This requirement confines us to the use of some form of plain piping, either straight or bent, in the construction of the trap.

In the second place it is evident that a sufficient amount of this piping will be required in the formation of the trap to provide the necessary water capacity for resistance to back pressure and evaporation.

In the third place all the piping used must be on a horizontal plane in order to preserve the required minimum depth of water seal.

Finally, for the purpose of insuring against loss of seal by capillary action, the seal of the trap must be separated from its connection with the drain pipe by a distance great enough to offset the maximum of capillary forces ever encountered in plumbing practice.

The first form of trap answering to these requirements with which we experimented was, therefore, the simplest form, namely, that of a straight pipe placed horizontally, as shown in Fig. 139.

This trap consists of the seal proper shown on the left side of the figure, which is made of 1 -inch bent tubing, the seal being not over a half an inch in depth; a long horizontal body consisting of a plain round pipe likewise 1 inch in diameter, or of exactly the same sectional area as that of all parts of the seal tubing; and at the opposite end of this pipe the sewer connection piece, which is again of sectional area everywhere equal to that of all the parts of the trap.

I call the first part of this trap the "trap seal proper," the second, the "reservoir chamber," and the last, the "outlet connection."

The outlet connection has its overflow point inch above the bottom of the reservoir chamber, so that when this chamber stands full of water the entire depth of water seal measures only 1 inches.

Evolution Of A Permanent Anti Siphon Water Seal Tr 151

Fig. 139.

First step in the evolution of a permanent anti-siphon self-scouring shallow-seal trap.

The trap was constructed of round metal tubing 1 inches in diameter, the reservoir chamber being ten feet long.

The principle of resistance of this trap to siphonage lies in the air space over the long reservoir chamber and in the shallowness of the water seal. The water constituting this shallow seal, yielding readily to the siphoning action, is thrown out of the seal proper and distributed over the surface of the water in the long chamber, and only slightly raises the level thereof. A part of it is carried out into the waste pipe. Then air from the fixture side of the trap, having ample room above the water to pass through the chamber without disturbing the water below it, breaks the partial vacuum in the soil pipe, and restores the atmospheric equilibrium in the pipe system.

The small seal in the trap proper is then quickly replenished by water flowing back into it from the long reservoir chamber without materially reducing its level.

A small amount of water is driven out of the trap by each subsequent repetition of the siphoning action, but less and less is lost each time because the air space above the water is each time correspondingly enlarged, and the resistance to siphonage accordingly increased until a point is reached when no further reduction of its level by siphonage is possible.

The reason why air and not water escapes through the reservoir chamber, is because the water thrown up from the seal proper by the siphoning action forms a spray which, striking the top of the reservoir, adheres to it and, in virtue of the greater attractive and cohesive force already referred to of the particles of this fluid, permits the lighter air to pass through it and escape.

But there are two very evident objections to this simple arrangement of the parts of the trap: the first being its inconvenient and unwieldy form, and the fact that a very slight sagging of the body would be sufficient to destroy its action; and the second, that in the event of the siphoning action being exceedingly powerful, a water wave is set up in the tube body which acts like a solid piston in driving out before it the rest of the water therein, so that this form of trap could only be relied upon to resist siphoning action of moderate intensity.

Fig. 140. Wave formed by siphoning action.

Fig. 140. Wave formed by siphoning action.

Our next inquiry was, therefore, to discover some way by which this water wave or piston might be broken up, and the air behind it allowed to escape to the outlet without exhausting the reservoir in its passage.

The most natural method was to simply bend the pipe body back and forth on itself abruptly, and the most compact form possible in which our ten feet of tubing can be bent in this manner being that of a square, our next experimental trap took this form, as shown in Figs. 141 and 142.* zigzag waterway through the trap, the water having at all parts a sectional area exactly equal to that of the trap proper shown at the left of the figures, and of its inlet and outlet arms.

Our reservoir in this case consisted of a metal box 13 inches square on the inside, having a glass top and eight partitions set in such a manner as to produce a continuous

*These and all the subsequent drawings of our horizontal traps have been made to the same scale of one-eighth the actual size in order to facilitate comparison between them.

Evolution Of A Permanent Anti Siphon Water Seal Tr 153

Fig. 141.

Evolution Of A Permanent Anti Siphon Water Seal Tr 154

Fig. 142.

Second step. Breaking up the waves by abrupt ends. Plan and section of square trap.