Floor Trap, with Tile Cover, set in Tile Work, and made tight by a large Elastic Gasket.

Table III Experiments on Water Scour 215

Fig. 205.

Floor Trap, shown with cover and Gasket removed.

Table III Experiments on Water Scour 216

Fig. 206.

Table III Experiments on Water Scour 217

Fig. 207.

Table III Experiments on Water Scour 218

Fig. 208.

Fig. 209a.

Fig. 209a.

Table III Experiments on Water Scour 220

Fig. 209.

Bath Tub, showing preferable arrangement of trap entirely above the floor.

pipe, and its outlet arm connects with the outlet of the water closet trap.

If now the seal of this closet trap is made deep enough, the shallow trap will protect it unfailingly from siphonage by supplying air through its seal to break the siphoning action.

This trap may be constructed under a considerable variety of forms, as shown in Figs. 210 to 224, to suit varying conditions, either the inlet or the outlet pipe passing through the centre of the reservoir or refilling chamber as desired. Or either arm may be placed out of the centre of the trap, as shown in previous drawings.

Table III Experiments on Water Scour 221Figs. 210 & 211.

Figs. 210 & 211.

Table III Experiments on Water Scour 223Figs. 212 & 213.

Figs. 212 & 213.

Table III Experiments on Water Scour 225Figs. 214 & 215.

Figs. 214 & 215.

Figs. 216 & 216a.

Figs. 216 & 216a.

Table III Experiments on Water Scour 228Table III Experiments on Water Scour 229

Fig. 217.

Figs. 218 & 219.

Figs. 218 & 219.

Figs. 220 & 221.

Figs. 220 & 221.

Table III Experiments on Water Scour 232Table III Experiments on Water Scour 233

Figs. 222, 223 & Figs. 225 to 227 represent the writer's earlier trap, the "Sanitas," which he developed from the pot trap as described in 1884, 5 and 6, in his little works entitled "Improved Plumbing Appliances" and "The Principles of House Drainage." This trap has been improved upon in the later studies herein described which developed the more scientific "Securitas" device, and in which were avoided the defects in the Sanitas of too great a vertical extension and too many abrupt and sharp turns. By doing away with these objectionable features the "Securitas" trap has attained a self-cleaning property equal to that of the simple S or ordinary siphon trap, for the bottom of the "Securitas" reservoir chamber can be curved upwards at the angle of junction with the small cross partition if desired. In practice, however, it is found better to leave this angle a little abrupt in order that small articles like rings or jewels, often accidentally finding their way into a trap, may not be swept into the sewer. The corner is too small to constitute an objectionable sediment pocket, but just large enough and conveniently enough located to safeguard small valuables without creating any corresponding objections. Being directly in the path of the strongest water flush ordinary sediment and greasy matters will not lodge there. This feature is sufficiently appreciated by users to justify its retention, although it might easily be done away with and all corners fully rounded if desired to complete the ideal round pipe section throughout.

Table III Experiments on Water Scour 234Table III Experiments on Water Scour 235

Fig. 225.

Table III Experiments on Water Scour 236

Fig. 226.

Table III Experiments on Water Scour 237

Fig. 227.

Figs. 226 and 227 show most clearly the objectionable feature referred to of the too great height of our first trap. In consequence of this unnecessary vertical extension, a larger proportion of its water seal is forced out under siphoning action than with the later device, which for this reason proves more self cleansing with even greater siphon-age resistance, and also has the very important advantage of forming a simple and perfectly effective back vent for a deep seal water closet trap, as already described.

Figs. 228, 229 and 230 show a few of the experimental traps made by the writer before the development of his Sani-tas trap, and some ineffectual efforts made in the wrong direction to take advantage of the superior specific gravity of water over air. by giving the two fluids a rotary movement within the body of the trap, and attempting to separate them from each other by centrifugal force in a vertical plane. In these early experiments the mistake was made of adhering to the perpendicular construction everywhere adopted at the time and even continued in the Sanitas trap. It was only on discovering that the strength and value of a trap in every way lay in its horizontal extension that success was finally attained. Had the trap shown in Figs. 228 to 230 been built horizontally rather than vertically the prob-lem would have been settled much sooner. It is true that the principle of the horizontal design was, to some extent, followed in the steps leading to the Sanitas trap, but it was only partially adhered to in the final form of the trap, and yet whatever of success it has had* I attribute chiefly to the horizontal element in its design.