space of this boiler an inch pipe, a, a, a, about 40 inches long was attached as shown in the figure. This pipe, led to an inch and a quarter tee b, to which were connected the several outlets used and the thermometer and steam gauge A. At B was a stop valve, and at Canother thermometer and steam gauge. All of the pipe a, a, up to the tee was heavily protected against loss of heat by asbestos paper, two inches of hair felt, and canvas. When the thermometer C showed 8 degrees of superheating, the loss of heat from the pipe would make thermometer A show two degrees of superheating, both steam ganges A and C showing exactly the same pressure. Thermometer B was graduated to degrees, a space of th of one inch being devoted to one degree. Thermometer C was equally coarse in its scale but the graduations were at intervals of two degrees. The steam gauges were graduated to pounds.

By raising the water in the boiler to the top of the gauge glass and increasing the quantity of steam generated by the boiler, the superheating at thermometer A could be made to vary from 2° Fahr. to zero, and the steam referred to herein as dry steam was such steam as flowed through the pipe a a when Thermometer A showed less than two but more than zero degrees of superheating, with reference to the pressure common to the gauges at A and C respectively.

EXPERIMENT 1.-The tee b was fitted at its under side with a draining pipe terminating in a petcock. Into its end one side were screwed -inch pipe plugs prepared as follows (see Fig. 60). The square hub to which a wrench is intended to apply was turned off, and a hole of an inch in diameter drilled through the center of the plug. A hole of an inch in diameter was then drilled in the inside end of the plug, so as to leave a thickness of metal at the outer end of one-sixteenth of an inch. The inch hole is then practically an orifice in a "thin plate," and removes the pos sibility of any of the heat of steam flowing through it, being employed in overcoming friction against the passages leading to the point where the steam issues into the atmosphere.

Thus arranged, dry steam, at 55 pounds gauge pressure, flows into the atmosphere of a boiler room in a jet which is perfectly transparent over about one-half an inch of distance from the orifice, as shown by Fig. 61. Fig. 62 shows similar jets for 95 pounds gauge pressure.

EXPERIMENT 2.-For the thin orifice in the end of teeb there was substituted a piece of -inch gas pipe d, Fig. 59, about four

Fig. 61. The jet issuing from the cock, f, is a bluish white color clear up to the orifice, due to the cooling loss of heat in passing through the six inches of 1-inch pipe and cock.

Upon passing water through the condenser so as to maintain it at an average temperature of 741° Fahr. the end jet became distinctly white, as per Fig. 64; the jet at the side certifying by its unchanged appearance that the steam operated upon was the same as for Fig. 61.

FIG. 63.-Dry Steam at 55 Pounds Pressure.-No Water in Cooler. The heat abstracted per minute was determined to be 12.75 British thermal units. The steam flowing* per minute was determined to be 1.13 pounds. The latent heat of steam at 55 pounds gauge pressure being 822 British thermal units, 1.13 pounds of steam would possess 929 units of latent heat, which if completely absorbed by refrigeration would cause the 1.13 pounds of steam to become liquid water at the temperature corresponding *This was determined by attaching Cock f to a Wheeler surface condenser, S, see Fig. 59, and determining the flow for a period of about half an hour, and weighing the condensed steam by a spring balance k. A sample record is the

following:

FIG. 64.-Steam at 55 Pounds Pressure Containing 1.4 per cent. Moisture. exhibits the appearance of a jet of steam at 55 pounds pressure containing 1.94 per cent. of liquid water.

A jet at 95 pounds gauge pressure being maintained at 76° Fahr., 26.27 thermal units were abstracted from it per minute, and the flow of steam was 1.75 pounds per minute. Fig. 66 exhibits the appearance of this jet, which by the above data contains 1.88 per cent. of water. Fig. 67 shows the same jet with no circulation of water through the cooling drum. It is slightly whiter than the similar jet of 55 pounds pressure, owing to the greater weight of steam flowing per unit of time. A similar difference is noticeable between Figs. 74, 69 and 66.

Fig. 68 shows a jet made by throttling steam at 95 pounds by