that with the long stack of pipe our ventilation has signallv failed. We will now cut off half the bends and half the length both of soil and vent pipe, leaving a medium length of each of forty-five feet. A discharge of the closets has lowered the seal to one and one-half inch. Four discharges have destroyed the seal. At the last lecture we found it possible to break the seal by discharging only one of the closets at a time. But it required eight repetitions of the dis- charge to do this, and we shall accordingly omit the experiment to-night for want of time and because our subject extends over a wider field than before. Our next experiment will be with an ordinary one and one-fourth inch vent-pipe which is really the size of the trap under consideration. Omitting the tests with a medium length of soil and vent pipe HOUSE DRAINAGE. 19 which broke the seal of this trap in two discharges, and also the test with a single closet, we will shorten the ventilation-pipe to fifteen feet by cutting off the two-inch iron pipe and all its bends alto- gether. This gives us a shorter vent-pipe than we should ever be likely to encounter in practice. Hence, if the friction produced in this short length of pipe is enough to prevent the effectiveness of the vent, anything longer than this would certainly de- stroy its power. The soil-pipe is of medium length. Discharge ofZ and J. Four simultaneous discharges of the closets have destroyed the seal of our trap, fully vented with a new in the manner required by the law, showing our expensive venting to be utterly untrustworthy, even under the simplest conditions. In the experiments made for the City Board of Health the same results were obtained by the dis- charge of a single plunger-closet. Four-inch by four-inch Y. We have, up to this time, used a four-inch by two-inch Y branch to connect our lead branch with the main soil-pipe. In our experiments for the Board of Health we were severely criticized by The Sanitary Engineer for using a four-inch by four-inch Y branch, which, we were told, would produce an action at least four times as powerful as the smaller branch. In order to test this we will connect our waste with a four-inch by four-inch branch immediately below the one we have been using, and repeat the last test under the new conditions. I would caution those of the audience who are seated nearest the trap to hold firmly to their seats, which have been tightly screwed to the floor in order to prevent them from being sucked bodily into the drains by the pro- digious siphonage power of our four by four-inch branch. 20 THE PRINCIPLES OF (Discharge of Z and 7.) We find 110 appreciable cliff ereuce in the two Ys, and I think those gentle- men can now safely release their hold upon the furniture. We have records of comparative tests made with two such Ys, made in exactly the same position on the apparatus, showing a greater rather than a feebler action produced by the smaller Y. Experiments with a Partially Clogged Vent-pipe. When the mouth of the vent-pipe has been partially closed by gradual deposit of sediment, the supply of air through it is proportionally retarded, and it be- comes less and less of a safeguard against siphonage. We have made a great many experiments in this field and found the resistance exactly proportioned to the size of the vent-pipe. In the tests for the Board of Health we used a straight stack of pipes without any bends. The siphouic action was somewhat more severe in all the tests. Secondary Office of the Vent-pipe. It remains now to examine the secondary office of the trap vent-pipe, namely, the aeration of the branch waste- pipes, pi'omotiug decomposition in them, and carry- ing off the gaseous products of such decomposi- tion. Some years ago, before it became customary to ventilate all the main lines of soil and waste pipe, as all sanitary engineers are agreed in recommend- ing now, there accumulated in the upper part of the pipe-system large volumes of dangerous and corro- sive gas generated by the decomposition of the heavy deposits in the large soil-pipes throughout their entire length. These gases, never liberated as they are now by a constant current of fresh air passing through the main pipes, sometimes formed HOUSE DRAINAGE. 21 in such large quantities as to eat through the metal and escape into the house. The water-flushing from the feeble pan-closets of that time was quite insufficient to purify the main-pipe lines, and seri- ous difficulties arose. Now, however, the case is very different. All our main pipes are thoroughly ventilated, and a far more liberal flushing is occasioned by the use of modern hopper-closets. This comparatively fresh air of the soil-pipe distributes itself by diffusion through the branch wastes, and gases can no longer collect to any harmful degree unless they are of unusual length and insufficiently flushed. Consider the case of a short-branch waste-pipe leading from a well-constructed washbasin and con- nected with a well- ventilated soil-pipe. Fresh air is constantly passing through the soil- pipe, carrying off the products of combustion as fast as they are formed. If the lavatory be fre- quently used and properly constructed the short- branch waste-pipe is scoured from end to end and kept very free from foul matter. Fresh air is dif- fused easily from the soil-pipe through this short branch as far up as to the trap. If the fixture is rarely used the last thin deposit of soap dries up on the sides of the pipe, and what little decomposition goes on then is inappreciable, and the products are removed by diffusion, or, if they are absorbed in the water of the trap, what could escape from its surface would, as we have seen, be absolutely infin- itesimal and harmless. Not so if we ventilate this short-branch waste, as now required by law. A few days is sufficient to evaporate out all the water from the trap, and soil-pipe air may then enter the house freely. This is no careless assertion founded on theory. It is the result of a series of very careful 22 THE PRINCIPLES OF experiments made by myself, and published in the sanitary journals, and it is the experience of ex- perts who have examined the working of the trap- vent law during the short period since its enforce- ment. Consider next the long-branch wastes of lavato- ries. We will suppose the fixture to be a washbasin or bathtub used every day. If the outlet be prop- erly constructed the discharge of the fixture will fill the pipe so full as completely to drive out the air that was in it and fill it with a volume of perfectly fresh air from the room. Every one has observed fresh air being sucked into the outlet of a lavatory, at the time of discharge, in volume sufficient to renew the air of the branch waste-pipe many times over, even with basins improperly constructed as they are. We will suppose the fixture to be very seldom used, say not oftener than once a week or month, as in a spare room. The last charge of water passes off and the pipe dries up. I believe that what de- composition would then go on in pipes connected with a properly constructed lavatory would be utter- ly harmless ; and more than this, I believe there is no case on record of harmful corrosion ever being found on such branch wastes. It certainly would not do to ventilate the trap of such a fixture left in periodical disuse ; for evaporation would unseal its trap in its intervals of rest, and far greater damage would arise than could come from the unventilated pipe. Those who do not possess this degree of con- fidence have only to arrange their fixtures in such a way as to avoid long-branch wastes, and the diffi- culty will for them be avoided. Consider now the question of branch wastes from kitchen and pantry sinks. Every one knows that grease and sediment from these fixtures will at once HOUSE DRAINAGE. 23 clog up in time any part of a trap not scoured by the water. We find the upper part of ordinary pot-traps always fouled with grease in such cases. The mouth of the vent-pipe taken from the top of such a trap also becomes similarly clogged, and it is probable that in whatever wav the vent-pipe be attached to the trap of a sink it will surely become clogged and inoperative in time. The only case in which trap ventilation can be recommended, as it seems to me, is in connection with certain kinds of water-closets. The consider- ation of this branch of my subject must be left for another time. I find, therefore, no advantage whatever in trap ventilation, with the above possible exception. There are several disadvantages, which, summed up briefly, are these : First, it destroys the seal by evaporation when ordinary S traps are used and when the vent-pipe is taken from the crown, as the law in some places requires. I find that if the vent- pipe be taken from some points six inches or more below the crown evaporation does not go on, or it goes on so slowly as to be harmless with traps hold- ing a reasonably large body of water. With S traps, however, it is necessary to ventilate at the crown, if they should be used at all, in order to prevent self-siphonage. Second. The vent-pipe does not accomplish its objects, and hence affords a false sense of security. Third. It increases the unsecured area of the trap, making it a cesspool. The ventilated S trap is used instead of a reservoir-trap by the advo- cates of trap ventilation for the sake of avoiding an unsecured chamber. But in doing so they add a sediment chamber, which is not only greater in extent of surface, more easily fouled and less easily THE PRINCIPLES OF cleansed, than that in the pot-trap, but one which is far more dangerous, inasmuch as its fouling, even to a limited extent, involves the destruction of the whole system. This chamber is as certain to become inoperative after more or less use as is any reservoir or cesspool in a trap to become clogged with deposit. It is so placed and of such a form that it must inevitably receive spat- terings from the filth-laden waste- water, without benefiting by its scour. I have found, by repeated test, that the water discharged from a washbasin with a large out- let and trap placed a foot or more below is thrown up over ten inches into the vent-flue at every dis- charge. Thus a very large sedi- ment chamber is formed. The de- Fig. 4. Trap vent posit of sediment may be rapid or clogged at the mouth, glow, according to circumstances, in some cases requiring years to reduce the size of the vent-opening to the point of inefficiency. In others this will occur in a few days. Fourth. It retards the outflow of the waste- water about thirty-three per cent. This is owing to the friction of the air-current entering with the water during the discharge. Fifth. It renders the discharge noisy. The same air-suction which delays the water produces a disagreeable roar when the water discharges rapidly. Sixth. It complicates the plumbing and adds to the danger of leakage through bad jointing and increased material. Seventh. It aggravates the danger arising from capillary attraction ; and, finally, HOUSE DRAINAGE. Eighth. It seriously increases the cost of plumb- ing, an increase which amounts to as much as from five to ten per cent, on the total cost of the Fig. 5. Complexity with insecurity. plumbing in new work and indefinitely in old work in which the trap ventilation sometimes becomes by far the greatest part of the work to be done. Fig. 6. Simplicity with security. Figs. 5 and 6 represent three fixtures plumbed by the two different methods, the first with, and the second without, trap ventilation. In the first draw- 26 THE PRINCIPLES OP ing the overflow-passage and the house-side of the trap are ventilated as well as the sewer-side, and the loss of the water-seal through evaporation is very rapid. This double trap ventilation is not common, but vet is occasionally carried into execu- tion by some of our more radical enthusiasts for branch- waste venting. In the second drawing " Sanitas " traps are used which require no ventilation to prevent siphonage. A washbasin, having an outlet large enough to fill the waste-pipe and trap " full-bore," scours them and keeps them free from deposit. The use of a urinal is rarely to be recommended. It is only introduced here for purposes of illustra- tion. Second Method. Let us now examine the second method of obtaining security against siphonage. This consists in the use of a large unventilated pot, or reservoir, trap. A small pot-trap will not resist siphonic action, but a large one will. Their power of resistance is exact!}' in proportion to their size. Nothing smaller than an eight-inch pot- trap, which I have here, can be relied upon in all cases. A six-inch pot-trap will sometimes be siphoned out by discharges occurring in common practice. A five-inch pot-trap siphons out much easier. An ordinary four-inch trap has very little resisting power unless its seal is unusually deep. Three-inch and two-inch traps are altogether useless. We will test practically the action of siphonage on a four-inch pot-trap of the usual depth of seal. (Discharge of Z and J. ) We see that three dis- charges are sufficient to break the seal of this trap. In our last lecture we found that either the Zane or the Jennings closet alone was able to destroy the seal in eight and six discharges respectively. HOUSE DRAINAGE. 27 Thus we see that only the largest sizes of pot- traps are reliable. To be secure in all cases, if we use pot- traps, we are required to have them as much as six or eight inches in diameter, and con- stantly inspect them to see that they are free from deposit. Traps of this size are veritable cesspools and as such are to be avoided wherever it is pos- sible. They are, moreover, expensive. A plumb- er's scale of charges for these traps is at the rate of one dollar for every inch in the diameter of the trap. Thus a five-inch, six-inch, and eight-inch pot-trap costs $5, $6, and $8 respectively. Fig. 7. D trap clogged. Fig. The pot-trap is, morever, bulky and unscientific in construction. Its cleanout cap is faultily ar- ranged at the top, where, if improperly adjusted, it will allow the escape of sewer gas without warning. The cleanout cap of a trap should always be wholly or in part below the normal level of the standing water in order that if an unsound joint occur it will at once be detected by an escape of water and the defect remedied. It is better to en- 28 THE PRINCIPLES OF danger the floors or plastering than the life or health of the owner. To ensure tightness the plumber is obliged to screw the cap on so hard that the house-owner is rarely able to unscrew it for examination or cleans- ing. Hence the plumber has to be sent for. Ill- feeling is aroused and the plumber is referred to in terms often lacking in refinement and politeness. Under the name " reservoir " traps I include all water-seal traps which are not self -scouring. It includes the old-fashioned D trap (Fig. 7), the Globe trap, and the Bottle trap (Fig. 8). Of all the reservoir-traps, the common pot- trap, bad as it is, is the best, as being the simplest. Balls, valves, and gates in traps add little or nothing to their power of resisting siphonage, and have no longer any value now that it is customary to ventilate the drain and soil pipes. They serve only as encumbrances and filth collectors. We come finally to the third method of obtaining security against sewer gas, of which Fig. 6 forms the general illustration. It is to give the trap such a form as to render it antisiphonic and self- cleansing at the same time. ANTI-SIPHONIC TRAPS. Let us first examine the action of fluids in traps when they are subjected to siphoning action and see if it is possible to construct a trap in such a man- ner as to accomplish these results. To better study the movement of the fluids we have had a number of S and pot traps constructed wholly of glass. We must make use of the natural forces at our command, the superior gravity and adhesive force of water over air, and construct our trap with refer- ence to the laws governing these forces in the move- ment of the two fluids. HOUSE DRAINAGE. Examining first our pot-trap. Under a powerful siphonage air is driven through the water in the body of the trap in the manner shown in this draw- ing (Fig. 9). A quantity of water is projected out of the trap in advance of the air-column, as shown by the arrows. If the action were continued long enough all the water above the inlet-mouth, even in Fig. 9. Movement of fluids in a pot-trap. the largest pot- traps, would be expelled. It will be observed that part of the water is forcibly thrown up against the top of the body of the trap, whence it is deflected back in the form of spray in all direc- tions. Part of the spray, however, falls across the outlet-mouth, and is sucked out. One of the prin- cipal reasons why the S trap is so easily siphoned out is that the curve at the top conducts the water directly into the outlet. Some form of reflecting surface should be used to throw the water back into the trap, and let the lighter air escape to supply the vacuum in the soil-pipe. Such a reflecting surface 30 THE PRINCIPLES OF is found in the flat top of the pot-trap above the outlet-mouth. We will therefore retain this useful feature, but reject the objectionable one of the ex- cess of sectional area in the body over that of the inlet and outlet arms, and we have our first modifi- cation, as shown in Fig. 10. The reflecting surface, however, should not be ar- ranged as here shown. The pocket increases the un- secured area of the trap. It is true it is no worse than the mouth of a ventilating pipe, which under the present law it is customary to put at this place. But it is just as certain that such a pocket will become clogged in time as it is that grease and filth will deposit a sediment on every- thing with which it comes in contact. The higher or deeper Fig. 10. First modification. the pocket the more rea dily will the deposit be formed. A shallow pocket might be partially scoured by the force of the water pro- jected upward against it by momentum. In this case a certain portion of each deposit of filth would be washed off by friction and the process of clog- ging would be somewhat retarded. But let the pocket be deep enough and there will then be parts which will be within the reach of the waste-water, but beyond its scouring effect. The spray thrown up by momentum will at this height have lost its power. The