If the indiarubber pipe to the detector or pressure gauge is placed in either of the traps marked 1, and the glass tube filled with liquid up to the data line, 5 pints of water poured into either of the traps marked 1, will produce a rise of 1 inch in the liquid of the detector, that is if all the drains are clear and joints tight, the drains being stopped off for testing at A. Should a trap be fixed anywhere between A and B a lesser quantity will be required to lift the liquid, and the position of the trap can be determined by comparing the exact quantity of water used with the capacity or quantity of gas in the drain. If a trap should be fixed or a stoppage formed in any part of the drain A B, the flushing of a closet or sink would, by the compression of the gas, force it in bulk through the weakest trap, or the one having the least dip or seal. The quantity which would pass through would depend on the amount of water used in the flushing and the fall of the drain. The drains to the building having been tested, and their defects ascertained, it will be necessary now to test the soil-pipe. On this plan it is fixed on the outside of the house, having a trap with an open grating just beyond the basement closet, and a ventilating pipe carried above the eaves of the roof. Whether the soil-pipe be fixed inside or outside of the building it should be perfectly gas-tight, and in this testing a person cannot be too particular. In testing the soil-pipe shown on plan, the easiest method is to put the detector or pressure-gauge at the grating of the trap 3, placing the indiarubber tube over the grating, and making a tight joint with clay. Then close the top of the ventilating pipe and pour water in the top closet, when, if the joints are tight, the liquid in the detector will rise suddenly, and then lower itself as the water leaves the trap, indicating that the soil-pipe is tight, but if it is not tight, no rising of the liquid will take place. Should there be no trap at the bottom of the soil-pipe, it will be necessary to excavate down to the drain to take out a length of pipe, to seal the mouth of the drain with clay for testing. If there should be leaky joints or holes in the soil-pipe, a little sulphur burnt in the pipe, or a little pungent essence thrown into it, will clearly denote where the leaks are. Having tested the soil-pipe and proved it tight, or effectually stopped all leaks as the case may be, no gas can be given off in these drains or fittings except through the ventilators (under ordinary circumstances) as no trap has been siphoned in the testing. As before stated, the ventilating pipe runs to the top of the building of the same diameter as the soil-pipe, in fact this is a plan of drains to a house recently built in the suburbs of London, and the planning of them would be considered perfect by many sanitary men, but before we testify them as perfect, let us carefully analyse the working of the ventilation.