carry the heated air through a separate system of ducts and flues to the rooms through specially constructed outlets placed directly beneath the radiators.

Thermostatic control is used to regulate the draft and check dampers in the boiler, the reheater coils and the fresh air supply. Additional dampers are installed for hand operation. In addition, the water supply for the humidity pan is automatically regulated by a float valve, the overflow draining into a small well which is used for no other purpose.

It will thus be seen that the system is entirely automatic throughout, at the same time being subject to a wide range of adjustment. For instance, the outdoor air supply can be entirely shut off in winter or this source of air supply can be used exclusively as in summer.

In further description of the operation of the gravity separator, it will be noted that the inlet pipe carrying the vitiated air to the first chamber leads downwardly in the chamber to a point below the separator box or chamber, while the outlet leading to the second box is at the top, the idea being that the

current of air will pass downward and then up again over the side wall of the separator box which is, in effect, a baffle plate or wall.

In this change of direction, the heavy noxious gases will gravitate with the cooling of the air to the bottom of the box and pass down and out through the opening in the bottom to the branch conduit leading to the chimney. The desired downward movement of this air current is assisted by an inclined or curved deflector wall arranged beneath the point where the vitiated air first enters the box. The desired suction for carrying off the gases to the chimney is secured through the chimney draft.

The idea of admitting the freshlywarmed air directly beneath the radiators is partly to avoid drafts due to the entrance of the air into the rooms. The outlets are designed to distribute the air as much as possible from the center towards the ends of the radiators, causing it to pass upward in approximately even quantities from end to end.

The system as installed is operated in connection with a hot water heating

plant. A Progress boiler of 1,300 sq. ft. capacity, made by the Thatcher Furnace Co., New York, is installed, supplying, in addition to the indirect heating stack, twelve radiators having a total of 632 sq. ft. of surface. The indirect heater contains 400 sq. ft. of surface and is divided into three stacks, each separately valved, affording further regulation of the system.

While no figures are available just yet regarding the capacity and operation of the "air separator" its limits of adjustment are so wide that its serviceability is assured in meeting any ordinary requirement. The system as installed has been in use since early Spring and has given notable satisfaction, especially in maintaining the freshness of the air.

Mr. Cluthe is enthusiastic over the operation of his system. "It has already given great satisfaction," he said, "and it undoubtedly operates as intended, receiving the lower stratum of air from each room, separating the foul air from the fresh air by the law of gravity and sending the foul air underground to the chimney. I consider it the nearest approach to the human lung and I can easily set the apparatus to take care of 20 people or 200.

END VIEW OF FIRST CHAMBER OF "AIR SEPARATOR."

Exhaust Air Enters at Left and Settles at Bottom of Chamber.

SECOND CHAMBER OF "AIR SEPARATOR" CONTAINING REHEATING STACK.

"This summer I have installed a 16-in. fan in the heating stack and operating it at the second speed, commencing at 9.30 A. M., I pulled down the temperature 3 degrees and later maintained an inside. temperature of 74° F. when the outside temperature was 87° F. Of course all doors and windows were closed."

The economy to be obtained from such a system depends, of course, upon the amount of air recirculated. In the present installation two-thirds of the air is recirculated, the other one-third being admitted through the fresh air inlet.

In point of cost, the system may be compared with that of an ordinary indirect system comprising air ducts, but in this case the radiator is concentrated in the center of the building and the extra cost of the "air separator," when commercially manufactured, will not be excessive.

Mr. Cluthe has applied for a patent on his design.

The architect for this residence is W. F. Staab, of Glen Ridge, N. J., who worked out the details of this unique installation. The heating work was done by Frank Larkin of Kaveny & Larkin, Montclair. Wm. Wickstrom, of Montclair, was the builder.

tween the inside and the outside of the building will be somewhat lower. In the third place, if the heat is not turned off, the temperature is liable to rise, causing a greater heat loss during the night and a general opening of windows and doors in the morning in order to lower the temperature to a comfortable point.

VALUE OF RECORDING THERMOMETERS.

