The foregoing chapters have been confined to defects contaminating the air supply to the building or the air in the building. The present chapter will deal with the contamination of the supply of house water. Figure 10 represents an arrangement of the water pipes which has been frequently described and condemned in print, but of which many examples at present exist in New York City as the inheritance of ante health board days. The fact that the same blunder is being repeated daily in towns and villages away from the larger cities is sufficient ground for again calling attention to it. The water supply pipe A A has branches B, .C, N, etc. at various levels. B supplies a kitchen or butler's sink, where drinking-water is drawn for the tank. C supplies and connects directly with the bowl D of a pan closet or other form of direct flushing closet. A valve E on the pipe C is connected by levers with the seat handle F of the closet. The size of the water pipe A and the pressure of the water supply is such that when B is open water will not only not flow to the other branches C and N, but will draw the water out of A A and draw air into the branches C or N, if either of the cocks on those branches is open. If the closet G is in use and the handle F is raised at the time B is opened for drawing drinking-water, air is drawn directly from the bowl D of the closet and enters the water pipe A A if it does not mingle directly with the water being drawn through B. If there are germs of disease in the air from the bowl D they contami" nate the water which flows through A A. To avoid this possibility, small flush tanks are required for each closet or group of closets, so arranged that this suction of air into the water pipe is prevented.

Figure 11 was an arrangement for providing for the overflow from a water tank in a Madison Avenue dwelling. It was not uncommon some years ago and may still be found in new work outside of the larger cities. As the street pressure was at first insufficient, the tank E was placed on the top floor of the building and was used for supplying water to all fixtures except the kitchen and first floors. A pump in the cellar was used for filling it, and an overflow pipe B was connected with the tank and branched into the main soil pipe A of the house. The overflow B was trapped at C to prevent air from the soil pipe finding access to the tank water. The trap was provided with a back-air pipe D to protect it against syphon-age. As long as the tank overflowed at frequent intervals the trap C was supplied with water and remained secure. But improvements in the city water supply were made and increased the street pressure so that water would rise at night to the level of the tank. The pump in the cellar was done away with, and the tank was supplied directly from the street pressure, with a ball cock on the supply pipe, which shut off the flow of water when the tank was full. In connecting this ball cock the plumber had set it so that the water was shut off before the tank had filled to the level of the overflow B; in fact it could not have been otherwise arranged without continual waste of water. As a result the trap C very soon evaporated and the air from the soil pipe had free access to the surface of the water in the tank. The water was used, among other things, for bedroom basins, and a case of diphtheritic sore throat which had occurred in the house was thought to be due to the use of the water in brushing the teeth. The tank overflow was cut off from communication with the soil pipe and an independent pipe led to the extension roof and provided with a flap valve.

Figure 12 shows the contamination of the water supply which was discovered in an isolated country dwelling. The water was drawn from a cistern A supplied by the rain falling on the roof. Under the pump B, which stood on the back piazza close to the kitchen door, there was a sink or hopper C to catch the drip from the pump. The hopper was connected with some lengths of clay pipe leading to a trough and ultimately into a cesspool. It was found that the hopper, being convenient to the kitchen door, had been used to receive much of the liquid waste from the kitchen and from floor washings. The connection of the hopper and the clay pipe and the pipe joints were found defective, as was also the brickwork of the dome of the cistern. As a result much of the waste water poured into the hopper found its way into the cistern, which in time became so foul that it was very offensive and was abandoned. Before this was arrived at two cases of diphtheria had developed in the house.

Figure 13 shows a very common blunder, found in new work as well as in old. The hot-water boiler A in the kitchen is almost always provided with a " sediment pipe "BB, for blowing out the sediment which in time collects in the bottom of the boiler. This pipe is controlled by a stop-cock C, and to provide a means of disposing of the water when the boiler is being blown off, the sediment pipe B is connected with the waste pipe D from the kitchen sink F. This connection is frequently made below the trap E of the sink. The danger lies in the fact that when the house is closed for any length of time it is • very common for the plumber to turn off the water from the house and to empty all the pipes, as well as to draw off the water in the boiler through the sediment pipe B, and to leave the stop-cock C open. In this case air from the drains enters the boiler A through B B, and from the boiler connects with all of the water.pipes in the house. When the water is again turned on it receives any contamination which the drain air may have lodged in the pipes or boiler. Two cases of typhoid fever followed the occupancy of a house which had been left, in the condition noted for a few months. The sediment pipe should be arranged to have a free discharge into a trapped fixture, or to be emptied through a hose temporarily laid to a yard drain.