When compressed air was first tried, it was found that the loss of power was enormous. It was difficult to store, for the air leaked rapidly away; it was expensive to generate, and there were thermo-dynamic difficulties in its use without number. When a thousand cubic feet of air is jammed into the space of one, a large amount of heat is developed, and in order to store and use the air this heat must in some way be drawn off. Similarly, air at high pressure, when released, cools rapidly. The result, if there be a sufficient moisture, is freezing and clogging. For a long time it was thought these difficulties were largely insuperable.

Now, however, these very difficulties are turned to a profit—to such excellent profit, indeed, as to afford an apparent paradox. It seems idle to assert that it is possible to get as much power out of a machine as you put into it—this means a frictionless and wasteless mechanism. And yet a very near approach to just this condition seems to have been made in the case of compressed air. This is due to the development of the reheating process. Lest the reader be not familiar with the technique of the subject, it may not be idle to explain its broader features. In the process of compression the air is sucked into a piston, and then rammed into a reservoir surrounded by a water jacket, the latter drawing off the heat generated in the compression. The machine which does this work is a beautiful affair of what is known as the four-stage type. That is to say, the air is first driven up to about eighty pounds pressure and cooled; then turned into a second cylinder, where it is compressed still further, then cooled again; and so on up to the desired point. Thus even at two or three thousand pounds pressure to the square inch the air within the reservoir remains at somewhere near the temperature of the outside atmosphere. But if the air be used in this condition, not only will a large share of the power employed in compression be lost, but it will, as already noted, have a tendency to freeze everything within reach. If, however, as it is released, it is passed through a heater or is shot through superheated hot water it will, under the well-known properties of air, enormously expand. In actual practice it has been found possible to add, by reheating, one horse-power to each horsepower developed by compression, at one-eighth or one-tenth the original cost of the latter. That is to say, if a given quantity of compressed air costs a dollar to generate, the further expenditure of ten cents in reheating will double its power to do work. Theoretically the total efficiency thus obtained is actually greater than if the same amount of coal had been burned in an ordinary steam-engine and the power thus generated used direct. In practical use it is slightly less.