drops of dirty water will simply rise to their turning-point, deposit their filth, and trickle back again into the trap. The ventilating outlet forms exactly such a pocket. At a certain height BOUSE DRAINAGE. above the crown of the trap the inner surface of this flue will receive the spatterings of the filth-laden waste-water, but never receive its scour. Hence the area of the vent-opening must infallibly con- tinue to decrease in size more or less quickly, ac- cording to the usage of the fixture, until the open- ing is too contracted to be of any value in resisting the action of siphonage on the water-seal. More- over, the cool ventilating draught helps to congeal the fatty vapors arising from hot waste-water in the trap and hastens clogging. We will, therefore, simply retain the reflecting sur- face but reject the pocket. Fur- thermore, we will slightly contract the inlet and outlet mouths at their junction with the body. This allows the air rushing through the body of the trap to pass through the water instead of driving it out before it. A very slight contraction is suffi- cient. These two modifications make the second step in our improvement, and are shown in Fig. 11. A trap was constructed in this manner, and proved to be Fi Secondstep . very much stronger in resisting Partial contraction of siphonic action than an S trap of ffi'^dSSt'IT'S equal depth of seal. pocket. Still our trap is very far from antisiphonic. Re- ferring to our glass pot- trap, we shall see that the water projected violently upward from the surface, by the air-bubbles rushing through the standing water under the influence of siphonage, is obliged to pass twice by the mouth of the outlet-pipe, once before and once after reflection against the THE PRINCIPLES OP top, and that it is at these moments that it is sucked out and lost. That part of the spray which happens to be thrown farthest from the mouth of the outlet-pipe will be seen to fall back safely into the trap ; but that which passes near this outlet, either in rising or after re- flection, is drawn out by the concentrated and powerful suction at this point and wasted. And we find that one of the principal reasons why a large pot-trap resists siphonage longer than a small one is that in the large trap the spray has more space above the surface of the standing water than in the small one, so that a smaller proportion of the water thrown up by the rushing air-bubbles passes within the influence of the suction at the outlet-pipe. If our reflecting surface could be placed below instead of beyond the mouth of the outlet this loss could be avoided. Our next step must, therefore, consist in so placing the reflecting surface. In Fig. 12 this has been ac- complished, but in an awkward manner. Before this surface can come into service the level of the water must evidently be reduced to the level shown by the shading in the figure. Hence the perpendicular part of the body of the trap above the lower reflecting surface is not placed to ad vantage. Never- theless, this trap will resist a Fig. 12. Third step, very powerful siphonic action, even as it is. The two reflecting surfaces, the lower and the upper, are so effective that this form of the trap has proved more tenacious of its last inch or HOUSE DRAINAGE. 33 two of seal than a four-inch pot-trap, although its diameter is nowhere greater than that of the outlet and inlet pipes. In this and in the preceding forms the depth of seal is too great to allow of a free and rapid dis- charge of the wastes. The air, in passing through the trap, disturbs nearly all the water in it. Our next step will therefore be to diminish the height of the water-column through which the air has to pass, and thus reduce the disturbance of the water with- out lessening its volume. It may be done by laying the body of the trap horizontal instead of perpendi- cular, as shown in Fig. 13. This immediately Fig. 13. Fourth step. gives us a very important improvement in resisting power. The area of the trap is no greater than that in Fig. 11, but it is found to offer double the resistance to siphonage. Moreover, while the volume of water is the same as in Fig. 10, the seal is not so deep. Hence the flow of water through this trap is more rapid than in the former, and its scouring effect correspondingly increased. As soon as the water in this trap has been lowered to the point indicated in the drawing, ample space is left above it for the passage of the air. It is evident that a much smaller body of water is dis- turbed by the passage of the air than is the case with the trap shown in Fig. 11. 34 THE PRINCIPLES OF Nevertheless, the trap thus made is not yet suffi- ciently antisiphonic. It is, moreover, awkward in form and difficult to set in such a manner that it shall remain firm in place. The long horizontal body is liable to sag and lose its form. Moreover, a single reflecting surface is insufficient to separate the water entirely from the air, and a strong and long-continued siphonic action destroys its seal. Other improvements are evidently necessary. A fifth step consists in increasing the number of reflecting surfaces, and in breaking up the long horizontal body by making it return upon itself in a quad- rangle, as shown in perspective in Fig. 14. In this form of the trap we have still further greatly increased the reflecting surfaces and the power of resist- ance to siphonic action, and we are now able to dispense with reflecting pockets, but we have obtained a trap exceedingly difficult to manufacture, awkward in appearance, and troublesome to clean out in case of accident, as when a match or any such foreign substance is dropped into the waste-pipe and becomes lodged in a bend of the trap. This form of trap must be simplified so as to render it practical, without losing any of the advantages we have thus far arrrived at. Figs. 15 to 19 show the manner in which this may be done, and the arrangement forms the final step of our improvement. We have here retained all the reflecting surfaces ; the horizontal body, which allows the air to pass above the water after a small quantity has been driven out, without disturbing the rest ; and the slight contraction of the inlet and out- HOUSE DRAINAGE. 35 let pipes at their junction with the body of the trap. We have added a cylindrical cleanout cap of glass, and obtained an apparatus which can be readily cast in lead in moulds of iron. The quad- rangular shape of the horizontal body is retained, but the two parallel cylinders are brought to- gether and merged into a single cylinder having a central partition about two thirds of its length, or extending from one end to the edge of the cleanout cap, which at the other end forms about one third of the total length of the cylinder. In ordinary use the waste-water passes through this trap in such Fig. is. sixth step, a manner as to act to the best " Sanitas " trap, advantage in scouring it. The partition wall in"the centre of the body causes the water to scour each side in succession. Thus while in outward appear- ance the body resembles a small pot-trap placed horizontally it has in principle the self-scouring form of the S trap. It must be understood, how- ever, that like the S trap it is only self -scour ing when properly set, namely : with a free outlet from the bowl somewhat larger than the inlet-arm of the trap at its largest part, or at its point of junction with the fixture. If set under a fixture giving a clear water-way of only an inch or of half an inch, this trap will not scour itself, nor will the waste- pipes with which the trap is connected. A good- sized washbasin holds, up to its overflow, about two gallons of water. This will escape through an average length of one and one-half inch waste- pipe, running full-bore and having a good fall, in about three seconds. Hence, through such a pipe 36 THE PRINCIPLES OF the water rushes at a rate of more than half a gallon a second and fully scours the pipes. With lava- tories constructed on this principle, the argument for trap ventilation based on the supposition that it is necessary to keep the branch wastes clean no longer holds good. Fig. 16. " Sanitas " trap. 2 Jg Let us examine now the action of the air and water in our trap under the influence of a very powerful siphonic action. We will suppose the trap to be placed in position under a fixture with the water standing in its normal condition up to the level of the outflow, as shown in Fig. 16. When, * This trap is manufactured by the " Sanitas " Manufacturing Co., No. 4 Pemberton Square, Boston, Mass. HOUSE DRAINAGE. 37 through siphonic action, a partial vacuum is created in the waste-pipe below the trap, the water in the inlet-arm of the trap descends under the influence of the atmospheric pressure on its surface tending to restore the equilibrium, until it reaches the dip Fig. 17. Body of trap. of the trap. The air then being lighter than water passes into and through the body of the trap and drives a portion of the water, not already driven out, before it into the waste-pipe. The water re- maining in the body falls back and maintains the seal. Subsequent siphonic action cannot re- move this water for the following reasons : The water standing in the inlet-arm after its partial removal from the body of the trap by siphonic 88 THE PRINCIPLES OF action, as described, is again lowered by a repetition of the action to the dip. Air again rushes iuto^the body to fill the partial vacuum and passes into and through the water standing therein. This water, though increased in depth by that which enters from the inlet-pipe, is, nevertheless, shallow enough Figs. 18 and 19. Glass cap and bridge. to give room for its passage. It projects upward a certain quantity of water in its passage, with greater or less violence, according to the strength of the siphonic action produced. This water strikes the under surface of the partition in the body, and is partly reflected backward by it, and partly follows the air-current toward the opening between the end of the partition and the cleanout cap. Owing to the greater weight and momentum of the water over that of the air, the water is reflected back, HOUSE DRAINAGE. 39 while the air passes on. A second reflection takes place against the surface of the cleanout cap. More water is thrown back, and a small remaining portion only succeeds in following the air into the passage above the partition. Of this small portion part again is reflected back by the upper inner curved surface of the horizontal body, and under very strong siphonic action a few drops may reach the last reflecting surface at the end of the body opposite the cleanout cap, whence it is once more arrested, and the air alone escapes into the waste- pipe. The spray falling upon the partition and upon the various reflecting surfaces collects at the bottom of the body and increases the depth of the seal. The greater cohesive and attractive force of the particles of water over that of air aids in the sepa- ration, since it causes a quantity of the former to adhere to the reflecting surfaces while the air escapes. This arrangement of the reflecting sur- faces evidently also prevents loss of the water-seal by the momentum of the water descending from the fixture. Although the seal is not excessively deep, yet the trap, owing to the considerable horizontal extension of its passages, contains a large enough body of water to protect it from the dangers of evaporation and back pressure. The contraction of the inlet and outlet arms at their junction with the body of the trap renders it secure against self-siphonage. The form also renders loss of seal through capillary attraction impossible, as will be hereafter shown. When used where trap ventilation is prescribed by law, this trap can, of course, be ventilated like any other. The vent may be applied at any part of the outgo, either at or below the crown. But since, unlike S traps, its seal cannot be destroyed by self- 40 THE PRINCIPLES OF siphonage or momentum, the vent need not be applied at the crown. It may be applied anywhere below the crown far enough away to quite avoid the injurious effects of evaporation. Hence, ventila- tion does not produce the destruction to the seal that it does with other self -cleansing traps, and may be used with impunity. Trap ventilation is, never- theless, in this case, as in most others, absolutely useless, and its installation is a total loss to the house-owner. Having now explained the theory of the construc- tion of the " Sanitas" trap, let us make a practical trial of its operation. The first discharge of both closets, the soil-pipe extension being forty-five feet, will lower the seal considerably, say to a point below the centre of the glass, but subsequent discharges will have very little further effect upon it, and when the seal has been reduced to about an inch and a half, or in the very severest possible cases long repeated, possibly to an inch and a quarter or eighth, even the most power- ful suction that can be applied with an apparatus used in practice will have no further appreciable effect upon it, even though the siphonic action be strong enough to destroy the seal of a fullv venti- lated S trap or of a six-inch pot- trap. Ordinary siphonic action will simply lower the water in the trap enough to permit the passage of air above it, leaving a seal of two inches or more permanently in the trap. The test which we are about to apply to this trap is severe enough to siphon out com- pletely in a single second a fully ventilated S trap or a four-inch pot-trap. (Discharge Z and J.) After this very severe test we find a seal left of one and one-half inch. (Repeated.) Only an eighth of an inch has been removed by a second discharge. HOUSE DRAINAGE. 41 (Repeated ten times or more.) Five repetitions of the discharges have lowered the seal less than an eighth of an inch, leaving a full seal of one and one-fourth inch. Five further repetitions pro- duce no further visible effect on the seal. In our previous experiments we have repeated the test fifty times without apparent diminution of the seal. Snow. We will now apply a test stronger than any we have tried this evening. It is a test severe enough, as we found in making the experiments for the City Board of Health, to siphon out a pot-trap eight inches in diameter. We will retain the full length of our soil-pipe and stop up the opening above the roof with oakum. Then, by discharging both closets together, we shall produce a suction as great as any which could pos- sibly be produced in practice as when the top of the soil-pipe is closed up by ice or snow. The first discharge has left one and one-half inch of seal. At the end of five discharges there is still one and one-eighth inch of seal left and five more produce no further apparent diminution of it. Thus we see the seal of this trap cannot be broken by any siphonic action we can try with or- dinary apparatus used in plumbing. Self -cleans ing Property of the " Sanitas" Trap. It remains now to determine if the " Sanitas " trap is actually as self -scour ing as it is claimed to be. We have had a " Sanitas " basin and trap set above a waste-pipe of glass in order to examine the scouring action of the water discharged from a basin with a properly proportioned outlet both on the trap and the pipe below it. (Discharge of the Basin.) We see that the water rushes through the trap and waste-pipe at a very rapid rate. The basin, when filled to the brim, holds 42 IHE PRINCIPLES OF about two and one-half gallons. It empties itself, when set with a waste-pipe having a good fall, in about three seconds. Hence, the water flows at the rate of nearly a gallon a second, and has an enor- mous scouring-force on all the branch piping con- nected with it. We will now throw various substances into the bowl and trap and see whether they are retained in them or not. We will first try a quantity of coarse coal ashes, and, to make the test somewhat severe, we will re- move the strainer and throw into the trap pieces of coal with the ashes nearly an inch in diameter. Filling now the basin we find a single discharge has removed all the dirt and a second discharge has left the trap and waste-pipe as bright as ever. I will now form a paste of softsoap and loani. The loam is a mixture of earth and clay. The combination of this and the soap forms in large quantity the kind of waste matter to which wash- basins are most accustomed. All this matter is instantly carried through the trap and waste-pipe, and, after a second flushing, it does not leave a stain behind. After trying a few other substances we will make a strong solution of soap and dirty water and let it dry on the pipe and then see if it will wash off after drying. Large pieces of hair-felt, strings of jute and tow, coarse gravel, pieces of stone an inch in diameter, nails, and matches are all whisked through the trap as easily as if they were nothing but house-flies. In short, every kind of substance likely to be met in usage, and a great many others, are carried through and away with speed and cer- tainty, and the self-cleansing power of the trap is demonstrated. HOUSE DRAINAGE. We have found the soap dried on the glass tube is completely removed by the strong flushing from the basin. '(This last test was tried on another occasion before the lecture.) Capillary Attraction. The seal of a trap is some- times slowly and silently drained off by bits of hair, sponge, or twine which get caught across the outlet of the trap, as shown in Fig. 20, and draws out its Fig. 20. Seal of S trap destroyed by capillary attraction. Fig. 21. Seal of pot-trap de- stroyed by capillary attraction. water by capillary attraction. Numerous experi- ments have been made of late on this insidious ene- my to the life of water- traps, and it has been found that there is a limit to the height which these sub- stances will carry the water above its normal level. We find this limit of height to be within three inches for small quantities of long and fibrous sub- stances such as might get lodged in traps. We must, therefore, form our trap in such a manner 44 THE PRINCIPLES OP that the water will have to travel along the fibrous substance more than three inches before its seal can be broken. The " Sauitas " trap has been so constructed (Figs. 22 and 23), and in no Fig. 22. " Sanitas " trap re- Fig. 23. " Sanitas." trap re- sisting capillary attraction. sisting capillary attraction. case has it been possible to destroy its seal by the capillary attraction of substances which could be lodged in it in practice. Back Pressure. Back pressure is a force now but little to be feared in plumbing. Before it became cus- tomary to ventilate our waste and soil pipes, pressure in the sewers, either from winds or tides, or the heat of steam or chemical action, sometimes produced a serious back pressure in our house-pipes. Now we no longer encounter the difficulty from these causes, since we are accustomed to have our pipes properly ventilated. It is only under certain rare conditions, such as when a trap is situated near the bottom of a tall stack of pipe and close to a sudden bend, that back pressure is produced by falling water com- pressing the air in advance of it. The bend in the soil-pipe prevents the escape of the air below as fast as it accumulates above under the falling water- plug. To resist this pressure it is only necessary to have a sufficient body of water in the trap and to set HOUSE DRAINAGE. 