It is ordained that access openings must be provided, in itself a desirable thing, but these access openings are described by a proprietary trade name, by which the great majority of makes are excluded. This illustrates the need for excessive care in drawing up any compulsory provisions. Doubtless in this case the fact of the name being a proprietary one had escaped the notice of the framer of the regulations, but the result is none the less indefensible. Bye-laws specifying position and construction of drains require much more careful consideration than they have hitherto got. The second provision to which I have referred, that all soil and waste pipes must be outside, has been for years a bone of contention between architects, whom I might call the more artistic class of sanitarians, and the more utilitarian, or at least less artistic classes. In many cases I confess that my sympathies are entirely with the architects. Perhaps in a new building it is reasonable to expect that the architect will either arrange the fittings so that his pipes may come down in some inconspicuous place, or design his elevations so that the pipes will not be an eye-sore. Perhaps he might even make them a feature. But in the case of an old building being remodelled, it is painful even to the ordinary eye, and must be doubly so to the artistic eye, to see the hideous monstrosities sometimes perpetrated in the name of sanitation. Another aspect was suggested to me recently. In small tenement houses it may be impossible to give each house a closet to the back, and rather than disfigure the front with a projecting soil-pipe the builder is tempted to resort to the expedient of a "common" closet, an arrangement, to my mind, much more objectionable than an inside soil-pipe.

The third provision, that no sink, bath, or basin should join a soil-pipe is another example of an excellent general rule, run to something like an extreme. I think I can trace its origin to the days when fixed basins in bedrooms were the rule in all houses with any pretence to luxury, and when at the same time sanitary construction was at a very low ebb. The early sanitary reformers found that many of these basins were connected direct to a soil-pipe, and that the connection was so bad that gases from the soil-pipe readily found their way into the bedroom. I have myself written many a report condemning this direct connection between a bedroom basin and a soil-pipe. In like manner a sink in a kitchen, which is virtually a living room, is so far open to a similar objection. But when in the same apartment we have a

closet, a bath, and a basin, each protected by a trap of practically the same resisting power, I never have been able to see that there was any absolute sanitary necessity for putting up two pipes. On the other hand, there is not only the financial objection, but the sanitary objection that two pipes, each of them more or less fouled, are used in place of one, and further, that the water which might be used with advantage to flush one pipe is divided between two. I am aware of two arguments in favour of the two pipes. One is that in some classes of houses the bath might remain so long unused that the trap would dry up, and the other is that the action of the hot water might tend to produce leaks in the soilpipe, if it were admitted to it. It may be questioned, however, whether it is desirable or even safe to put baths into a class of house where there is not a reasonable chance of their being used, and the other argument has lost much of its force in view of the much higher standard of construction which has been developed in the last few years. I am not advocating the abolition of waste pipes, but I believe that, as a compulsory requirement, the present enactments are too sweeping.

In these examples I have endeavoured to show that it is possible, by stringent and unbending application of rules which from a general point of view seem to be very desirable, to produce results which are not in accordance with what is desired and expected. The remaining provision, that which specifies the testing of drainage systems, introduces a new element.

The provisions with regard to drains and soil pipes being outside, and waste pipes separate from soil pipes, suggest a sort of nervous timidity, which refuses to believe in the possibility of such appliances being made and kept perfectly air-tight. The provision. for burying a presumably good pipe in concrete suggests a ludicrous and humiliating confession of its inefficiency. Fifteen or twenty years ago, unfortunately, such a state of mind was more than justified, but the science and art of construction have made some progress since then. Judging from drainage bye-laws, one would scarcely come to that conclusion. Not only have we such regulations as those to which I have referred, but the methods of testing are no more up to date.

