At each return bend of the pipe the wave created by the siphonage is partly reflected back, and broken up, and the air thus finds an opportunity to force its way through to the outlet without appreciable reduction of the water level in the reservoir chamber.
Tested on the apparatus shown in Fig. 241 this trap was found able to withstand indefinitely the most powerful siphonage which could be applied, a strain which in a single discharge, destroyed the seal of a fully vented S trap, Fig. 143, even though the vent pipe was new and smooth, and of the full size of the bore of the trap and only ten feet in length!
S-trap emptied by one siphoning action, leaving only a few drops in the bottom of the trap, at the level of the dotted line. A vent pipe 10 feet long was attached to the vent outlet shown.
The same strain easily emptied an unvented 4-inch pot trap, Fig. 144, and siphoned out an S trap having a seal six feet deep! Fig. 145.
Pot trap, which lost its seal after three siphoning actions on our apparatus, shown in Fig. 221.
Yet our new trap easily withstood this enormous pressure ten times repeated, losing only 3/8 inch in the first two applications of the strain, and nothing more in the remaining applications. In another test it lost ½ inch in the first three applications and no more thereafter.
The trap will hold its seal more securely than a pot or cesspool trap of more than nine inches in diameter, whether the pot be vented or unvented (Fig. 146), because in either case the movement of the water at the top of the pot is very sluggish and this favors early clogging with greasy scum, gradually converting it into an S trap.
Fig. 145. S-trap, with seal 6 feet deep, which lost its seal, tested on the same apparatus.
Fig. 146. Pot trap nine inches in diameter.
Hence we have, here obtained an absolutely antisiphon self-cleaning trap of practical form.
But there are yet two serious objections to this form. The first being its great size, and the second the resistance which the numerous return bends make to the outflow of the water in its normal use. Small eddies are generated at each bend, which retard the quick and free escape of the waste water without in any way increasing the scouring action thereof.
Tested for friction it was found that it required 35 seconds for the water of a 12-gallon tub to escape through this trap as against 21 seconds for the same amount of water under the same conditions passing through a trap of this same size but constructed as shown in Fig. 138 without the partitions.
In order to obviate the first objection of the inconveniently large size of the trap, we next experimented with smaller sizes, successively reducing the horizontal dimensions from 13 to 10½ inches, and then to 9, 8, 7, and finally 6 inches. Moreover the shape of the partitions was varied in order to ascertain the most effective and easily constructed arrangement.
Figs. 147 to 161, inclusive, show the various forms we have examined, and in Table I the test made on some of these forms are recorded. The results of the tests may be briefly summed up as follows:
Seal Retaining Traps, (i) With traps of the kind under consideration the power of resistance to siphonage is in proportion to the horizontal length of the waterway in the trap. The largest is capable of resisting the most powerful siphonage that can possibly be brought against it on any apparatus which can be built for making a plumbing test of which we are aware,