45 the trap at a distance below the fixture it serves sufficient for this water to form in the pipe when subjected to back pressure, a column from twelve to sixteen inches long. (Fig- 25.) The weight of such a column is ample to withstand any back press- ure ever now encountered in good plumbing. Fig. 24. S trap emptied by back pressure. The " Sanitas " trap is made to contain a body of water heavy enough to easily resist any back press- ure it can ever be called upon to bear in modern plumbing work. Evaporation. When traps are not ventilated evaporation goes on with such slowness as to be scarcely perceptible. Nevertheless, it is best to have the trap contain as large a body of water as is 46 THE PRINCIPLES OF consistent with its self-cleansiug properties. An ordinary one and a half-inch S trap holds about three eighths of a pint of water. A one and a quarter-inch S trap holds less than one fourth of a pint. The " Sanitas " holds one and a half pint, Fig. 25. Deep seal S trap resisting back pressure. or about as much as an ordinary three-inch pot-trap. This is sufficient to last, under ordinary conditions, over a year without renewal when the trap is un- ventilated. Where the trap is ventilated, however, in the manner customary under our present plumbing laws, the seal of an ordinary machine-made S trap is HOUSE DRAINAGE. 47 licked up by the air-current in a very short time, varying in my own experiments from four to eleven days. Size and Material. Traps for the smaller fix- tures should be manufactured in one size, that is, of a capacity sufficient to fill the usual one and a half or one and a quarter inch waste-pipe full- bore. In other words, the size of the traps should be governed by the size of branch waste-pipes. These pipes should never exceed one and a half inch in diameter, except for water-closets. Waste-pipes should not be less than one and a quarter inch in diameter. Hence, the capacity of the trap should not be less than this at any part, and to be self-scouring should not exceed this capacity at any part. The cleanout cups should be made of glass or of metal. Glass should be used for washbasins only, and then only when a possible fracture will not produce serious damage to frescoed ceilings below. With bathtubs, sinks, laundry -tubs, and all other fixtures metal cups should always be used, since even the best annealed glass is liable to be broken by sudden changes of temperature or by careless usage. PART II. WASHBASINS. THE character of our lavatories is a matter of very much greater importance than is usually sup- posed. We have been in the habit of selecting our washbasins and bathtubs purely from a standpoint of convenience, ap- pearance, and economy. Sanitary ? .[25. ordinary considerations have been quite ner ' B1K overlooked, in the belief that they have little or nothing to do with the form of these particular fixtures, so long as their traps and waste- pipes were properly made. This is a very serious error, and particularly so in relation to washbasins, in the choice of which sanitary considerations should outweigh all others. We say this advisedly, and for the following reasons : As usually constructed, the outlet is altogether too small in proportion to the size of the trap and waste-pipe. The result is imperfect flush- ing of these pipes, gradual accumulation of filth in them, and the various serious evils to which such ac- cumulations give rise. Fig. 25 shows the actual dimen- sions of the ordinary basin-strainer. It will be found, by accurately measuring these figures, that the amount of water-way is just equivalent to that of a three-fourths inch pipe. A very short usage soon reduces this meagre opening, through the 50 THE PRINCIPLES OF collection of sediment and lint, to a still smaller stream. Accordingly we find that by far the greater part of the ordinary basins now in use discharge a stream not over half an inch in diameter. The waste-pipes are usually an inch and a quarter or an inch and a half in diameter, a capacity which is given for the purpose of ensuring the safe removal of the water delivered by two supply-faucets run- ning full force, under medium or high city pressure, and escaping through the outlet and overflow pas- sages combined, together with a possible simultane- ous discharge of the adjoining fixtures entering the same waste. Now a half -inch stream of waste- water, trickling through pipes capable of delivering ten times as much, fouls, but does not scour, them. I have taken out such pipes and found them more than half filled with slime and filth, and in places where the pipe ran nearly horizontal, or made sharp bends, I have found them nearly filled with the putrefying mass. No amount of ventilation can cleanse such pipes. But the sediment was soft and gelatinous, and would easily have been swept away by the powerful discharge of a basin filling the pipes full-bore. Besides the important sanitary advantage of a rapid discharge, we have others of economy and convenience. To empty an ordinary basin requires a very considerable amount of time and more patience than the majority of people possess. The result is that people fall into the habit of washing from the faucet rather than from the basin, and a great waste of water is involved. A quick waste and convenient method of operating and controlling it results in a saving of water and very great con- venience in usage. A knowledge that a sudden dis- charge of a basinful of water through the pipes acts HOUSE DRAINAGE. as an important sanitary measure, after the manner of a flushing tank, in cleansing them from end to end, leads to a legitimate use of the basin and an economy of water, a consideration which the water- companies and the public in time of drought will not be slow to appreciate. WASHBASINS HAVING CONCEALED OVERFLOW- PASSAGES. This class of fixture violates one of the first con- ditions of sanitary plumbing. A portion of the apparatus intended to carry off waste-water at the irregular and uncertain intervals of overflowing becomes fouled without the chance of cleansing through flushing action, and is placed in such a position that it cannot be seen or reached without disconnecting the whole fixture. Our first subdivision of this class is the ordinary PLUG-AND-CII AIN OUTLET-BASIN . We see here (Fig. 26) the concealed overflow-pipe constructed of lead and so placed as to be altogether inaccessible. Being above in open com- munication with the air of the room, it taints it with the de- composing soap and filth with which the sides soon become coated. The ordinary washbasin has no proper flange for con- nection with the lead overflow - pipe ; the joint has therefore to Fig. Ordinary plug-and-chaii outlet-basin. 52 THE PRINCIPLES OF be made with ordinary putty, which can never be made permanently and surely tight. The lead pipe must be connected with the main waste-pipes above the trap, and the joint here must be wiped with sol- der. Thus, to set an ordinary washbasin, the plumber has two extra joints to make, which add both to the expense of the work and to the chances of im- perfection and leakage. It is an exceedingly com- mon thing to find the overflow-pipe wrongly con- nected ; it is sometimes entered below the trap, sometimes attached directly to the trap-vent, and sometimes connected with the wastes of other fix- tures in such a way as to open, through the vent- pipes, an indirect avenue into the house for sewer- gas. It forms, in short, an unnecessary and dan- gerous complication to the plumbing, and these basins should never be used. Many house-owners stop up the holes in the earthenware leading into the overflow-pipe at con- siderable inconvenience to themselves, in the hope of avoiding the chance of the entrance of offensive or injurious gases into the house through this chan- nel. With defective traps, or with traps whose seal is liable to be quickly destroyed bv evapora- tion, siphonage, or other cause, this precaution against danger would not be useless if the overflow- pipe connections could be made certainly tight, es- pecially when the fixture is left for some time unused. As they are made, however, it is probable that no such precaution would form any reliable security. The use of the plug and chain, which charac- terizes this type of basin, is another serious defect. The chain, lying in every successive formation of dirty water, collects gradually in the recesses of its links an unknown quantity and variety of filth, HOUSE DRAINAGE. 53 which cannot be entirely removed on account of its irregular form, without the use of special acids or constant scrubbing with a brush, a process never applied to it. The length of wire used in an ordi- nary basin-chain averages six feet, and has a sur- face of about fourteen square inches, a surface which, in consideration of the peculiar adaptability of the form of the links for retaining dirt, presents a very formidable area of pollution. To those per- sons who use their reasoning powers in these mat- ters the idea of washing the face in water defiled by a chain transferred immediately from the dirty water of some unknown predecessor is with good reason exceedingly repulsive. The chain, more- over, frequently breaks, and then the hand must be plunged into dirty water to remove the plug. The position of the chain and plug at the bottom of the bowl is, moreover, peculiarly inconvenient, inasmuch as they are in the way of the hands, which should meet a smooth, unbroken surface of earthenware rather than the hard and irregular out- lines of the brasswork. If this latter consideration appear to some trivial, it does so only because habit has rendered us callous to such defects ; the defect none the less exists, and acquires importance through the frequency of its repetition and the con- stant use of the fixture in which it occurs. The fact that it is altogether unnecessary is a sufficient reason for its abolition. CONCEALED WASTE-VALVE BASIN. Fig. 27 represents a basin fitted with the so-called " Boston Waste," which is very popular. There is probably no form of basin-fitting more faulty in principle than this. It contains two independent, inaccessible, and invisible foul-water passages, one forming the overflow-passage, and the other the 54 THE PRINCIPLES OF outlet passageway^ 1 between the strainer and the waste-cock. This latter passage forms an elongated cesspool for the defilement of the clean water en- tering the basin. After using the fixture, the waste- water escaping through this channel deposits part of its dirt, particularly floating matter and soap- suds, all along its sides, and leaves it there to be taken up and applied in a diluted solution to the hands and face of the next comer. Six wiped Fig. 27. Concealed waste-valve basin. solder joints, one putty joint, and five threaded joints, making twelve in all, are required to adjust the waste-pipes of this apparatus and its trap be- low the basin-slab ! No wonder the plumber is con- stantly in requisition to keep in repair such a com- plicated machine so long as the owner allows it to remain in his house. Not the least of its defects is that the passageway for the waste-water through the ground-cock is usually so small (scarcely a quar- ter of an inch wide) that a small deposit of sedi- ment will entirely prevent the outflow of the water. The ' ' Boston Waste " cannot be too strongly con- demned. The great extent of its use in spite of its HOUSE DRAINAGE. 55 high cost shows how little knowledge the public have in these matters, and how important it is that their attention should be called to them. THE STAND-PIPE OVERFLOW-BASIN. It is not sufficient that every part of our appara- tus should be visible and accessible from without, and devoid of all fouling chambers and corners, but it is above all necessary that, combined with the utmost convenience and simplicity of action and economy of construction, it should be so formed as to ensure the complete automatic scouring of its waste-pipes and trap without detriment to the water- seal of the latter. A suitable enlargement of the basin-outlet is all that is necessary to produce the requisite scouring action ; but the force of the out-flowing water-col- umn is so great when the pipes are charged full- bore that it will siphon out and completely destroy the water-seal of an ordinary S trap unless it be fully ventilated at or very near the crown, and it will dangerously lower it even then. This is the action which we have called u self-siphonage." No injurious effect is produced on the seal of an anti- siphon trap by self-siphonage, but the water is some- what reduced in the trap below its normal level. It is therefore desirable, and when S traps must be used extremely important, that such a basin should be so constructed as to enable it automatically to restore the water, and in the following description it will be seen that this has been accomplished. The next important point is to obtain the utmost simplicity of form and to provide for an overflow- passage which shall be both visible and accessible. We have established as the second datum of our problem that the basin be fixed and single, and have an independent, visible, and accessible overflow-pipe. 56 THE PRINCIPLES OF It is important both for convenience and economy that the opening in the marble slab covering our basin should be circular or elliptic. These open- ings are cut by machinery, and any form other than these requires manual work, and at once increases the cost of manufacture. Moreover, this form of opening occupies the least space on the slab and presents the most agreeable effect. A third datum in our problem is, therefore, that the usual round or elliptical form of the opening in the basin-slab be retained. Finally, as the overflow-outlet must be near the top of the basin, some form of passageway which shall extend from the top to the bottom must be provided, and since this cannot be on the outside of a fixed basin without being concealed by the slab, it must be on the inside. Hence, as a fourth datum, our overflow must have the general form of a stand-pipe, and to be completely out of the way of the user it must set in a recess under the slab at the back of the basin, which must be perpendicular at this point to receive it. Fig. 28 represents the elevation and plan of our stand-pipe overflow-basin, designed in accordance with these data. The opening in the marble slab is circular or elliptical. A smaller circle represents the stand-pipe in a small recess at the rear of the bowl under the slab. The recess is large enough to allow easy cleansing without moving the stand-pipe ; and yet not so large as to injure the appearance of the bowl. The bottom of the basin pitches slightly from the centre toward the outlet ; enough to thor- oughly drain off the water at each discharge, and yet not so much but that the last part is retarded until the siphon formed by the main body is broken by the air. The result is a restoration of the water- seal of the trap, in case the trap be of a kind which HOUSE DRAINAGE. 57 would in basins of ordinary form be destroyed by self-siphonage. The diameter of the brass outlet at the bottom of the basin measures 2 inches in the clear, in order to allow for the obstruction caused by the strainer. The stand-pipe is raised and low- ered by a simple weighted cam, arranged as shown Fig. 28. The " Sanitas " stand-pipe overflow-basin. in Fig. 29. The cam is pivoted close to the rod which raises the stand-pipe, and its bearing sur- face has the form of a parabola. Its proportions and arrangement are such as to enable it to raise the stand-pipe without perceptible friction. The weight of the handle, about 6 ounces, under a leverage of only an inch and a half (the length of the horizontal bar) , is sufficient to overbalance the stand-pipe and hold it raised. When the weight is lifted, the stand-pipe is lowered and the outlet closed. In this position the weight is directly over 58 THE PRINCIPLES OF the pivot, so that the plug and stand-pipe overbal- ance the weight and remain closed. Thus a single movement of the hand will open or close the out- let and cause it to re- main in the position in which it is left. The stand - pipe rod passes through, and is guided by. a short tube threaded on the lower end and provided with a nut by which the brass- work is secured to the marble slab. This rend- ers the brass work adjust- able, that is, permits it to be used with slabs of any thickness, the slabs be- ing perforated by a sin- gle hole, as is usual for the chainpost of ordinary basins. The centre of the hole in the slab comes over that of the basin-outlet. In order to ensure the plug on the stand-pipe falling always into its socket, the strainer is attached permanently to the stand-pipe plug, as shown in the drawing. It thus serves as a guide for it, and allows its being lifted out at pleasure for cleansing purposes. It is thus possible to remove the strainer, and reach the inner surface of the waste-pipe as far down as to the trap itself. Thus we have an apparatus, every part of which, both inside and outside, is visible and accessible without unscrewing or taking down any part of the set fixture. Fig. 29. Brasswork of the " Sanltas " basin. HOUSE DRAINAGE. .09 The stand-pipe with its plug and strainer may be lifted out by simply unhooking it from the stand- pipe rod. P'igs. 30 and 31 show the appearance of the apparatus in perspective. By these it will be ob- served that the recess in the basin and the stand-pipe are covered by the marble, and do not interfere with the general form seen from above. The stand-pipe measures about 3^ inches in height and 1^ inches in diameter. Hence its exte- 3wrior surface measures 13 square inches. Its superficial area is, therefore, not so great as that of the ordi- Fig. 31. Basin-set " open." nary basin-chain. But while the chain cannot be 60 THE PRINCIPLES OF cleansed on account of its intricate form, the smooth surface of the stand-pipe, on the contrary, can be surrounded and polished by a single move- ment of a cloth or sponge. Ample room for the scrubbing-cloth is provided between the stand-pipe and the walls of its niche, so that both may be cleansed without lifting out the former. WASHBASINS, BATHTUBS, SINKS, ETC. The criticisms we have made on washbasins apply equally to bathtubs. No better flushing apparatus could be devised for the branch waste- rig. 32. " Sanitas " ball). pipes than a properly constructed bathtub with a large outlet. As we have found the case with washbasins, so with bathtubs : the best form of dis- charge and overflow is the open stand-pipe ; and the most convenient method of operating it is by HOUSE DRAINAGE. 