In studying the operation of heating systems, the installation of recording thermometers has been found very useful; and our recommendations for the operation of particular systems have been largely based upon the knowledge which we have obtained from a study of recording thermometer charts. Such a record is also very valuable for demonstrating to the customer the principles of operation recommended. If the building is large, the proper location of the thermometer is not always a simple matter. It should be placed where it will show the average condition, and some experimenting is usually required to determine

this.

The Rochester Railway & Light Company has been able to effect very striking economics in the operation of the heating systems of two of its large steam customers, and it is my purpose to give some details regarding them.

ECONOMIES EFFECTED IN AUTOMOBILE FACTORY.

The first case is that of a factory used for the manufacture of automobiles. This

is a well constructed brick building, four stories in height, and having a total floor area of approximately 60,000 sq. ft. All the heating is by direct radiation. The installation of a recording thermometer showed that the steam was left on the building all the time, and there was a decided rise in temperature during the .night hours. Upon our recommendation care was taken to see that at the end of the working day all windows were closed and the steam entirely shut off, turning the steam on in the morning in time to make the factory comfortably warm at seven o'clock.

Fig. 1 shows a typical week's record of the recording thermometer, taken after our recommendations had been put into effect. The "sawtooth" character of the chart is very marked. Note how the

temperature falls off gradually during the night hours and on Sunday, and how readily it responds when the steam is turned on about six o'clock in the morning. The temperatures of 70° F. and over shown during the afternoons were unnecessarily high for a machine shop. This is accounted for by the fact that the thermometer is placed in the center of the room, and the temperatures near the windows are without doubt somewhat lower than recorded. In general, the present operation of the system is satisfactory. The temperature record is for the week. commencing Saturday, January 18, 1913, and the outside temperatures as recorded by weather bureau, are as follows:

Outside temperatures. Deg. F.
Maximum. Minimum. Mean.

In order to show the effect on the total steam used for heating, I have calculated the average pounds of steam per day used during a number of months, and have plotted these against the average outside temperature of the same months. The results are shown in Fig. 2. The upper curve is for the operation before our recommendations were put into effect. The lower curve shows the

present method of operation, and covers a period of over two years. The saving averages 37%.

RESULTS OBTAINED IN GROUP OF FACTORY BUILDINGS.

The second case is a large group of factory buildings with a ground area of approximately 75,000 sq. ft. All the buildings are well constructed brick buildings, varying in height from five to seven stories. The heating is mainly by the fan system, and there is provision made for recirculating the air in only one of the buildings, which forms approximately one-quarter of the total heating load. Our preliminary investigation showed that the steam was left on the heating coils all the time, that the cold air openings were not closed at night, and that, if the weather was at all severe, the

FIG. 3. TEMPERATURE RECORD IN GROUP OF FACTORY BUILDINGS CONTINUOUSLY HEATED.

fans were run continuously. The installation of recording thermometers in several places showed that the night temperatures often ran up to 80° F. and over. Several tests were run on the operation of the system during the night hours, and the following facts were developed.

With the cold air openings closed and the steam entirely shut off the heating system, the temperature in the buildings dropped at a rate of approximately 1° per hour in severe weather.

When the steam was turned on the coils but the fan not started, the natural circulation would hold temperature approximately constant. When the fans were started up, the temperature of the buildings responded rather slowly.

COMPARISON OF TWO CASES.

The conditions of this establishment are entirely different from those in the first one described. In the first case the work requires considerable muscular exertion, and if the temperature is a little low in the morning, no particular harm is done. In the second case, a great deal of the work is light bench work, a great many girls are employed in certain departments, and some of the work in progress would be ruined if the temperature were allowed to get too low. It was the opinion of the management that the temperature should not be allowed to get much below 65° F. at any time, and that 70° was the desirable temperature during the working hours.

FIG. 4.-TEMPERATURE RECORD IN SAME GROUP OF FACTORY BUILDINGS WHEN INTERMITTENTLY HEATED.

SCHEDULE RECOMMENDED FOR OPERATION OF HEATING SYSTEMS.

Based on the data given above, we recommend to the management the following schedule for the operation of the heating system, which schedule was put in force. As we did not feel that we could afford to take any chances, we "played safe" in making up the schedule, with the idea of making modifications after more experience had been gained. General Instructions.

5:30 P. M. Close cold air intake. Stop all fans.

Turn off steam from all fan coils.