To test the effect on traps below of emptying the tank after the manner of a flush tank, a 4-inch outlet valve and waste pipe were fitted up in the manner shown.

Outlets were left on each story below the water-closet for testing the traps at various heights on the stack. The soil pipe was ventilated at the top full size, and had the usual foot vent. Back pressure was generated by the bend just below the basement floor.

In order to permit also of a series of experiments on evaporation a 4-inch galvanized iron flue was erected by the side of the soil pipe. This flue terminated just below the first floor in a galvanized iron lantern, with a glass door on its front side. A 1-inch rubber tube was connected with the bottom of the lantern, and an anemometer was placed above the point of connection in an enlargement made to receive it. The anemometer was so arranged and placed that it could measure accurately the current of air passing through the rubber tube in either direction. The galvanized iron flue could be tested either cold or heated by gas-jets, as shown in the drawing. A second lantern was placed on the third floor with a similar appliance for heating the flue.

A 1-inch lead waste-pipe was connected with the soil-pipe just above the basement floor. This branch waste had a number of ventilating openings made upon it, and a deep seal S trap at its end. The trap had three ventilating openings in its outlet arm, one at the crown and the others below the crown, as shown. All the vent openings both on the trap and on the branch waste were provided with small connecting tubes, so arranged that the rubber ventilating flue could be readily attached to either. The openings were, furthermore, all provided with closely fitting corks so that they could be hermetically sealed.

By this arrangement the effect of ventilation at different points of the trap or its waste-pipe upon its water seal could be accurately tested. Further tests in evaporation were made by connecting a second branch waste below the first with a brick flue heated by a stove.

In order to make an accurate record of these experiments the diagrams shown in Fig. 244 were made. In these the trap seal is represented by a vertical line between two circles. The upper circle represents the outlet arm of the trap in section, and the lower circle the inlet arm. The horizontal lines show the level of the seal after each discharge.

The small diagram (Fig. 243) illustrates a simpler form of apparatus upon which I made a large number of experiments on siphonage. I assumed that as severe a test for siphonage to which a trap could be subjected in practice would be that which would be sufficient to siphon out an 8-inch pot trap or a ventilated S-trap constructed in the usual manner. Such a test may be made by connecting the trap with a 2-inch waste-pipe from a large bath tub, emptied through an outlet large enough to fill the waste-pipe full-bore, the waste plug being successively raised and lowered a number of times while the water is escaping. The siphoning action produced on a 1-inch branch connected with such a waste at a point six feet below the tub is sufficient to destroy in one second the seal of a 1-inch S-trap of the ordinary construction, having a vent opening at the crown, of the same size with the base of the trap (1 inches), and connected with a 1-inch ventilating pipe of smooth, new lead, sixteen feet long. It will also destroy the seal of an ordinary S-trap having a vent opening at the crown inchin diameter, without any vent-pipe attached thereto, and will siphon out a pot trap 8 inches in diameter having a seal four inches deep.

Fig. 243. Simple Apparatus for Trap Testing.

Fig. 243. Simple Apparatus for Trap Testing.

Fig. 244. Diagram Showing Test for Siphonage.

Fig. 244. Diagram Showing Test for Siphonage.

Trap Testing Apparatus.

The tank in this little apparatus, which in principle resembles ours here, had a capacity of 100 gallons. The waste-pipe was 2 inches in inside diameter like ours, but only 6 feet long to the trap branch, while ours is 17 feet long to the testing branches.

The outlet plug, like ours, was large enough to fill the waste-pipe full-bore.

The next figures (245 to 249) give the sections of some of the traps tested, the horizontal lines corresponding with those in the diagrams. Each test was repeated a number of times, the results being each time almost absolutely identical. A single discharge of 15 gallons destroyed the seal of a 1-inch S-trap vented with a 1-inch pipe 25 feet long, attached at the crown. With this vent-pipe shortened to 15 feet two discharges of 15 gallons each broke the seal. Shortened to 9 feet 7 discharges broke the seal.

In the pot traps tests 15 gallons were used at each discharge. They all lost their seals, as shown, except the 8-inch pot, tested in the later experiments for the Board of Health, to be described in another chapter, which lost all but a quarter of an inch of its 3-inch deep seal after the tank had been emptied 16 times. The 6-inch pot required four and the 4-inch pot two tanks full to break their seal.

On the Board of Health apparatus, also several other traps were tested at the same time with the pot traps, but as only two were able to preserve their seals against the tests applied, and as most of them had already been tested in the experiments made for the National Board of Health and their tests published, by which their power of resistance as compared with that of a ventilated S-trap and to a pot trap, was made known, it was thought unnecessary to record the failures again in our tests.

Fig. 245. S Trap.

Fig. 245. S Trap.