61 the weighted cam already described. The cut, Fig. 32, shows the arrangement in perspective. Part of the floor in front of the bathtub is designed with hinges to lift upward and give access to the trap. Fig. 33. Transverse section of tub. The stand-pipe sets in a recess at the back of the tub. The whole length of the tub is thus rendered available, and no part of the outlet mech- anism stands in the way of the feet of the bather. In ordinary bathtubs the stand-pipe is set six or eight inches from the back into the tub. It has to be lifted out by hand when the tub is emptied, and a place found for it. This constant shifting of the stand-pipe is not only troublesome but liable to cause injury to the tub itself. In careless handling the heavy pipe is often dropped into the tub, where- by the thin copper is permanently dented and marred. By the "Sanitas" fitting all handling is avoided. The supply-cocks may, of course, enter the upright end of the tub, or be placed above it out of the way as shown in the drawing. THE PRINCIPLES OF Figs. 33 and 34 show the tub in section, and Fig. 3o in plan. The finished wooden top or slab of the tub has the usual form, as shown by the dotted lines in the latter figure. The tub itself has also the usual shape, except as to the semicircular recess for the reception of the stand-pipe. In all plumbing work it is of the first importance that every inch should be accessible, and, where possible, every inch visible without unscrewing or undoing any part of the work. Therefore when- ever a bathtub is cased in, all parts of the piping should be accessible through hinged panels. Fig. 34. Longitudinal section of tub. It will always be found best to follow the custom of many plumbers and builders of raising the bath- tub four inches or more above the floor on pieces of studding, as shown. This renders access to the trap much easier. A bathtub raised six or eight inches would be still better. It should stand 2 feet, 2 inches, or 2 feet, 4 inches, above the floor up to the top of the finished woodwork. This would require HOUSE DRAINAGE. 63 the use of studs 6 inches or 8 inches thick under the rough woodwork of the tub. The writer has, for many years, used a tub set even higher than this, and finds it equally if not more convenient in usage. Fig. 35. Plan of the " Sanitas " tub. A small piece of sheet lead, about eight inches square with its edges turned up an inch all round and soldered to form a water-tight tray, may be set under the trap, as shown, to catch the water when the cup is removed for examination of the interior. Or a saucer or dustpan may be used for the purpose. But the best plan is to place the bathtub trap in full sight below the plaster of the ceiling of the bathroom or china-closet below. It is not desirable, as some writers on sanitary plumbing urge, that the bathtub should stand open on legs without wooden casing, because the bath- tub must set so near the floor that the space under it if left open would be hardly more than a crack. Dust would collect under it, and though, with prouer care, it would be quite possible to remove it with the proper form of swab or broom, yet we know that places so difficult of access would be neglected bv servants. The very reason which would induce us to leave, in some cases, the washstand without casing would lead us to case up the bathtub, 64 THE PRINCIPLES OF that is, to avoid dirt and dust corners. In either case, whether the tub be sheathed in or not, its trap, being wholly or partly under the floor-level, would have to be approached through movable panels, so that no advantage on this score would be gained by leaving the bathtub open. Fig. 36. Iron bathtub with "Sanltas " waste. In Europe where copper baths are made of metal heavy enough to stand alone without a wooden frame, the custom prevails of allowing them to stand free on the floor. But in this country where tinned and planished copper sheets weighing only from ten to twenty-four ounces a square foot are used, a rough frame is required to support it. This requires a casing of finished wood and the casing should extend to the floor. Porcelain baths of English manufacture are also used in this country to a certain extent and gener- ally stand open. Their high cost and great weight prevent their general adoption. They have, more- over, the disadvantage of being cold to the touch and of lowering the temperature of hot water when HOUSE DRAINAGE. 65 it is used in them. Porcelain-lined iron tubs are also manufactured here and are made to stand open without casing. The liability to " scaling" of the enamel is their only serious objection. Slate and marble tubs are also made, the slabs being con- nected with cement. Fig. 36 represents a porcelain-lined iron bath- tub standing open. SHOWER-BATHS. Fig. 37 represents a shower-bath. It stands free in the corner of the bathing-room, which has a dished or sunken floor to receive it. Jets are arranged on all sides as well as above and below. The shower-bath is an agreeable luxury for sum- mer use ; but for winter houses it is not so much to be recommended. The shock produced by sprays of cold water upon the body standing inactive is dangerous ; and the use of warm water in the shower-bath in winter, without intelligent precau- tions, is also objectionable. For these reasons shower-baths are now seldom used in city houses in the North. The usual form of shower-bath consists in a simple rose-nozzle arranged overhead to throw per- pendicular jets downwards. There are, however, a very great variety of forms and arrangements of shower-bath in use. Sometimes, in combination with the simple rose-nozzle above, a lower nozzle is provided with a rubber hose in such a manner that the lower jets may be applied in any desired direc- tion. Sometimes the end of the bath-tub is raised high enough to enclose the shower-bath as in a niche. Lateral as well as perpendicular jets may be used within this niche, in which the bather may sit or stand. Sometimes the rose-nozzle is fur- 66 THE PRINCIPLES OF nished with special " needle" outlets throwing strong sharp jets in addition to the drops from the sprinkler. To give varying pressure the cistern is Fig. 37. Needle shower-bath. sometimes hung in such a manner that it may lie raised or lowered at will bv means of pulleys. Fig. 38 illustrates the manner in which the overflow-passage of a bathtub may become foul. Colonel Waring, in his article on " The Principles and Practice of House Drainage," in the November and December numbers of The Century Magazine for HOUSE DRAINAGE. 67 1884, speaking of concealed overflow -passages, says : ' ' They are practically never reached by a strong flushing stream, and their walls accumulate tilth and slime to a degree that would hardly be be- lieved ; . . . they are more often than any other part of the plumbing work, except the urinal, the source of the offensive drain-smell so often observed on first coming into a house from the fresh air. ... It will, perhaps, be instruc- tive to illustrate by a diagram the reason why the usual hidden overflow is so objectionable. . . . If we suppose the tub to be filled to the level of the overflow and Fig 38 Colu . eal( , (1 its waste-plug to be removed, the overflow, water will immediately rise in the overflow-pipe to very nearly its height in the tub. It is, of course, impregnated with the impurities of the water in the bath. Furthermore, the lighter particles of organic matter flowing through the waste will, some of them, rise by their levity into the overflow- pipe. The water rushes up into this pipe with much force, but it descends only very slowly as the level in the bath descends, so that at each operation there is a tendency to deposit adhesive matters to the walls of the pipe. What is so deposited decomposes, and escapes little by little in a gaseous form through the perforated screen into the air of the room. The amount of these decom- posing matters is somewhat increased, though prob- ably not very much, by floating particles passing through the screen when the overflowing is perform- ing its legitimate function." "This is the simplest statement of the proposition, and this is, perhaps, the least objectionable form of 68 THE PRINCIPLES OK hidden overflow. Where the waste-pipe is closed at the bottom of the overflow by a plug or valve attached to a spindle rising through the overflow- pipe, a very favorite device with some plumbers, the difficulty is in every way aggravated, and the amount of fouled surface is much increased. The inherent defect here illustrated attaches to every overflow of this general character connected with any part of the plumbing work. In the case of a bathtub it may very easily be avoided, as shown in the next diagram." Colonel Waring then illustrates and explains the ordinary stand-pipe arrangement, and says : " Unfortunately such a substitute for the ordinary overflow is not applicable to washbowls as now made. 1 It may be made available for pantrv- sinks if the pipe can be so placed in a corner as not to interfere with the proper use of the vessel. If its universal adoption for bathtubs could be secured, a very widespread source of mild nuisance would be done away with. Fortunately, it is far cheaper than any arrangement for which it is a substitute. It is one of its incidental uses that it enables us to get rid of the dirty chain attached to the ordinary bath- plug." The size of the trap for bathtubs should never exceed that of the outlet and waste pipe. A 1 J inch trap is large enough for any bathtub or washbasin. The proper size for traps and waste-pipes of differ- ent fixtures is a matter which is very little under- stood. This ignorance has its origin in the faulty construction of the outlets, which are always entirely too small. A very simple and self-evident rule is that no trap except a water-closet trap should be larger or smaller than the waste-pipe at its most con- 1 This was written before the " Sanitas " basin was put on the market. HOUSE DRAINAGE. 69 tracted point and no fixture outlet should be smaller (in its clear water-way) at any part than the trap. In no other way can the pipes and traps be properly flushed or the fixtures emptied with the desired rapidity. Hence, since we find the best size for waste-pipes for all lavatories is 1 inch or 1^ inch in inside diameter, except in certain cases hereafter to be referred to, no trap should be greater or smaller than this. Moreover it is very important to bear in mind that the measure of a trap is the diameter of its inlet and outlet pipes at their smallest part. Traps are arbitrarily measured without regard to their real discharging capacity, but generally from the diameter of the inlet-pipe at its lower end, which is in most cases much larger than the smallest part of the trap. Many traps are sold as IJ-inch traps which are contracted at some point to seven eighths of an inch or even five eighths of an inch. The size of a trap, then, is evidently the size of its smallest part, since this part governs its capacity for discharge. Guided by these principles we have the following rule : The discharging capacity of the outlet of every fixture should be great enough to fill its waste-pipe "full-bore," and the size or capacity of the trap should equal that of the outlet. PANTRY-SINKS . Pantry-sinks, like bathtubs, may be classified in the same manner we have done for washbasins ; and here again the open stand-pipe overflow is by far the best and most convenient form. These sinks should be constructed of tinned and planished cop- per weighing from sixteen to twenty ounces to the square foot. Iron and earthenware sinks are made ; but they are objectionable as exposing the dishes 70 THE PRINCIPLES OP and glassware to greater danger of breakage. For this reason, also, cherry slabs are preferable to marble. Fig. 39 represents an oblong pantry-sink with a stand-pipe overflow set under the slab as has been recommended for bathtubs and washbasins. The whole is set open to allow of easy inspection of all parts. Fig. 39. " Sanitas " pantry-sink. The waste-pipe should never exceed one and one- quarter or one and one-half inch in diameter ; and the outlet should have a capacity large enough to fill this pipe full-bore. The sink should be trapped with an antisiphon trap of the size of the waste- pipe. There are plumbers who still ignorantly insist that a pot-trap is suitable for sinks on account of the formation of grease. Let us examine the man- ner in which a pot-trap deals with this material. Does it collect the grease and preserve it in its body until such a time as it may become convenient to remove it, or does it in some wav alter its chemical DRAINAGE. 71 constitution so as to deprive it of its power to clog the drains? It is evident, upon the slightest reflec- tion, that it can do neither. It isnotlarge enough to materially cool the grease nor to retain any consider- able amount in its receiver. It could not intercept as much grease, even if it were completely stuffed up with it, as would pass through the sink of a large establishment, such as a hotel, or clubhouse, in a day. or of a small house in a fortnight. The outlet-arm leaves the body of the trap very near its top, as near as the solder jointing will allow. Hence, since grease is lighter than water, it will rise to the top of the trap and the first that collects there in any large quantity will necessarily obstruct the passage of the water. The trap can only retain in its body the small quantity of grease which would fill the corners remote from the outlet -arm, as shown in Fig. 40. The heavier matters carried into the trap along with the grease fall, on the other hand, to the bottom. These matters consist of bits of meat, bone, and vegetable. They form a foul sediment on the bottom of the trap and rapidly putrefy. The force of the water is generally suffi- cient to keep a passageway open for itself for a considerable length of time. Hence, after the corners have become filled with the black and rot- ting mass, the grease will pass through this trap exactly as it would through an ordinary S trap. Having no proper tools to unscrew the tightly sticking cleanout cap, the owner, after vainly ham- Fig. 40. Pot-trap fouled. 72 THE PRINCIPLES OF mering at, and disfiguring, the brasswork with such unsuitable instrument as he can lay hands on, is obliged to send for the plumber. In short, it stands to reason, and is borne out in fact, that an ordinary pot-trap has really no merit whatever as a sink or grease-trap. It is neither large enough to cool and retain all the grease nor small enough to let it all pass. To cool the grease enough to harden it before it passes into the waste- pipes beyond the trap requires a large cesspool or regular grease-trap. In many cases it is better to dis- pose of the grease by sudden and powerful flushing. The best apparatus I know of for this purpose is the flush-pot, which is nothing more than a sinkage with strainer in the bottom of the sink, so arranged as to hold the waste-water until it may be be dis- charged, like a flush-tank, to scour the pipes and hurry away the grease in an irresistible torrent to the drains. 2 Many plumbers are very fond of the pot-trap, just as in old times they were fond of its near kin- dred, the foul D trap, and, until lately, of the 8 trap, because they were able to make them by hand in spare moments on a "rainy day." But, in the long run, what is best for the public is best for the plumber. Every additional complication of the plumbing, and everything which detracts from its convenience, safety, and reliability, diminishes the amount of plumbing the public will allow in their houses, casts discredit on the art, and distrust on the plumber. Pot-traps, like S traps, are, however, now also made by machinery. Moreover, new conditions have originated new occupation for the plumbers' 8 The "Dececo" flush-pot is manufactured by the "Drainage Construction Co.," Newport, R. I. An excellent form of flush-pot is described by Wm. Paul Gerhard, C.E., in his " Domestic Sanitary Appliances." HOUSE DRAINAGE. 73 spare moments on " rainy days." The use of the flanged iron soil-pipe furnishes him with this de- sideratum. The lead rings for the joints are cast in the plumber's shop as occasion demands in small, simple moulds furnished by the manufacturers for the purpose. The plumber casts in leisure moments a stock of these rings large enough to last him through the busy seasons, and finds no reason to complain of the machine-making of the pot- trap. Ultimately, it is probable that the S trap, except Fig. 41. Laundry-tubs. for water-closets, and the pot- trap, machine as well as hand made, will disappear from good plumbing work altogether. LAUNDRY-TUBS. Laundry -tubs are made of wood, soapstone, gal- vanized iron, enameled iron, and porcelain. Wooden tubs should only be used in places where their use is constant ; otherwise they shrink in the intervals of disuse and become dirty and leaky. 