This was a water-closet in a tight, unventilated compartment in a pri vate house. Odors were often present which no one could account for. The job was new and first-class. The house was well built— too well. After many others had failed to diagnose the trouble, a plumber with some philosophy in his make-up examined the job. He stood in the hall, and slammed the closet-room door. It failed to latch, the room being so tight that the air-pressure kept it from seating on the rabbet of the frame. The door, of course, was instantly thrown partly open again by expansion of the air, and the plumber caught a glimpse of the water in the closet-bowl bobbing up and down. By repeating the experiment and measuring the depth of water between times, he discovered that, as suspected, the sudden closing of the door of the small, tight room was thrusting the water down in the bowl and causing enough to flow over into the soil pipe to break the seal. The trouble was remedied by cutting inch off the door at the bottom.
Evaporation has been described elsewhere. Leakage of seals has been mentioned in conjunction with types of fixture taps. Siphon age of traps is simple. The conditions necessary to start a common siphon being established in a waste pipe, the seal will be drawn out. The discharge of water from a fixture will siphon its trap (self-siphonage), if no provision against siphonage is made. The crown vent pipe, as described, breaks the siphon in a trap when its fixture is discharging, and prevents other fixtures from siphoning or waving out the seal. Capillary loss of seal occurs through hair, lint,
and strings hanging over the weir of the trap. Dipping into the seal on one side, and ending in the pipe on the other, water will, climb through or between such matter by capillary force, and will drip by gravity into the pipe. This is indicated by the tangled lines at R, Fig. 210, represent ing capillary material banging over the outlet neck D of the trap. The trap indicated is for a lavatory with horn overflow bowl, V being the overflow connection, I the waste, B the crown vent, and 0 the outlet. Traps are some times locally vented at V.
Materials forming a porous coating on the inner walls of the trap through chemical action or otherwise, are now and then responsible for the loss of water-seal by action of a capillary nature. The shape of a trap may favor the accumulation of matter that will lead to capillary loss of seal. This is one reason why the plain, open-wall, cylindrical-bore traps are best. It is found that no matter how the trap is shaped, its surface is, as a rule, not used except at the points which conform to the simplest, most direct course—as before said. Other shapes, then, present needless fouling surface and space for accumulation of matter that interferes with the proper service of the trap. De parture from the shape mentioned is necessary to secure an unvented trap that cannot be siphoned. Any trap that must necessarily be connected so as to put the air of the sewer side against the gasket of the clean-out cap, should not be used.