Referring again to Fig. 147, air takes the course offering the least friction; and F branching out of and into E, which is the same size pipe as shown at X and Y, the greater part of a current of air pass ing upward through them will travel by pipe E. For this and other reasons it is best to take the branch pipe F for the soil pipe. Then, whatever offset may be necessary to reach the closet openings will be washed; and the straight, vertical stack left for the vent affords no chance for the lodgment of rust or other obstruction. When water is discharged into the soil pipe at G, pipe V protects the closet trap from siphonage; and the tendency to form a vacuum above the water in the soil pipe by the piston action of the discharge water, is neutral ized by a proportional draught of air from vent pipe E through branch Y. The air in the vent pipe between Y and B tends to continue its course to the roof, while that below the branch Y is traveling toward branch Y. A partial vacuum formed in soil pipe F by a discharge from a fixture, will be checked by a supply of air drawn from vent pipe E between branches X and Y. The vacuum formed behind the discharge water in soil pipe F increases the upward velocity of air in vent pipe E below Y; and the air pushed down in front of the discharge attempts to reverse the current below X. The increased velocity of the air in pipe E demands more air than was passing through it by natural draught. This demand is supplied by the extra volume which the water is pushing before it.
As long as the discharge water is above branch X, the air simply revolves in the two pipes which form the loop. The air in pipe F travels downward before the water, and up through pipe E and branch Y, and down pipe F behind the water. This revolution of air in the loop continues until the water reaches the junction X of pipes E and F, without causing any perceptible "puff" at the fresh-air inlet opening.
When both the connections are in vertical lines as in Fig. 147, after the water passes X, it will probably reverse the current of air in the fresh-air pipe in some instances; but, were it possible to shove out every atom of air in the soil pipe between the trap and point X, there still would not be a particle of foul air puffed out at the fresh-air opening, if the fresh-air pipe is of greater length than the distance between X and the trap.
The object in not connecting the loop stacks as close together as fittings will permit, is to keep the water, as it turns into the hori zontal main, from interfering with the entry of air to the vent. By giving some distance to travel before reaching the loop connection, the discharge of water will be well spread in the main line before passing it. From this point on, it may cause violent eddying of the air in the main, but no actual reversal of the current will take place.
The force of air in front of water in down spouts that connect inside of the intercepting trap, may at times reverse the air in the fresh-air inlet proper. The loop pipe is an aid in this respect, too, as more air is at hand to cushion the rush of a sudden downpour; and the various fixture trap seals are, if affected at all, left much more stable. It would, if necessary, be
better to have soil-pipe air expelled from an inlet, at times, by the action of storm water, than to incur the risk of siphonage or waving-out of fixture trap seals for lack of it.
No pipe of any building should open to the air with less than a 4-inch end. Small pipes should be increased to 4 inches before passing through the roof, as shown in Fig. 148. Pipe 4-inch and larger, up to 6-inch, should be increased to 6-inch. The object in all cases being to prevent closure by hoar frost. With 6-inch and larger pipe, it is doubtful if it is ever xecessary to increase the size at the roof, excepting in buildings with cold roof space, no matter how high the building may be; yet some city ordinances call for an increase of one size regardless of size, which is manifestly foolish, as it permits ing 2-inch to 21 or 3-inch on any type of job, and this is known to be inadequate in any but southerly latitudes. The velocity of air up the line is, of course, higher in tall buildings than in low ones; hence, in them, more moisture is carried through any given opening, and the theory of increasing large pipe at the exit is based on the tion that smaller openings would, as a result of this excess of moisture, be closed by frost. The great amount of warmth over large buildings must often, however, be considered as reducing the chances of closure by hoar frost. In tropical climates, no increase of any size is necessary. In southerly temperate latitudes, no special attention is given precau tions against hoar frost, beyond in creasing the size of small vents to at least 4 inches in diameter.
Flashings. There are patent devices for flashing around pipes, usually made of copper; but the plumber will do well to command the skill necessary to manipulate sheet lead to suit conditions as he finds them. In any location where warm air will always be seeking an outlet from the attic through chance openings, the sleeve of the ing may be made two to four inches larger than the outside diameter of the vent, and capped with an annular V-ring of lead in the manner shown in Fig. 149. The cap ring need only be tacked to the sleeve with solder. The top edge of the sleeve should be notched or some other provision for air-exit made, so as to insure constant changing of the air in the sleeve. If, on account of braces or projections necessary to hold the pipe rigid where it passes through the sheeting, it is inconvenient to let the sleeve extend below the sheeting as shown in the engraving, it may terminate at the roof line. If the building is a storage warehouse, or for any reason the attic will not be very warm, or conditions are in favor of cold air being drawn in through chance ings in winter, then the method of flashing and packing the sleeve with felt or mineral wool as shown in Fig. 150 should be employed. In all cases the vent and flashing must rise above the possible snow-level for flat roofs. The snow-level on a steep roof will be less, but drifts may obstruct the vent if left at the snow-level. Some latitude for settling of the roof under the weight of snow and ice, and for expan sion of lines supported by brick piers or other supports far below the roof-level, must be allowed in fitting flashings. If they are too closely drawn or capped, trouble will soon follow.