74 THE PRINCIPLES OF HOIM: DKAIXACK. Soapstone sinks are most widely used on account of economy and their general serviceableness. The handsomest and best, as well as the most expensive, are the heavy porcelain travs shown in Fig. 41. A single one and a half -inch antisiphon trap is sufficient for an entire set of laundry trays, but no pot- trap should ever be used. The discharging capacity or size of the bore of a trap should always be very nearlv as great as that of the waste-pipes to which it is connected, in order that its water discharge may thoroughly scour the pipes. The fixture-outlet should also always be large enough to fill the waste-pipes and trap full- bore, in order to allow of their proper flushing every time the fixture is used, and also to effect a rapid emptying of the fixture. When a washbasin is constructed with a very contracted outlet, the discharge will be very slow and the pipes will accumulate sediment. If a trap having a very small outlet, smaller even than that of the basin, be used with such a basin, the discharge will sometimes be more rapid than when a properly constructed trap is used, because the waste-pipe between the basin-outlet and the contracted trap will be filled full-bore and create a strong suction which will assist in emptying the basin. When the plumber is obliged to set a " Sanitas " trap under one of these ill-constructed basins with contracted outlet, and finds the discharge sluggish, the rapidity may be increased by contracting the waste just at its point of junction with the body of the trap, until he trap has a discharging capacity less than that of the basin-outlet. This practice is, however, not to be recommended. It is better to have a full-sized trap, even if the basin discharges slowlv. But the only proper course is to obtain basins which have properly constructed outlets. WATER-CLOSETS. We have used plunger-closets in our tests for siphonage because these closets produce the se- Fig. 42. verest siphoning action. It is probable that some forms of valve-closets would come very near them, however, in this effect. But valve and plunger 7fi THE PRINCIPLES OF closets, like pan-closets and mechanical seal traps, are now condemned by sanitarians as unsanitary and far inferior to simple hoppers. Fig. 43. Valve-closet. Fig. 42 represents in section the ordinary pan- closet. Enough has been said by all who have any knowledge of, or interest in, sanitary plumbing about the evils of this form of fixture : its large and foul receiver, which never becomes cleansed except when the closet is taken^apart and subjected to ere- HOUSE DRAINAGE. 77 mation ; its complicated form and noisy action ; its flimsy construction and its numerous vent-holes for the admission into the house of the unwholesome odors generated in the receiver. All are becoming aware of the dangers arising from these defects, and it is unnecessary for me to dwell upon them. The one single reason why the pan-closet, with its com- plicated machinery and the fifty-one distinct pieces Fig. 44. Plunger-closet. required to construct it, is sold so much cheaper than other kinds is because these materials are of the thinnest and^ flimsiest character, and no attempt is made to prevent the diffusion of its foul gases through the pan-journal bearings and the opening made by the pan in usage. Figs. 43 and 44 represent respectively types of the valve and plunger closets. These closets are equally complicated in construction with the pan- closet and are equally objectionable in the theory of their construction. Practically they are better made, because most of those in use are patented articles. Being of more solid and honest construc- tion they command a higher price ; but they are 78 THE PRINCIPLES OP very liable to leak and get out of order, and are no longer recommended by unbiased judges. A few years ago, before systematic ventilation of the sewers and soil-pipes became universal, a tight- fitting valve or plunger might have served a good purpose in resisting back-pressure, and, as these closets were at first built without overflows, the valve or plunger performed in a measure an actual service in reducing the chances of sewer-gas leak- age. Now, however, the circumstances are altered. It is found that an overflow is necessary in these closets and this overflow-passage is rarely provided with a mechanical closure. Hence any gases which could pass an ordinary water-seal could pass through these closets by way of the overflow-passage quite regard- less of and quite as easily as if the valve or plunger in the trap never existed. Moreover, the ventilation of the sewer and soil pipes renders back-pressure under ordinary conditions impossible, so that the onlv useful office which the valve or plunger could perform in relation to sewer-gas is no longer called for. The valve and plunger evidently cannot prevent the loss of water-seal from siphouage, momentum, evaporation, or suction, even where the overflow- passage is closed by a ball, as is the case with some of the Jennings valve and plunger closets ; for siphonage, momentum, and suction act in the direction in which the overflow-ball or valve is opened, and evaporation is chiefly due to trap ven- tilation. Moreover, the tightness of a valve or plunger against its seat can never be implicitly re- lied on. They are always liable to leak, and could never be fitted with such microscopic accuracy as to prevent the passage of any microorganisms the HOUSE DRAINAGE. 79 bacteria, or disease-germs, or their spores which might be iu the water, through the minute openings which exist between the particles forming the valve and its seat. The only object of the valve or plunger, there- fore, is to retain a certain quantity of water in the bowl so long as they remain in working order. But it is found that this result can be accomplished equally well and much more reliably by simpler means. The receiver or container of these closets is open to the same objections as that of the pan, differing only in degree, and the overflow-passage, not required in the latter, forms a second filth-collector, and increases the cost and complexity of the closet. HOPPER-CLOSETS. For the before - men- Itioned reasons sanitarians are united in condemning all mechanical seal-closets and in recommending the improved hopper - closet. The old style of hopper, commonlv designated the Fig.45. The 8hort hopper, i. l ong " and short" hoppers, are objectionable as providing no sufficiently large body of standing water for the reception of the soil. The sides of the bowl in these kinds be- come rapidly fouled, and this form of hopper is never to be recommended except where the circum- stances require the use of pails for flushing. They are not fit for the better class of houses because the trouble necessary to keep them clean will not be endured ; nor for the poorer class, because the trouble will not be taken and the closet soon be- THE PRINCIPLES OF comes a nuisance in the house. Or if, by exception, cleanliness in this direction be insisted upon, the extra labor and consumption of water soon offsets the saving in first cost. IMPROVED HOPPER-CLOSETS. There are several forms of improved hopper water-closets, among which the best are the follow- ing : a^oi Fig. 46. Ordinary " wash-out " closet. Fig. 46 represents an ordinary wash-out closet. It contains the large surface of standing water for the reception and deodonzation of the soil. The flushing-stream sweeps across the bottom of the bowl with great force and drives the wastes before it into the trap. Whether or not the trap itself be emptied depends upon the length of time the flush- ing is continued after the bowl is cleared. HOUSE DRAINAGE. SI The objections to this form of hopper are : (a) the presence of the extended pipe surface between the bowl and the trap, and the inaccessibility and invisibility of the latter ; (6) its extravagant con- sumption of water, the waste matters often whirling about some time in the bowl before they are driven out ; (c) its excessive noisiness in action ; and finally (d), the spattering occasioned by the violence of the flushing. i ; } Fig. 47 repre- sents a closet con- structed with a double trap, one below the other. This water-closet, which is called the "Tidal -Wave," works on the principle of the siphon. The air between the traps is exhausted by the action of the valve and cistern. Fig. 47. Uoyle's " tidal-wave " water-closet. -phis unites the two bodies of water in the traps and forms the siphon which empties the bowl. This apparatus has lately been considerably improved by its manufacturers, and as now made forms a very excellent closet. Fig. 48 represents the " Dececo " closet. This is a simple and effective apparatus, and works on the principle of the Field's flush-tank. A weir-chamber is used below the trap to assist in charging the siphon. The weir-chamber is just below the floor. In order to charge the siphon the water is let into the basin through the supply-pipe and the flushing-rim until it overflows the outlet of 82 THE PRINCIPLES OP the trap and falls into the weir-chamber below. This closes the inlet of the weir-chamber before it can escape through the outlet and prevents the air from entering the siphon. The air already there is carried out by the current of water and the siphon is formed. As soon as the water in the bowl descends to the bottom of the dip of the trap air follows it and breaks the siphon. Fig. 48. " Dececo " water-closet. The bowl is then refilled by the afterwash. This closet is an ingenious one ; it is simple and durable, and the later and better forms seem to produce in- variably the siphonic action in the manner described, giving the requisite flushing without spattering or waste of water. It should not be used as an ash- barrel or receptacle for all kinds of rubbish. When HOUSE DRAINAGE. properly used it is a closet which never needs repair. Compared with the " wash-out" closet of Fig. 40, these points of advantage arc to be noted here : (1) The depth of water in the bowl is much greater where depth is needed to receive and deod- orize the soil. (2) The trap is in sight and the walls of the outlet are under water and are odorless instead of the reverse, as in the " wash-out" closet, (o) The water-seal in the trap is twice as deep and therefore better able to resist the influence of siphonage, etc. THE " SANITAS" WATER-CLOSET. Figs. 49, 50, and 51 represent in section and plan the " Sanitas " water-closet. The form is absolutely Rear: Fronf Fig. 49. Longitudinal section. simple. The bowl and trap are one and the same thing, inasmuch as each forms the other. The Hushing is accomplished without machinery of any kind in the closet, but by the pressure of the water only, and the quantity of water required is reduced to a minimum. The supply-pipe (see Fig. 50) enters the bowl below the normal level of the standing water therein, and stands permanently full of water THE PRINCIPLES OP Fig. 51. Plan. HOUSE DRAINAGE. 85 up to the cistern- valve. This water is held in the supply-pipe by atmospheric pressure, the pipe being closed at the 'top by the cistern-valve and at the bottom by the water in the closet-bowl. The lower end of the supply-pipe is perforated at two places . 52. Front view. independent of each other : first, at a point inter- mediate between the overflow of the trap and its dip ; and, second, at the bottom of the trap. The first supplies water to the flushing-rim, and the second furnishes a jet which lifts part of the water out of the trap and bowl by its propelling power. Since both jets enter below the level of a large body of standing water in the bowl, they act noise- lessly, and, as the supply-pipe stands always full, they act instantly, and the flushing of the closet is 86 THE PRINCIPLES OP very rapid. The operation is as follows : Upon opening the cistern-valve the water in the supply- pipe is instantly set in motion by the pressure of the atmosphere on the surface of the cistern and escapes through the two orifices in powerful jets. The lower jet removes part of the water from the trap and causes the water and waste matters in the water-closet to sink into the neck of the bowl. Meanwhile the upper jet fills the passage leading to the flushing-rim, and, descending into the neck of the bowl, falls upon and drives out the waste matters collected in the neck without noise or waste of water. The action is almost instantaneous. The cistern-valve being again closed, movement in the supply-pipe immediately ceases, and the water in the flushing-rim and passages leading thereto falls back into the closet and restores the normal level of the standing water in the bowl and trap. The form of the closet-bowl is such that the sur- face of the standing water therein is very large. It has the shape best calculated to receive and deodorize the waste matters falling into it. The water is deepest at the back of the closet, and very deep at the point where the wastes strike. All parts of the trap and bowl are easily accessible from the bowl itself, and there is no superfluous space and no surface which is not thoroughly scoured by the flushing-streams in the normal usage of the closet. There is no invisible trap below the bowl, and when the closet appears to be flushed clean it is so. This closet can be easily flushed in one second by less than a gallon and a half of water. There are several advantages in having the supply -pipe of a closet enter below the level of the water in the bowl and closed above with a valve without air-pipe so HOUSE DRAINAGE. 87 that it shall remain always full of water. In the first place its action is instantaneous and noiseless. The water does not have to fall from the cistern to the closet before it begins to work. In the second place the friction of air in the pipe is avoided and the water exerts at once its full power in discharging the waste matters. Hence a very considerable economy of water is the result. As already stated, the upper orifice is placed below the level of the standing water in the closet-bowl, but above the dip of the trap. This position of the upper jet gives us another very important advantage. Should the water in the closet be lowered by evaporation or siphonage below the upper orifice, air will at once enter the supply-pipe through this orifice and water will then descend from the pipe into the closet through the lower orifice, until the upper orifice is again covered, and the seal of the trap is thus auto- matically maintained by the water in the supply- pipe. This pipe may be made capacious enough to restore the seal as often as it is likely ever to re- quire it. A pipe 1 or If inches in diameter and six feet long will contain water enough to secure the seal against destruction by evaporation for a great many months, even in the dryest and hottest weather. Hence the closet may be left to itself in city houses for the entire summer's vacation, without fear on this score, and the danger of a loss of seal through siphonage is also reduced to a minimum. The seal of this closet is over three inches deep. Such a seal is difficult to break by siphonage, even without the use of our automatic supply-pipe, which I have called the " Sanitas " water-closet supply-pipe. It will be observed that the closet is provided with a ventilation opening near the crown of the trap. This ventilation will seldom if ever be required to 88 THE PRINCIPLES OF prevent external siphoning action. But it is useful to break the siphon which would be formed in the closet-trap itself during the flushing, and thereby prevent the noise which the formation of such a siphon would occasion in use. Hence, where it is desired to have the closet act noiselessly, a short, two-inch vent-pipe should be used to connect the crown of the trap with the soil- pipe immediately on a line with it or with any other convenient point of the soil-pipe near by. Moreover, the law at present in some cities requires every trap to be vented regardless of consequences. So that such a vent-opening may at times be needed to con- form to the requirements of this arbitrary and ill- considered provision. It will also be observed that the closet is provided with a cistern-overflow con- nection which may serve also when desired for a bowl ventilation pipe connection. An important advantage in having the trap and bowl of a water-closet combined in this simple form is that they may be easily emptied in winter to pre- vent freezing. This is particularly desirable in the case of summer residences which are closed up in winter. The water may be easily sponged or pumped out of this closet without taking it apart, whereas closets having inaccessible traps under the bowl or floor cannot be emptied or cleansed without taking the apparatus to pieces, and in the case of many forms of wash-out closets where the trap under the bowl is in a single piece of earthenware with the bowl, the emptying or cleansing of the trap is either very difficult or altogether impossible. The upper flushing is accomplished without spat- tering because the pressure of the upper jet is relieved at the upper orifice, and the water quietly overflows the rim of the bowl. Fig. 52 gives a front view of the u Sanitas " closet. HOUSE DRAINAGE. 