If the regulations as to testing have any meaning at all, their object is to ensure that every system constructed in accordance with them will be air-tight. The provision usually is that the whole system is to be tested to the satisfaction of the Sanitary

Inspector, sometimes to be smoke-tested to the satisfaction of the Sanitary Inspector. All drains are to be left uncovered till this is done. How does this official set about the test? Almost invariably by filling the system with smoke under a slight pressure, and then by sniffing and searching along drains and soilpipes, using nose and eyes to detect any escaping smoke. If none is found, the drain is passed; if any is found, something must be done to repair it. Does the failure to find any such escape prove the system to be air-tight? Speaking from the experience of literally thousands of such tests, I say without hesitation that it does not. In many cases it is strictly accurate to say that nothing was found wrong; I know of no case where it could be said that everything was proved to be air-tight. The result depends on the wind, the weather, the inside temperature, the outside temperature, the eyes of the inspector, the nose of the inspector, his general intelligence, his experience, and I know not how many more contingencies. To decide whether the drainage system of an existing house is sufficiently tight is one thing, and for this purpose the smoke test is the best thing we have got. To decide whether a system which we are constructing afresh, and which we aim at making air-tight, is so in reality; is a totally different thing, and for this purpose the smoke test is utterly inadequate.

In March, 1897, a paper was read to the Architectural section of this society by Dr. Neil Carmichael, to whose scientific researches sanitation owes much. It contained the following:"If our pipes were made in their substance and in their joints as tight as they ought to be, then a pneumatic test-air pressure with a pressure gauge-would determine absolutely if a pipe were sound. If defects were found that is to say, if it were found not air-tight-then the location of the defect might be sought by the smoke test." I was present at the reading of this paper, but I fear had been somewhat inattentive, for this part had entirely escaped my memory till I came across it in the Transactions a few days ago. Meantime, about the close of 1897, circumstances had led me to make some practical experiments exactly on the lines therein indicated. I embodied the result of these investigations in a paper which I read before the Royal Scottish Society of Arts last November, under the title, “A suggested standard for drain testing." In it, while referring to the work of various others, I overlooked Dr. Carmichael's important suggestion, an omission which I am glad now to have the opportunity to repair.

My suggested standard was practically the same as Dr. Carmichael's, namely, air-tightness under a given pressure, as tested by a pressure gauge. I had, however, by that time gone a step further, and put the suggestion into actual practice, embodied in the instrument which is here illustrated. I do not of course assert that this is the best instrument that could be devised for the purpose, but I do assert that it effects the purpose, and enables me to refute any allegation that the proposed standard is impracticable. A gasfitters' gauge, or, in more scientific language, a manometer, a few drops of water or of mercury, a bicycle pump, a bicycle valve, and two or three yards of rubber tubing, with a little brass T piece, constitute the whole affair. As it stands, with water in the gauge I can test any drainage system with a pressure up to 4 in. of water. By using mercury, the pressure can be increased up to about 4 feet of water. When it is applied to test a system which is sealed by means of water traps, of course the maximum pressure that can be applied is that which is just sufficient to force the weakest trap. Incidentally, therefore, it will be seen that the instrument measures, and measures accurately, the strength of the weakest trap. If any more air is forced in, it simply bubbles through the said trap, and the gauge will rise no higher. Before beginning operations, all the legitimate openings, air-pipes, unfinished branches, and the like, are carefully closed up. The pump is wrought until the desired pressure has been reached, or until the gauge will rise no further. Pumping is then stopped, and the gauge is watched. If there is any leak, of course the gauge gradually falls, and, as Dr. Carmichael suggested, the leak is sought by using a smoke machine. If it is considered necessary to test to a pressure greater than that of the weakest trap, then the traps may be tightly plugged, and the pressure got by using mercury, combined, if you like, with a longer glass. So much I could have said without any practical trial. But it is nearly two years since I made this instrument, and since then I have used it on every opportunity. The results have more than confirmed my previous opinions, and a good deal of what I put even last year as probable, I now give as the result of experience. I have applied the test to every new drainage system which has been carried out under my supervision, and for a very considerable time I have for such a purpose entirely abandoned the smoke test. Again and again I have found that a system of pipes which had triumphantly withstood the smoke test was so far from