Whether this plan is practicable or not must be left to elevator manufacturers, but it seems to me that with the Hale-Otis elevator for instance (which is conceded to be one of the best) it could easily be accomplished. Certainly some such arrangement would effect a great saving of water, and perhaps bring water bills to a point that this class of consumers could afford to pay.

Hydraulic elevators where the water is used over and over again, by being pumped from the discharge to elevated tanks, cut little or no figure in connection with a city's water supply. When fuel, first cost, attendance of an engineer, and the poor economy of the class of pumps usually employed to perform this work are considered, the cost of operating such elevators is greatly in excess of what it would be if power were supplied direct from water mains, at any reasonable rate. The following remarks will then relate almost exclusively to that class of hydraulic elevators supplied with power directly from the water mains.

Let us now consider whether they are a desirable source of revenue, and in this my knowledge does not exceed my actual experience. Few elevator users appreciate the great quantity of water their elevators consume. Even in Kansas City, where, on account of the high pressure carried, much smaller cylinders than ordinarily are required, it is found that passenger elevators frequently consume 500,000 to 800,000 gallons of water per month, which will make a very considerable bill, at the most liberal rates. I have, therefore, concluded that the quantity of water was so large that, unless liberal concessions were made, it would be a hardship to consumers to pay their water bills, and have therefore made a special schedule, according to quantity, for elevators and motors, these rates standing below our regular meter rates, and running to the lowest point at which we think we can afford to furnish the water. This schedule brings the rate below what we would receive for almost any other legitimate use of water; and, in view of our rapidly increasing consumption, and the probability of soon having to increase all our facilities, it is an open question whether this will continue a desirable source of revenue.

In Kansas City we have elevators of various manufacture: the Hale-Otis, Ready, Smith & Beggs, O'Keefe, Kennedy, and perhaps others, each having its peculiarities, but alike demanding large openings in the mains for supply. These large openings are objectionable features with any waterworks, and especially so with direct pumping. An occurrence from this cause, about two years ago, is an experience I should not like repeated, but is one that might occur whenever the pressure in the mains is depended upon to throw fire streams. In this instance a large block of buildings occupied by jobbing houses and having three elevators was burned down, and the elevator connections broken early in the fire, allowing the water to pour into the cellars in the volume of about twelve ordinary fire streams. This immense quantity of water had to be supplied from a 6-inch main, fed from only one end, which left little pressure available for fighting the fire, and as a matter of course failure to subdue the fire promptly was attributed to the water-works. We have since had up hill work to restore confidence as to our ability to throw fire streams, although we have demonstrated the fact hundreds of times since.

From this time we have been gradually cutting down on the size of openings for elevator supply, but under protest of the elevator agents, who have always claimed that they should be allowed at least a 4-inch opening in the mains, until we have found that under 80 to 90 pounds pressure two to four 1-inch taps will answer the purpose, provided the water pipes are of ample size.

The "water hammer" produced by the quick acting valves of elevators has always been objectionable, both in its effect at the pumping-house and upon water mains and connections. To obviate this, Engineer G. W. Pearson has suggested the use of very large air chambers on the elevator supply, and still smaller openings in the mains, his theory being that the air chambers would not only materially decrease the concussion or "water hammer," but that they would also act as accumulators of power (or water under pressure) to be drawn from at each trip of the elevator, and replaced when it was at rest. This plan I have never seen put to actual test, but believe it to be entirely practicable, and that we will have to ultimately adopt it.

All things considered, the plan of operating elevators from tanks in the top of buildings, supplied by a small pipe connected with the water-mains and arranged with a float valve to keep the tank filled, I believe to be the best manner of supply, except for the great additional cost of putting up such apparatus. By this arrangement the amount of water consumed is no less, in fact it would ordinarily be more than with a direct connection with the mains, but it has the advantage of taking the water in the least objectionable manner. Still, if this mode of supply were generally enforced, the large first cost, an additional expense of operating, would undoubtedly deter many from using elevators.

Another evil in connection with the use of elevators, and which no doubt is common, is the habit many parties have of keeping a key or wrench to turn on and off the water at the curb. This we have sought to remedy by embracing in our plumbers' rules the following: "All elevator connections in addition to the curb stop for the use of the Water Company must be provided with another valve where the pipe first enters the building for the use of occupants of the building." Without this extra valve it was found almost impossible to keep parties from using the curb valve. In most cases the persons were perfectly responsible, and as there was no intent to defraud the company by the act, they would claim this privilege as a precaution against the pipes bursting or freezing. This practice was very generally carried on, and was the direct cause in at least two cases of very serious damage. In the instances referred to, the pipes burst between the elevator and the area wall of buildings, and the valves outside had become so worn from frequent use that they would not operate, allowing the water to literally deluge the basements before the water main could be turned off.