89 LATRINES AND TROUGH WATER-CLOSETS are designed for use in public places where an attendant can be employed to take constant charge of them, and where water is so abundant that its extravagant consumption is no disadvantage. Trough water-closets consist of a long reservoir or trough, inclined toward one end, where a discharge- plug is placed, and having a single or double row of water-closet seats placed over it, so that all the Fig. 53. Latrines. closets are flushed together, or, in other words, so that the flushing of one necessitates the flushing of all the rest in the series connected with it. They are constructed in different manners, either of brickwork having vertical sides and rounded bottom, or of iron, usually enameled. Latrines (Fig. 53) are practically trough water- closets having the trough diminished in size, and a bowl or funnel discharging into it under each seat. The bowls are constructed of earthenware or white enameled iron, and the trough or pipe with which they are connected is made of iron, and has a trap at its end under the discharge-plug. In the figure 90 THE PRINCIPLES OF the discharge-plug is hollow, and consists of a stand-pipe with overflow-passage through it. The height of the overflow regulates the position of the standing water in the bowls. The plunger or dis- charge-plug is under the control of the attendant, who flushes the closets as often as he considers it advisable. The bowls are so constructed that the waste matters fall directly into the standing water, and nothing strikes their dry sides ; they are thus partially deodorized. But the liquid and soluble portions of the solid wastes, which are allowed by the faithful attendant to remain for some length of time in the latrines, as well for the sake of economizing water as to enable him to attend to his other duties, soon precipitate a slimy deposit all along the inner surface of the closet, and particu- larly around the plunger-chamber. This is not easily removed, and always forms more or less of a nuisance. In most cases it will be found much better to provide, instead of latrines, a row of good hopper-closets, with treadle, door, or seat attach- ment for automatic flushing, if desired. SLOP-SINKS, SLOP-HOPPER SINKS, AND SLOP-HOPPERS. Figures 54 and 55 represent two kinds of fix- tures designed for the reception of slops. These have no means provided for the flushing of the walls of the sink. Either may be provided with a flushing-rim for the purpose. But the use of the flushing-rim in private houses is oftener neglected than observed. Servants will not take the trouble to thoroughly cleanse the slop-hopper at every usage, and it soon begins to emit a disgusting odor. In hotels or large clubhouses, where their use is constant and under systematic supervision, where special attendants are detailed to take charge HOUSE DRAINAGE. 91 of them, and where each story is independently pro- vided with a separate slop-hopper, their use may be recommended ; but in private houses they should Fig. 54. Slop-sink. never be allowed. A good hopper water-closet, with a strong enameled iron drip-tray to protect Fig. 55. Slop-hopper sink. the bowl, is much better, inasmuch as, while it serves the purposes of the slop-hopper equally well, it escapes its objections in ensuring a periodic flush- 92 THE PRINCIPLES OF ing. Every time the closet is used for the pur- poses of nature it is thoroughly flushed, and even slops are much seldomer allowed to stand in the bowl, because their pi-esence would be immediately detected by the next regular user of the water- closet, and the damage would be likely to "recoil upon the head of the offender." It is customary in private houses to place the slop-sink in the attic, but no house-owner can give any better reason for its existence than that he had seen it in some other houses. When valve, pan, and plunger closets were used to the exclusion of the more modern hopper, the slop-sink had a certain raison d'etre. In these closets, especially those requiring an overflow-pas- sage, the closure of the outlet is apt to cause an overflow of the slops when a large pailful is poured in quickly. But the modern hopper-closet has a clear, open passageway into the drains, and, being provided with the most improved form of flushing apparatus, is, in fact, the best form of slop-hopper that has been devised. Some persons who have insisted, even contrary to the advice of their archi- tect or sanitary engineer (who now unite in con- demning them), upon having the customary slop- sink duly installed in their attics, wishing to have at least an appearance of a reason for their way- wardness, urge that the virtue of the slop-sink lies in the strainer : this serves to prevent the obstruc- tion of the drain by scrubbing-brushes, rags, large cakes of soap, or other household articles used in scrubbing, capable of clogging the soil-pipe, which a careless servant might throw with the slops into the sink. This office of the strainer is certainly a use- ful one, and if every story in the house contained a slop-sink provided with such a guard, and every water-closet had a movable or portable strainer HOUSE DRAINAGE. 93 endowed with sufficient intelligence to close the out- let only when slops were poured in, the soil-pipes might really be protected from the gross careless- ness our friends so much feared. But as such a profusion of slop-sinks and strainers is evidently impossible in private houses, and as slops are col- lected in every story of the house as well as in the attic, and as no servant careless enough to throw scrubbing-brushes into a water-trap would take the trouble to lug slops from the lower stories up to the attic, in order to protect the neighboring water- closet trap from such an accident, or, in other words, mount one or more flights of stairs to avoid the trouble of removing the scrubbing-brush from the slop-pail before emptying the slops into the nearest water-closet bowl, it is evident that the argument of protection to soil-pipes has little weight. For hotels and some other public buildings, the slop-sink should have a good flushing-rim. It then becomes the so-called " slop-hopper," and the bowl should be properly protected by a stout iron drip- tray, properly supported, to receive the frequent blows bestowed upon the hopper by the careless pail. As they are generally made, urinals are very objec- tionable things in private houses. Urine undergoes rapid decomposition, and gives off a powerful and very disgusting odor. When in this state, it has the power of turning fresh urine into the same con- dition almost immediately, so that unless the urinal is so formed and placed that its surfaces are thor- oughly cleansed after use, it soon becomes a very foul and disagreeable fixture in a house. Fig. 56 94 THE PRINCIPLES OF represents the most economical form of urinal as they are now made. The bowl is generally constructed of glazed earthenware, with some form of fan or flushing- rim for spreading the flush- ing stream over its entire interior surface. The urine escapes through numerous perforations in the bottom and back of the bowl, into the waste-pipe. In some forms the trap is made in a single piece of earthenware with the bowl. There are a number of different forms of urinals, both swinging and stationary, and they are flushed either by a stop-cock directly on the supply-pipe, to be turned by hand, or by a special cistern. The for- mer method of flushing is open to the same objection Fig. 56. uriuais. as the direct supply to water- closets, and is now forbidden in some places by law. The pressure may be at times insufficient to fill the pipes, and the foul air from the surfaces of the urinal, perhaps contain- ing disease-germs, may be sucked into the supply- pipes on opening the stop-cock. In the figure, an automatic flushing cistern is used, which has within it a tilting vessel arranged to give a periodic flush as it slowly fills under a small faucet kept constantly open. This is perhaps the only certain method of ensuring a sufficient flush for single urinals con- structed in the usual way ; but it involves a great HOUSE DRAINAGE. consumption of water and is very wasteful, inas- much as the flushing goes on always, whether it be required by the use of the urinal or not. For private houses it is much better to construct the urinal in the manner shown in Fig. 57. It is a simple hopper-closet raised to the height of a urinal. By this arrangement, all of the advantages of a urinal are obtained, without any of the objec- Fig. 57. Combined urinal and slop-sink. tions. Moreover, by stepping on the steps or foot- rests at the floor in front of the fixture, the device serves equally well as a water-closet. The writer has found by experience that this form of urinal never becomes foul, nor is its use as a water-closet accom- panied by the least inconvenience. He has used it both in public and private buildings with equal 96 THE PRINCIPLES OF HOUSE DRAINAGE. success. The bowl, containing a large body of stand- ing water, dilutes the urine, and prevents it fouling the sides. Habit, with water-closets, leads to its flushing after its use as a urinal at times when the ordinary form of urinal would have been left un- flushed. But should, by any chance, the flushing be neglected, the next use of the fixture as a water- closet would insure its cleansing. Moreover, by combining the two fixtures in one, economy both of space and first cost is obtained, while the offensive appearance and smell of the urinal is avoided and the consumption of water is greatly diminished. Not the least of the advantages of this arrangement is that it is suitable for use by both sexes, a consider- ation of some importance, especially in the hall of a private house, where the want of space limits one to the use of a single fixture. In public buildings, however, such as hotels, rail- way-stations, manufactories, school or club houses, where proper and systematic attention may be ex- pected to be given to them, urinals may become not only desirable, but absolutely necessary. Stall urinals should also be constructed in various places in the main thoroughfares, easily accessible to the public, as an important sanitary measure. PART III. SOIL AND DRAIN PIPES. GENERAL CONSIDERATIONS. The material for our pipes naturally f oi-ms the first subject for consideration, inasmuch as upon it their proportion, treatment, and arrangement in a great measure depend. By far the most suitable material yet discovered for soil and house drain-pipes is iron, and the most important matter connected with its use is the for- mation of the joints between the separate pieces. Lead has been almost entirely abandoned in this country for soil and drain pipes, on account of its want of strength and rigidity, its comparative high cost, its liability to be perforated by vermin, nails, or corrosion, and of the greater time and labor required to make the joints. Large lead pipes often sag of their own weight and tear away at their points of support. The action of alternating hot and cold water also produces a destructive effect upon the material. In England lead soil-pipes are still used, but it is not customary to use the soil- pipe for the conveyance of all kinds of waste, and hot water from lavatories and sinks is carried into separate pipes, so that the material as used abroad is less objectionable. S. Stevens Hellyer, the well-known and popular English writer on sanitary plumbing, says, in speak- ing of the question as to the material most suitable 98 THE PRINCIPLES OF for soil-pipes : " This may scein a curious question to ask of plumbers as well ask a shoemaker of what material should boots and shoes be made ! Everybody knows that the latter would say : ' There's nothing like leather,' as the former is sure to say : ' There 's nothing like lead.' . . . Allowing experience to be my schoolmaster, I answer lead, especially for our climate." Mr. Hellyer claims the following points of superiority for lead : its greater smoothness, greater resistance to corrosion, greater ductility for bending to suit the various positions it has to occupy, more perfect jointing, greater adapt- ability for connecting with branch wastes, and greater compactness, which allows it to be placed in slots or niches smaller than those which are required for iron. He admits the following objections : its deterioration under alterations of temperature which tend to work it until it breaks, its sagging, its ex- pensiveness, its liability to be perforated by rats or carpenters' nails, its greater weight, and the require- ment of greater skill in making the joints. The advantages which Mr. Hellver claims for lead have within late years lost their force. Improved methods of protecting and jointing other materials have placed them in these respects far ahead of lead, as will be shown hereafter. White enamel is now applied as an inner coat to cast iron in such a manner as to render the inner surface as smooth as that of new lead. But in use lead soon loses its smoothness, the sewage adheres to the surfaces of the pipe to a greater or less extent and roughens it, in time, with a hard deposit of greater or less thick- ness according to the usage of the pipe, so that the difference in smoothness at the outset in favor of lead as compared with an iron pipe, properly coated, is of small consequence after a few years' use. HOUSE DRAINAGE. 99 The numerous cast bends and fittings now made and adapted to every possible turn or angle liable to be encountered in arranging the pipe renders the duc- tility of the lead pipe no longer of any advantage. Finally, other and more suitable materials are now jointed in such a manner as to render them quite as compact as the lead pipe. Stone and brick drains cannot be effectively flushed on account of the roughness of their interior surfaces. Moreover, they are porous to a certain extent, and the cement with which they are laid is always more or less pervious to water. Wooden drains soon decompose and leak, and when made of plank must be of such a section that scouring is impossible. Copper is easily corroded by the acids of sewage and decomposition, and it is, moreover, too ex- pensive when made heavy enough for the purpose. Zinc, tin, and galvanized iron are totally un- suitable, and not to be considered for a moment. In the worst kind of so-called ' ' Gerry buildings " they are, however, occasionally used. CAST IRON. Cast iron is the material which in this country for the last twenty years has been most generally used. It has in this time proved itself to be a most reliable and excellent material for soil-pipes. It is light, cheap, stiff, and strong, and it corrodes so slowly that, if of the proper thickness and quality of iron, and properly cast, coated, and put together, it will last as long as the house. The inconveniences at present attending its use are not inherent in the nature of the material. As now made, the pipes are often cast of uneven thickness, and they are always improperly jointed. Neither defect is necessary. 100 THE PRINCIPLES OF The experiments of M. Gaudin, made in 1851, show the maximum rate of loss by rust of uncoated cast-iron pipe exposed to the action of clean, fresh water on both sides to be a little over an eighth of an inch a century. His experiments extended over a period of thirteen years. With the present methods of protecting iron, its life can be very greatly prolonged ; indeed, even the use of the ordinary bituminous coating (coal-tar pitch) has proved, when it is properly applied, to be able to keep the pipe quite intact for twenty years. The life of a soil-pipe, even when quite thin and uncoated, has been found by experience to be so great that it is not unreasonable to suppose that the greasy matters contained in sewage serve to protect the pipe in a measure from the water and from the corrosive action of the acid components of the sewage. JOINTS. Equally important with the question of the mate- rial is the manner in which the several parts are put together, inasmuch as upon this depends not only the safety of the work, but also, in a measure, the choice of the material itself. The question of joining or coupling the pipe will therefore next be considered. The ordinary joint is neither tight nor permanent ; it cannot be made to resist water or gas under pres- sure, and it is soon destroyed by alternations of heat and cold in the pipes, such as are often produced by the passage through them of hot water or steam. It is expensive both in time and material. It re- quires expert labor to adjust, but defies expert labor to take it apart again without more or less destruction of the piping. Even the process of HOUSE DRAINAGE. 101 putting together involves a hammering which en- dangers the integrity of the pipe, and the most experienced and careful workman often cracks it in the process. The safe use of white enameled pipe is out of the question with the calked joint, be- cause the jarring produced by the calking-tool cracks the enamel. Fig. 58 shows the ordinary hand-calked joint. It is made with lead and oakum or jute. A gasket of jute or other similar fibre is inserted into the , _-L cavity of the bell or hub, and the spigot end of the length next above it is set firmly down upon it, or the II B gasket is rammed in with a tool W after the lengths are set up. The Bi gasket is used to prevent the lead P? from running out of the joint and obstructing the bore of the pipe at some point below, besides wasting ^ the lead. The lead is now poured Fig. 58. The ordi- ,, , , i jf nary beii-and- upon the gasket trom a ladle and spigot joint. shrinks as it cools. The calking- tool must then be used to expand it again and drive it into the cavities and pores of the iron. A faith- ful and skilful operator can by perseverance suc- ceed in fitting the lead into the iron at all parts of its circumference, so as to make it tight for a time, just as a painstaking dentist can drive the gold by patient labor into the cavities of a tooth, and tem- porarily arrest its decay. But the process in both cases is slow and uncertain. The dentist confines his calking to a single small spot well within his reach, and he labors with extraordinary care. Yet the filling often fails when put to the test. The plumber must work quickly over an extended field often in awkward positions ; he must perform a deli- cate task with clumsv tools. 102 THE PRINCIPLES OF The metals to be welded together are often so placed that it is impossible, without the utmost patience and skill, to reach them properly. The result is that when put to the test the joint almost always fails. Extra heavy pipe and hubs are re- quired to withstand the blows of the calking-tool. Lighter pipes cannot be made tight without danger of cracking the iron. It is now generally recognized and acknowledged that a plumber's calked joint is rarely either air or water tight, though a vast amount of lead and labor is spent on them to make them so. When we reflect that the sole aim and object of a soil-pipe joint is to make a gas and water tight connection between pipes, we see that the method commonly employed is an absurdity, and reflects little credit upon human ingenuity. Even supposing that, by chance, a calked joint has been made to stand the test which is now prop- erly required of it when new, its tightness is very soon destroyed by the expansion and contraction of the pipes caused by the passage through them of hot water or steam. The expansion of the spigot is in such cases greater than that of the hub, be- cause it is on the inside nearer the heat and not protected like the latter from the hot fluids passing through the pipes. Hence the lead is temporarily compressed between the spigot and hub, and, being inelastic, does not resume its original bulk when the pipes cool again. A minute opening is thus formed all round the spigot, as shown in the lower branch of the pipe in the cut, and the joint leaks. The object of requiring the whole system of pipes used in plumbing a house to be filled with water as a test is not only to determine the tightness of the joints in a manner which is impossible with the peppermint or smoke tests, since these can be HOUSE DRAINAGE. 