Some drains have an outlet-shaft on the soil-pipe, but are not "intercepted" from the sewer. Even in this case, if the outlet-shaft is closed after the smoke has begun to issue from it, the test will be good, although the lower end of the drain is open to the sewer. If no smoke escapes at the shaft, or at any or all of the unsealed traps, - a sign that the draught is towards the sewer, - then the test is not reliable, and must be repeated under different atmospheric conditions.

In drains which hare intercepting traps, and where the vent-shafts can be stopped, the smoke-test will, if continued a reasonable time (say an hour), invariably discover a fault if it exist. Where cement or asphalt covers the yard, small holes may, with the permission of the occupier, be "picked" through it over the supposed line of drain. This, however, need only be done if no result be found without it, and if the experienced inspector is led from other observa-tions still to suspect the soundness of the drain. The holes must, of course, be afterwards filled with cement

If the light of a taper or candle be held over a hole or crevice in pavement or wall, or passed along a skirting, whilst a smoke-machine of the kind already shown is being worked, the flame will pulsate with every beat of the machine if the drain is not sound, and if their is ever so small a connection through the ground between the hole or fissure and the faulty part of the drain. This test is as unerring as it is delicate, and often reveals a defect when even the smoke refuses to traverse the passage, seeing that it does not depend upon the transmission of anything, either scent or smoke, from the machine to the flame, but upon a wave motion pneumatically imparted to the intervening air.

The water-test consists in filling the drain with water to the top of the highest gully. This is a posit ire test, whereas smoke, as already stated, gives in many cases only negative results. It is, in some cases, too positive, because it will make a nuisance where none existed before, as in a well-made clay-jointed drain. The water-test can only be applied to old drains if access-chambers are provided upon them. Inspectors are not empowered to dig down to and sever drains, in order to plug them for the water-test, and although drains ought to stand chokage without emission of sewage through the joints, the inspector has no right in his search for nuisances to blow out by his tot the joints of a clay-jointed drain that has not been choked, and which are efficient whilst the drain remains clear. If therefore, where access can be had, he applies the water-test to an old drain, without the consent of the owner, he risks rendering himself or his authority liable to an action for damage, as it may be alleged that the drain now leaking was previously tight, and not in such a state as to be a nuisance, which allegation he may be powerless to refute. The smoke-test, however, is applicable to all drains at all times, and, if it be properly applied, the fault which it will not find, even with the ends of sewer and vent-shaft open, must be very minute.

All new drains ought to be submitted to the water-test, and in this work, and this only, ought it to be used, unless by the express desire of an owner anxious to have drains of the very best kind. It is customary, however, because of obvious difficulties, only to apply the test to that part of the drain which is within the private premises, and not to that in the public street between the premises and the sewer. The usual method is to plug the drain from the interior of the access-chamber, stop all the lower gullies, and fill the drain to the top of the highest one. Bach drain entering the chamber is similarly treated. It is well to allow the water just to appear in the lower gullies before stopping them, otherwise the branch-drains, to which they are attached, will be air-locked. This of course, will not matter if the drain is sound, but if it is not, time may be lost in waiting until the air escapes from, and the water reaches, the defective point

The plugs used in stopping drains, before the application of the water-test, have been described and illustrated in Section VIII. Vol. I., and nothing need here be added on the subject It may. however, be mentioned that gullies may be stopped at the neck by passing into them a strongly-made india-rnbber bladder with tube attached, through which the bladder may be inflated with a small force-pump, stub as is used for inflating bicycle tyres. When all is ready, the drain is tilled with water.

The entire drain must be left exposed, until this test has proved it to be sound, but each pipe should be supported on concrete in the middle of its length to prevent its being disturbed if trodden upon, and during the filling of the trench, as a sound drain may be injured during the latter operation if this precaution be neglected. After the test, the remaining portion of the underside of the drain - beneath the sockets - should likewise be concreted. The expense of this concrete is very small, and. in addition to holding the tube steady, it ensures proper support at the part most needing it. It is of little consequence how tin-upper part of a drain is covered compared with its nether side; when that is solid the arch formed by the upper side of the pipe is well supported to bear the weight of the superiucumbent earth.

Concrete is not applied to the exterior of a drain to make, as it were doubly sure of the soundness of the joints, but to ensure the stability of a drain perfect in itself. If the joints are not sound, concrete cannot be depended upon to make them so; and if they are, they require no additional perfection from that material. The drain may, of course, be entirely covered with concrete after the water-test, if so desired, but that on the upper part is of far less use than that underneath, adds somewhat to the expense, and may be an inconvenience when alterations to the drain are required.

It will be observed that the access-chamber itself was not included in the foregoing test, but as it forms a part of the drain, and will probably be filled with sewage if the interceptor-trap becomes choked, there is quite as much reason why it should be tried as severely as the drain itself. But where, as is often the case, the top of it is below that of the highest gully, it could only be included in a second test. Again, in order to embrace it in the trial, the drain must be plugged on the sewer side of it, for which purpose three or four pipes would have to be left out of the drain between it and the sewer.1 The drain may be plugged from within the sewer if the latter is large enough to allow of a man passing up it, in which case the access-chamber could be tested together with the street portion of the drain. Because of these difficulties, therefore, although the chamber ought to and can be tested, it is usual to insist and rely upon perfection of workmanship.

Access-chambers as usually made, - i.e. with open channels,- are liable to be filthy if not made in the best form, whilst the retention of a large body inert air - which they make unavoidable - is contrary to all the canons of successful drain-ventilation. Again, their walls are liable to be not water-tight, in which case, if a chokage takes place in a trap beyond or within them, they become leaky cesspools, hiding, because they leak, for a long time, the fact of the obstruction having taken place. In the meantime, the ground outside them becomes saturated with sewage, which no subsequent clearanee of the drain will remove. It is better therefore for the drain, in its passage through the access-chamber, to be inclosed or continuous as an air-tight tube and fit with screw-down removable covers. An access-chamber should be intended merely for access, and not be liable, through the choking of an intercepting trap, to become part of a drain, - in fact a cesspool. Even at the best an access-chamber (because of its great capacity) is incapable of perfect ventilation through a 4-inch pipe. Where drains in small systems are properly made, and therefore unchokable with ordinary use, access-chambers are unnecessary.

1 The trap itself may be plugged from the chamber, but the plug mart be of such a form as to admit of being released from the surface of the ground, when the chamber is full of water. - Ed.

Some people appear to think that the word "ventilation" in connection with drains means relief from pressure, and that pressure of a distinct kind exists in well-ventilated sewers, and therefore in the drains connected with them. We often find a shaft on the soil-pipe relied upon for such "ventilation", no intercepting trap or air-inlet being provided. Such shafts are sewer-ventilators only, and bring increased danger to the occupiers of houses possessing them. The general principles and methods of drain-ventilation have been described an<l illustrated in Section VIII., Vol. I., and only a few remarks on this important subject need now be made. The motion of air in drains is very sluggish, and all bends tend to make ventilation more difficult The inspector will therefore discountenance the use of all unnecessary turns, especially in the outlet-shafts, advising that all needful ones be as long and slow as possible, prohibiting those at right angles, and utterly con-demning such arrangements and positions as involve two or three such angles following each other. The ridge of the roof is a good position for the outlet shaft to terminate, but when this can only be reached with bends like those shown in Fig. 674, ventilation is nearly impossible, for the bends constitute an air-trap. If there are skylights or dormer-windows on a roof near the vent-shaft, this must be continued above them, but in other rases the vent of a well-constructed drain may safely terminate about a foot above the eaves.

Drains 60010Drains 60011