* "Lectures on the Science and Art of Sanitary Plumbing," London, 1882.
(7) Had he known the inevitable consequence of the law in introducing unintentional defects and by-passes, mistakes which have alone vastly more than offset any possible advantages back venting could provide.
Mr. Hellyer would have found several other very serious objections to trap venting which I have referred to in other places.
Mr. Hellyer's conclusions and recommendations resulting from these experiments were that every trap should be "back vented," and it is probable that this hasty conclusion and the wide circulation his publications enjoyed are largely responsible for the tremendous mistake of the trap vent law, a mistake which, regarded from the standpoint of pecuniary loss alone, has already cost the public hundreds of millions of dollars.
Mr. Hellyer would have found his vent pipes would have failed to protect the trap seals under any of the following conditions:
(1) Had the vent pipes been partially reduced in area by sediment deposit at the bottom or by frost at the top, or by rust anywhere in its length.
(2) Had the overflow pipe of the bath tub been plugged as is very frequently the case.
(3) Had the modern siphon jet or other closets having a strong and rapid flush been used in the tests.
(4) Had the vent pipe been very long and rough or contained an unusually large number of sharp bends.
(5) Had the experiments been prolonged sufficiently to try the effect of evaporation in the trap seals produced by the ventilating current.
(6) Had he investigated capillary action.
Fig. 271. Author's Pneumatic Trap Testing Apparatus.
It must be borne in mind that the value of our tests lies in showing not only the absolute but also the relative power of resistance of unventilated anti-siphon traps and the ventilated S trap. In making a comparison it would not be sufficient nor conclusive to show that unventilated anti-siphon traps are as efficient as the vented S under certain moderate strains. Their value can •only be fully demonstrated when it is shown that they are superior to the vented S, not only under moderate but under every possible condition that can be encountered, and even under strains greater than are met with in ordinary practice, and clearly the larger the range of tests the more conclusive the comparison. We should be justly open to criticism if, in making our tests we stopped anywhere short of a thoroughly comprehensive and exhaustive comparison of all forms of traps under every form of strain, from the mildest to the severest which could be applied.
We have found that a single discharge of about four gallons from our tank was able to break the seal of our 1¼-inch S trap with a 1¼ -inch clean new vent pipe only 6 feet long and having upon it a single return bend. And we found that the same trap could be siphoned out by two such discharges with a 6-foot vent pipe roughened on the interior surface as by rust even without any bend at all on the pipe. This test was made with the air pipe under the tank valve closed as by sediment or frost. Our air pipe takes the place of a ventilated trap under a bath tub, to which our tank corresponds. The overflow to the tank was closed as is sometimes done with the overflow pipe of bath tubs.
We then tested the vented S, with the air pipe open, and found that its seal could be broken by two discharges of four gallons each, the length of the trap vent pipe being 18 feet and having two return bends.
Now it is not necessary for us to show whether or not this strain was more severe than was ever possible in plumbing practice, because our chief purpose then was to compare the vented S with other traps unvented, and we found in our experiments that our 5-inch pot trap was able to withstand without loss of seal more than a dozen discharges made under precisely the same conditions as were applied to the vented S, including the very severest.
The superiority of the large sizes of our unvented pot traps over the vented S in the matter of resistance to siphonage was, therefore, absolutely demonstrated under conditions in which the whole apparatus was comparatively new.
Our contention is, then, that an unvented pot trap of large size is safer as a protection against siphonage than a vented S, and that the former will retain its power of resistance to siphonage longer than the vented S. In places where grease or other sediment is liable to collect in the scouring waterway of traps, it is still more liable to collect in the unscoured mouth of the vent pipe, and the pot trap cannot be clogged by frost, whereas the vent pipe can and often is so closed.
In every other respect the superiority of the unvented anti-siphon trap system over the vented S system is too evident to need further discussion.
Each state board of health should be equipped with some simple form of testing apparatus like that shown in Fig. 271 or like the pneumatic apparatus recently devised by the writer and shown in Figs. 272, 273 and 273a.
They may be made by any plumber or coppersmith without difficulty in a few days. A very considerable economy can be realized if several duplicates of the apparatus are made at the same time.
In all the demonstrations in siphonage of which I am aware up to the present time, the tests have been made by hydraulic apparatus, such as have been herein exemplified. But of late I have found that a pneumatic apparatus permits of a much greater accuracy of determination, especially in making comparative tests. By using a suction pump, as shown in Figs. 271, 272 and 273, we are able to reproduce at will any degree of rarification in the soil pipe desired to correspond with the varying conditions encountered in plumbing practice, as well as strains far beyond the usual ones for the purpose of comparing different traps and systems of plumbing with one another, and a vacuum gauge enables us to apply precisely the same degree of vacuum to every trap tested, so that results of great accuracy and scientific value are obtainable.