103 eluded by a temporary coat of paint or putty, but also to try the quality and thickness of the metal. If a pipe is very defective in casting, its weakness will be revealed by a strong pressure test, and the faulty piece rejected. Another serious objection to this joint is the difficulty of disjointing pipes in which it is used. The usual way to take out a pipe, once so put together, is to break it to pieces, and then remove it by degrees. There is, in fact, no practicable alternative ; for to melt off the lead would not only be expensive and dangerous, but involve the dis- jointing of quite a considerable length of pipe in order to enable a single spigot to be lifted two inches, or enough to disengage it from its hub. Now alterations in our plumbing arrangements are necessarily so frequent that this objection becomes a serious one. The necessity of using fire in a house in process of construction for melting the lead necessary to make this joint is also a formidable objection to it, on account of the danger of igniting the surround- ing carpenters' litter and burning down the house. It is true that lead or solder melting would have to be carried on for other purposes, such as wiping the joints on the smaller pipes, but the less use we have for the solder-pot the less will be the danger, and the less the temptation for the workman to carry on the melting in dangerous places in order to save himself the trouble of running up and down wearisome flights of stairs to a place of safety. Still another very serious objection is the tempta- tion this joint opens for fraud. The lead may be partially or even wholly omitted without very great risk of detection, since it is out of sight, and fre- quently immediately covered by a coat of paint. 104 Tin: PRINCIPLES OF The calking may be still more easily slighted. If the hydraulic test is not demanded, a coat of paint or a little putty will easily make the joint stand the smoke or peppermint test. A few of the joints well within the reach of the house-owner may be filled with genuine lead, while all those which are covered by floor boards, or are not easily accessible, may be composed of paper and sand, and covered with putty. Possibly a thin coating of lead may be poured on top to present an honest appearance, and satisfy the suspicious and shrewd house-owner who goes about probing the nearest joints with his pen- knife in order to ensnare " the rascally plumber." Finally the bell-aud-spigot joint, when faithfully made, is very expensive both in material and labor. The amount of lead required for each joint, includ- ing waste, is estimated at about a pound for every inch in the diameter of the pipe. Thus an ordinary four-inch soil-pipe consumes four pounds of lead in each joint. The average length of time required by a skilful pipe-layer to make a single joint is estimated at twenty minutes, not including, of course, the plan- ning of the pipe system or the cutting and general arrangement of the pipe sections for their proper positions, a part of the work which has no connec- tion with the kind of joint used. THE " SANITAS" PIPE JOINT. Our " Sauitas " joint has been designed to obviate these defects and enable lengths of cast-iron pipe to be securely and economically united. In general terms it may be described as an adjustable flanged joint with lead washers or gaskets for packing. It is a steam-fitter's joint, with im- provements which adapt it for use in plumbing. HOUSE DRAINAGE. 105 The leaden gaskets are star-shaped in cross-sec- tion, and are crushed between the flanges of the Fig. 60. Detail of joini. Fig. 59. The " Sanitas " pipe. Fig. til. The half-ring. pipes to be connected by means of bolts and nuts. Figs. 59 and 60 illustrate the joint. In Fig. 59 the lower form of the joint is to be used in con- necting straight pipes, and the upper form in con- 106 THE PRINCIPLES OP necting branches and bends with each other and with straight pipes. The latter differs from the former in the addition of a half -ring just above the flange of the bend or branch, and is illustrated on a larger scale in Fig. 60. This half-ring is shown in detail in Fig. 61, and permits the bent or branched pipe to be revolved about its axis in setting, before the bolts are tightened up. The entire pressure is brought upon the ends by the small shoulders on the half -rings in such a manner as to prevent their fracture when the bolts are tightened up. The bolt- heads set in square recesses in the flange ears to prevent them from turning when the nuts are screwed home. Fig. 62 shows the lead packing-ring in perspec- 'ig-. OJ. Lead pa. tive, and Fig. 63 shows its star-shaped section in actual size. It is crushed to less than half its thickness by the pres- sure of the two half -inch bolts screwed Fig. 03. sec- up easily by a man of ordinary strength uonofpac ig- w - t k twelve or fourteen inch wrenches. The bolts are left-and-right threaded. Two ratchet- wrenches working in opposite directions are used to correspond with the reversed threading of the bolts. The pressure exerted by one wrench is thus resisted by the other. This avoids the necessity of securing the pipes while the nuts are being screwed up, and causes both sides to be compressed alike, HOUSE DRAINAGE. 107 since the wrench which has given and received the greatest pressure ceases temporarily to turn until the other has caught up with it. The " Sanitas " wrenches made for this pipe are so formed that the joint may be made in the most Fig. 64- contracted places, as shown in Fig. 64, where the stack stands at the bottom of a slot a foot deep and only eight inches or one brick in width. Fig. 65. " Sanitas " ratchet-wrenches. Fig. 65 shows the ratchet- wrenches. By the use of these ratchet-wrenches the joint may be thoroughly calked by a single ordinary unskilled workman, after the pipes are once set in place, in less than twenty seconds. To calk an 108 THE PRINCIPLES OF ordinary bell-and-spigot joint in the usual defective manner is estimated by good authorities as requiring on the average, when the pipes are once in place, as many minutes. The amount of lead used for calking our flanged joint is about one eighth that required for the ordinary joint. The lead gasket for four-inch pipes weighs half a pound ; and for two-inch pipes, one fourth of a pound, while the rule for calking ordinary joints is to use one pound of lead for every inch in the diameter of the pipe. We also save the fuel, oakum, etc., used in making ordinary joints, and avoid the danger of lead-melt- ing in houses. Figs. 66 and 67 show in section and elevation the simple method of connecting lead and iron pipes when the flanged joint is used. With ordinary bell-and-spigot pipes the proper connection between lead and iron is both laborious and expensive, re- quiring the use of brass ferrules. The lead pipe has to be wiped on to the brass ferrule, and the brass must be calked into the iron hub. A double joint is thus required, and this, especially with the larger pipes, in- volves the use Of Fig. 67. Perspec- considerable skill tive vlew - and valuable material. With our flanged joint all this is done away with. The lead is simply flanged out to correspond with the flanges of the iron pipe to which it is to be connected, and bolted to the pipe by means of a cast- Fig. 6. Sectior HOUSE DRAINAGE. 109 iron ring furnished with the pipe-fittings, and having ears and bolt-holes corresponding with those of the pipe-flanges. The lead packing-ring is used be- tween the lead and iron flanges exactly as if the flanges were both of iron. In this manner a per- manent steam-tight joint is formed between the two metals without hand-calking, brass ferrule, or joint- wiping. In bell-and-spigot pipes comparatively few bends and branches are made. Should the angle required to reach a certain fixture in laying the pipe be a different one from that given by the bends fur- nished, the desired direction must be obtained by canting the spigot slightly in the socket, a move- ment different from the axial rotation we have already described and provided for. To accomplish the same result with flanged pipes a greater variety of castings are made, furnishing bends of a larger number of angles, in the same manner as is done in wrought-irou piping when used with screw-joints for plumbing purposes. With the flange-joint a certain play is obtained by screwing up that side of the pipe upon which the greatest inclination is to be given slightly more than the other. But the variety of castings furnished enables every requirement to be met, without resorting to the method of unequally compressing the packing-ring. "We find , , , T X 7 , ^ 7 , and -^ bends corresponding to angles of 90, 60, 45, 22, lli, and 5f respectively. By using some one of these bends, or a combination of two or more, any desired direction can be obtained. The half -rings are required only for bends and branches. Straight pipes are screwed together directly, and have ears and bolt-holes at both ends. The straight pipes are manufactured in lengths of 1ft., 2ft., 3ft., 5ft., and upward; 10 inches, 9 inches, 8 inches, 110 THE PRINCIPLES OF 7 inches, and 6 inches. Pieces having a bell or a spi- got on one end and a flange on the other are also made for connection with old bell-and-spigot pipes, or for substituting bell-and-spigot pipes at any point de- sired in case suitable " Sanitas " fittings cannot be obtained at momentary notice or in out-of-the-way places. Moreover, by using these connections, any one desiring to try the " Sanitas " pipe for the pur- pose of comparing it with the ordinary pipe before using it exclusively can incorporate in his piping a few lengths of the " Sanitas " pipe, and observe its action under hydraulic pressure and otherwise, as compared with that of the ordinary pipe, of which the remainder of his work may be composed. Al- though this joint requires less space for setting than any other, it is still always best to give ample room for it, especially if it be set in slots, and particu- larly if the plumber is accustomed to the old bell- and-spigot pipe only. It is recommended never to set soil-pipes in slots. But if it must be done, the slots should not be over four inches deep or less than one foot wide. As already explained, the hydraulic test, which should in every house be required before the work can be pronounced safe, is on ordi- nary bell-and-spigot pipes very diffi- cult of application, because there is no easy method of temporarily closing the outlets. Here again our flanged joint presents an advantage of great importance. In order to close the Fig. <>8. cupped P en i n g it is only necessary to screw end. on caps provided with ears and bolt- holes corresponding with those of the pipe-flanges, as shown in Fig. 68. The regular packing-ring is used between the cap and flange, so that the joint HOUSE DRAINAGE. Ill is steam-tight, like the rest of the piping. When the test has been made the caps can be removed and used again and again by the plumber. They are furnished with the rest of the pipe and fittings. The lead rings after use can be used for old lead or recast into new rings. It is sometimes required in practice that each pipe used be tested at the foundry before coating it, in order to ensure soundness. With ordinary bell- and-spigot pipes, the application of a pressure test is difficult, if not impossible. The straight lengths can be tested under pressure, but the branches and bends offer serious difficulties on account of their form. Hence the oil test has to be resorted to, and the strength or thickness of the pipe is not by this method made known ; moreover, the oil test is in many other respects obviously inferior to a strong pressure test. A simple machine has been devised to test these flange-pipes. It consists of two plates with rubber disks on one side, which are pressed against the flanges of the pipe to be tested, by means of clamps and wedges of peculiar form, designed for rapid application. One of the plates is perforated and connected with a water-pipe and pressure-gauge. A simple force-pump is added, so that where the water-pressure is subject to consider- able fluctuation, each pipe may be tested under pre- cisely the same pressure. By the use of this device, flanged pipes of any desired size, and all the branches and fittings, may be quickly and accurately tested at the foundry before coating. In the case of ordinary bell-and-spigot pipes, the expense and imperfections of the jointing are so great that the pipes are cast very long, in order to save joints as far as possible. The attempt to cast pipes of small diameter, say two-inch, three-inch, 112 THE PRINCIPLES OP and four-inch, in lengths of five feet, is almost certain, unless special precautions are taken, to result in an inequality in the thickness of the metal. The writer has found bell-and-spigot pipes of five-feet lengths, made by the best firms ^^p^^ and sold for extra heavy weight, no ^^^ ^^^ thicker than a piece of thick paper J ^ on one side and half an inch on the I other. Fig. 69 is an accurate draw- ^ J ing of a two-inch pipe which he has ^k J recently been obliged to reject, ^^i \-^ among a large number of others from Fig. 69. imperfec- the best makers, upon testing them tions of long before they were laid by the plumber, castings. j t j g muc ]j more unusual to find pipes of equal than of unequal thickness throughout. This is a very important consideration. The strength and thickness of a line of piping is equal to its thinnest part, as the strength of a rope is equal to its weakest part. Hence all the metal used in the piping of a house beyond the thickness of its thinnest part is thrown away. Of what use is it to pay for extra heavy pipes, when one side of most of them is extra light ? It is not for the strength of the piping that we require the thickness, since they are not used like columns to support floors and walls, but for security against leakage and decay. Now since, as is very well known by plumbers and engineers, the majority of long pipes of small sizes are uneven in thickness, the chances of obtaining only the even pipes throughout an entire stack are obviously infinitesimally small, and it is not proba- ble that one house in a thousand exists in which one or more of the lengths of pipe are not very seriously uneven. The enormous waste of metal and the great danger of leakage which this condition of HOUSE DRAINAGE. 113 things implies renders it of the utmost importance to employ some means of remedying this great defect. We find a remedy in using short castings as far as possible, and in casting the long pieces with un- usual precautions and in a different manner from the short ones. The plumber would find great advantages to offset the inconveniences in using a variety of short lengths of pipes, instead of fre- quently cutting the usual five-feet lengths to fit the spaces between the floors and fixtures. Cutting cast-iron is an extremely difficult and tedious pro- cess. Were these various lengths manufactured from six inches upward, he would find it possi- ble to avoid cutting entirely, and probably add very few joints to the number now required, for it must be borne in mind that each time a pipe is cut a new joint is necessitated, so that the saving in the number of joints in ordinary plumbing practice, by using no other than five-feet castings, is much smaller than is at first supposed. Now, however, that we have found a simple, safe, and economical joint to take the place of the clumsy, uncertain, and expensive one in vogue, we have no further need of long castings. The saving in pipe-cutting, to say nothing of the other advantages, far more than offsets the labor of making an extra joint or two, and we have a stack of pipes whose thickness can be relied upon as being uniform throughout. Where it is found necessary for any purpose to remove a piece of pipe from a stack already set up it is only necessary to place temporary supports under the pipe above the one to be disjointed, un- screw the bolts, remove one of the lead rings by means of a chisel or saw, and slip out the length to be removed. 114 THE PRINCIPLES OF To replace a pipe or fitting several methods may be employed, of which the best is that in which short flanged or threaded brass pipes are used. To the lower end of the pipe to which the new piece is to be connected is bolted a piece of short flanged brass pipe. Another short flanged brass pipe is then slipped over the first, being made just large enough to do so, and the fitting to be connected is afterward bolted to its flange, and to the main piping below. The two brass pipes are finally connected bv means of an ordinary wiped solder joint. Or brass pipes may be screwed to iron flanges, to save the more expensive metal. Another method is to substitute iron for the brass pieces, one of the pices being provided with a hub, and calk the joints by hand in the usual manner. This latter method, however, has the objections of all hand-calked joints, and is for this reason not to be recommended when the first can be applied. By this means the flanged pipes may be connected with old work in which bell-and-spigot pipes have been used. Or bell-and-spigot pipes may at any place be inserted in the line of flanged pipes in this manner, if desired. As we have already explained, the bell-and-spigot joint is incapable of withstanding the effects of sudden and severe variations of temperature. The spigot being nearest the heat expands more than the hub and compresses the surrounding lead, per- manently diminishing its bulk, and forming a passage for the escape of gas. The principle of the construction of our flange-joint is such that this trouble is overcome. The flanges are affected equally by changes of temperature, and the lead packing is never compressed by expansion or con- traction. Thus supposing, when the pipes are cool, HOUSE DRAINAGE. 115 steam is suddenly allowed to pass through them. Both upper and lower flanges and the lead firmly imbedded between them expand alike outward under the same degree of heat, and return again unaltered as the pipe recools. The bolts expand and contract with the changes of temperature proportionally with the flanges, and do not affect the packing. To give these theories a practical test, I have had some four-inch piping connected and closed up at the ends with our flanged joints, and coupled the whole with the boiler of a steam-engine, the steam- gauge indicating about thirty pounds pressure. The steam was left on until the pipe-flanges and bolts had all become thoroughly heated through. The coupling was then immediately transferred to the cold-water supply from the city main, and after the steam had been let out the cold water was suddenly turned on until the piping was filled. As the experiment was performed in midwinter, the test was as severe as possible. The cold watjr was then poured out and steam again immediately applied. This alternating application of steam and cold water was repeated successively a dozen times. During the entire process no sign of a leak either of steam or water was obtained. The bolts had been screwed up in the ordinary manner without extra care. It is well known that no bell-and-spigot joint will stand such a test even after the most careful calking. The same variations of temperature cause the pipes to expand and contract also longitudinally. But in this direction there is always ample play left in setting the pipes for this action, and the lead is obviously not affected by it. Each packing has 116 THE PRINCIPLES OF upon it the weight of all the pipes above it, as well as the pressure exerted by the bolts. The weight is therefore constant, whatever be the temperature or length of the pipes, provided thev are properly set. The expansive force of iron is so great that if free play is not allowed for it in a building, it will make way for itself by tearing away its bonds. Mr. Bayles says : "Insetting up a line of soil-pipe, intelligent provision should always be made for ex- pansion and contraction of the metal resulting from changes of temperature. These changes, however, are seldom sudden or extreme ; but when the pipe is at any point rigidly fastened to the wall it ex- pands in both directions. The amount of motion at the ends is small, but it must be provided for, or it will provide for itself. The power with which iron expands, as its temperature is raised, is practi- cally irresistible. The end of a pipe may not move more than an eighth or sixteenth of an inch, but the power with which it moves that distance is so great that it can only be resisted by a power great enough to crush the metal. This would be, in ordinary cases, equal to about 75,000 pounds per square inch, the strength of cast-iron to resist crushing strains being from 60,000 to 90,000 pounds per square inch. Consequently, we see that unless the fastenings at the ends of aline of cast-iron pipe are of such a character as to admit of slight movement, something must give way, and it is not likely to be the pipe. This, then, must be provided for in the character and position of the fastenings, which must be so arranged that, while allowing for some movement, they shall not develop a tendency to break or loosen the joints. Under ordinary condi- tions the amount of expansion is seldom great enough to give much trouble, but when steam or a HOUSE DRAINAGE. 117 great volume of very hot water washes into an iron pipe it is sometimes great enough to loosen joints and even crack the pipe." Accordingly, if a line of pipe is rigidly fixed at the bottom, the hooks which hold it against the walls should be placed a short distance away from the flanges, so that the line of piping is free to slip up and down slightly under the influence of expan- sion and contraction. Otherwise these hooks are liable to be loosened from the mortar or woodwork into which they may be driven, since it would be easier for the pipe to loosen the hooks in the mortar or wood than to further compress the packing-rings, or to stretch out the heavy bolts of wrought iron. The opportunities furnished by the ordinary bell- and-spigot joint for careless or fraudulent work are avoided in our flanged joints. The entire thickness of the lead is visible from the outside between the flanges. As the lead is the only packing used, and as this is in open view, nothing can be fraudulently omitted. The bolts and nuts are also visible and, moreover, must be of the standard size and strength in order to furnish the requisite amount of com- pression to stand the hydraulic test. Thus we find in our " Sanitas " joint all the char- acteristics demanded for plumbing purposes. 1. It is water, gas, and steam tight, even under heavy pressure. 2. It is unaffected by the expansion and con- traction of the pipes. 3. It is capable of resisting severe jars and strains, both compressive and tensile, such as are occasioned by the weight of the pipe, or by settlement and movement in the building. 4. It requires neither skilled labor nor machinery in its manufacture or in its jointing. 118 THE PRINCIPLES OF HOUSE DRAINAGE. 5. It is of such a form and nature as to admit its being as easily taken apart for repairs or alter- ations as it is put together, and this without damage to any part. 6. Its form and construction are such as to allow it to be made and put together rapidly, to follow easily the irregular contour of the construction, and to be used immediately after fixing in place. 7. It requires no hand-calking or hammering, which are liable to fracture the pipe or its lining. 8. It is so formed that any imperfection, either in the materials used or in the manner of putting them together, can easily be detected at a glance from without, without expert aid. 9. It is compact enough to permit its use in the most contracted spaces. 10. It causes no obstruction to the water-way, and leaves no appreciable space or pocket for deposit. 11. It is simple, durable, indestructible, economical, and unobjectionable in appearance. It is, therefore, suitable for water, gas, and steam under pressure, as well as for drain and soil pipes. SIZE AND GENERAL ARRANGEMENT OF THE PIPING. HAVING thus described a safe and economical method of jointing our cast-iron pipes, it remains to consider their proportions and general arrange- ment. We shall, as treating of house drainage, confine our attention to the piping of the house proper ; the consideration of the drainage beyond the house limits belonging more properly to the subject of sanitary engineering. The size of soil and drain pipes should not ex- ceed four inches. This is ample to carry off every possible form of discharge or combination of dis- charges to be met with in plumbing, even in the largest buildings, except for special hotel, laundry, or manufacturing purposes. It is a mistake to sup- pose that because the fixtures are multiplied, the diameter of the soil-pipe must be multiplied corre- spondingly. It is a rare occurrence, even in a hotel- building, that several water-closets are flushed at exactlv the same instant, and even if they were, their distance from each other and the capacity of a four-inch pipe would give ample room for the escape of the water. The choking of a pipe is far oftener due to its being too large, or to faulty construction, than to its want of sufficient size. The great per- pendicular extension of plumbing-pipes also facili- tates the discharge, and pipes which for land-drain- age would be much too small, will be found ample for plumbing purposes on this account. 120 HOUSE DRAINAGE. When we consider how important it is that the soil and drain pipes should be as thoroughly scoured as possible by the discharges sent through them, and remember that the smaller the pipe the more perfect the flushing, we should be inclined to reduce our soil-pipes to a size even smaller than four inches, were it not for the careless usage slop-hoppers and water-closets are so often subjected to. The traps of water-closets are very frequently made of pipe as much as four inches in diameter, and large pieces of newspaper are often used where toilet-paper alone is suitable. This, and the care- lessness of servants who will throw into a closet any- thing which is small enough to pass through its trap, would cause a great amount of annoyance and ex- pense if the soil-pipe were smaller than the trap of the slop-hopper or water-closet. Hence we have fixed upon four inches as both the smallest and the largest size of soil and house-drain pipes, and be- lieve that no other size should be used except for rare and exceptional cases. All the piping of a house should be in full view. Nothing should be walled in or covered over and rendered inaccessible. One of the first rules of modern sanitary work is to bring everything out of the darkness into light and air, where defects, if they occur, can at once be detected and removed. We are accustomed to running our steam-pipes in plain sight, and rendering them, by gilding or silver- ing, as ornamental as possible. The same custom is now beginning to apply to our plumbing-pipes. Where they pass through parlors or reception-rooms, they should stand behind movable panels or doors : a little ingenuity on the part of the architect will generally enable this to be done with good effect. HOUSE DRAINAGE. 121 The piping should be arranged to run as direct as possible, and should be concentrated. It will be found very convenient, especially in city houses, to build a broad recess, or slot, in the masonry of the party- wall, on the line of the bath or toilet rooms. for all the plumbing and ventilating pipes which can be collected together in this neighborhood, and, if possible, to run up in this slot the smoke-flue of the furnace, in iron. The heat of the smoke-flue will create a strong circulation in the ventilation-pipe, and at the same time radiate a useful heat into the bathrooms. The brick recess should be enclosed in masonry on all sides where it passes through the floor, and as high as three or four feet from the ground. This serves to protect the woodwork from clanger of overheating. Between this height and the ceiling the recess is open in front, exposing the pipes and admitting the radiant heat from the flue. The various stories are separated from each other by brick platforms, built across the recess on the line of the floors, and made tight around the pipes with cement or mortar. Above the upper bathroom the iron smoke-flue enters a regular brick flue, and the soil-pipe ventilator runs up independently through the roof. The writer has adopted this system in several city houses and found it very satisfactory. The furnace smoke-flue may be con- structed of tile instead of iron, if preferred, for greater durability ; but if iron be used, it may be made heavy enough to last as long as desired. The recess being, moreover, accessible, the pipe may be renewed at any time without difficulty. Every stack of soil-pipe should be thoroughly ventilated, by being extended full-size from the bottom to, and through, the roof. No ventilating- pipe running through the roof should be of less 122 THE PRINCIPLES OF diameter than four inches, inasmuch as smaller pipes are liable to become clogged in winter by snow and frost. The extensions above the roof should not be less- than two feet high and the tops should never open near a chimney - top, ventilating - shaft, dormer- window, or other opening, for obvious reasons. It is generally sufficient to allow the pipes to remain wide open, without return-bends or ventilating-caps, which only serve to obstruct the circulation. Wire nettings may be put over the opening at the top, to prevent objects from falling into the pipe, through accident or malice. These iron ventilating and soil pipes form the best possible lightning-conductors, because they are always sure to have a good and moist ground- connection, and are composed of a bodv of metal heavy enough to carry the most powerful charges of electricity without danger of melting ; their presence, therefore, in sufficient number, renders the usual form of lightning-rod superfluous. The soil-pipe should be firmly suported at the bottom. The best support consists in the projection of the foundation-wall, or in a stone or brick pier made for that purpose. The junction between the soil and drain pipes should be made with an easy bend, of as large radius as possible, to prevent the accumulation of obstructions and the powerful back- pressure on traps caused by the friction of the air in attempting to pass round a sharp bend in front of a descending column of water in the soil-pipe. The main-drain should run along in full view on the foundation- wall, if possible, or supported by piers resting on the concrete, or hung from the joists by strong iron hangers. Clean-out openings should be provided at all places where sediment or obstructions are liable to collect. HOUSE DRAINAGE. 123 Sometimes it is found convenient to rest the drain directly on the concrete. In this case it is custom- ary to 'form the concrete in a trench whose bottom pitches with the proper grade to accommodate the drain. The drain should have a fall of half an inch to a foot, if possible, or at least a quarter of an inch to a foot ; of course, the greater the pitch the better. The main-drain should be trapped with a running- trap of iron just inside the cellar-wall, or, if this is impossible, outside the house, in a manhole. The trap should always be accessible and should be pro- vided with clean-out caps with air-tight covers. It is a good plan to run a water-conductor into this trap, to ensure its occasional flushing. To provide for a complete circulation of air through the soil and drain pipes, a fresh-air inlet of the full size of the drain should enter it just inside (on the house side) of the main-trap above described. The mouth of this inlet should open outside of the house, at some little distance from any door or window. Where a fixture is connected with the rigid iron soil-pipe stack, provision must be made for a certain degree of movement or play on the part of the fixture, in such a manner that the movement shall not crack the joint or in any way loosen it. A settlement of the masonry, a jarring or shrinkage of the floors is certain in a new building to alter to a greater or less extent the relative positions of the fixture and its soil-pipe connection. Injury to the joint from this cause may be prevented in two ways. One of these is to use a sufficient length of lead pipe in all cases between the fixture and the iron stack, and the other is to support the fixture directly on the stack itself, and make it entirely independent of the floors and woodwork. 124 THE PRINCIPLES OF The first method is used with cast-iron piping and with all fixtures having waste-pipes of small calibre. Lead piping has so much flexibility that a section of even moderate length will permit of a considerable movement on the part of the fixture without injury to the joint. Where the fixture is a water-closet, a length of lead pipe having a horizontal extension of two feet, or its equivalent, in any inclined direction, will permit the utmost shrinkage of joists or settlement of walls liable to occur in good plumb- ing after the plumbing is connected, without injury to the work. The second method, that of supporting the fixture directly on the stack, has been successfully employed with wrought-iron piping. The closet sets on a cast-iron base firmly attached to the piping, so that a shrinkage of the floor may take place without affecting the joint. It is important that all angles and bends in our piping should be as smooth and gradual as possible. No sharp angles should be allowed. Thus, if it is ever necessary at any point to use T-connections, they should not be formed as they are in common bell-and-spigot fittings, but should be made with a curve at the branch junction. But Y-branches are to be preferred to T-brauches, and it is very seldom, if ever, that T's are required. The peppermint and smoke tests are useful for application at any time after the house has been occupied, when it is desired to ascertain if the pipe system has remained sound throughout, especially in places where a leakage of water might not occur. To apply the peppermint test, the vent-openings are first to be closed with plugs. A two-ounce bottle of oil-of-pepermint is then carried up to the roof by an assistant, and its contents poured into the soil- HOUSE DRAINAGE. 125 pipe at its mouth above the roof. A pail or pitcher of hot water is immediately poured down after it, and the opening is then plugged up. The assistant remains upon the roof until the examination within the house has been completed ; otherwise the odor clinging to his clothes will be likely to follow him into the house. The peppermint is volatilized by Fig. 70. Asphyxiator for applying smoke test. the hot water, and should any leak occur it will at once be detected and located by its pungent odor, unless the pipes have been improperly imbedded in the walls or are so covered up that access to them is impossible. If they are set as they should be, everywhere in open view, no difficulty will attend the detection and repair of the minutest defect or leak. The smoke test is applied by means of special bellows manufactured for the purpose (Fig. 70). It enables those whose sense of smell is not acute to operate instead with the sense of sight. RUSSELL & BURKE, No. 18 Howard Street, BOSTON. Sectional Steam Boilers, WROUGHT AND CAST IRON AND BRASS PIPES, FOR STEAM, QAS, AND WATER PLTJMBINGK Tubular and Artesian Wells Sunk. Brass and Iron Fittings, Coils. Heaters, Pumps, and all kinds of Steam and Gas Fitters' Tools, Hot Water and Steam Wanning and Gas Lighting Appa- ratus. Also, Plumbing Materials furnished and put up. ESTIMATES GIVEN FOR HEATING PRIVATE HOUSES PUBLIC BUILDINGS, ETC. DANIEL RUSSELL, JR., T. F. BURKE, Everett. E. Somerville. PORTLAND fftone Vta (Jo. SALT GLAZED SEWER PIPE, GREASE TRAPS, FIRE BRICK, TERRA-COTTA VASES, Send for Catalogue and Price-lists. No, 8 Liberty Sq,, BOSTON, MASS. GEO. C. DUNNE, General Agent. JfiMES TUCKER. FRflNK R. TITUS. TUCKER & TITUS, plumber 1 ^ panita) Engineer, No. 9 Pemberton Sqnare, Orders by Telephone No. 739. THE LIBRARY UNIVERSITY OF CALIFORNIA Santa Barbara THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW. UC SOUTHERN REGIONAL LIBRARY FACILITY A 000587914 3