It has been demonstrated by experiments made in France that the power required to compress moist air is less than that for dry air. A table showing the power required to compress moist and dry air has been prepared from the data of M. Mallard and shows that for five atmospheres the work expended in compressing one pound of dry air is 58,500 foot pounds, while that for moist air is 52,500 foot pounds. In expansion also moisture in the air adds to the economy, but in both cases the saving of power is not great enough to compensate for the many disadvantages due to the presence of water. Mr. Norman Selfe, of the Engineering Association of N.S.W., has compiled a table which shows some important theoretical conditions involved in producing compressed air. So much for the theory of compression. We now come to the practical production of compressed air. The first record that we have of the use of an air compressor is at Ramsgate Harbor, Kent, in the year 1788. Smeaton invented this "pump" for use in a diving apparatus. In 1851, William Cubitt, at Rochester Bridge, and a little later an engineer, Brunel, at Saltash, used compressed air for bridge work. But the first notable application of compressed air is due to Professor Colladon, of Geneva, whose plans were adopted at the Mont Cenis tunnel. M. Sommeiller developed the Colladon idea and constructed the compressed air plant illustrated in Fig. 2. [Illustration: FIG. 2.] The Sommeiller compressor was operated as a ram, utilizing a natural head of water to force air at 80 pounds pressure into a receiver. The column of water contained in the long pipe on the side of the hill was started and stopped automatically, by valves controlled by engines. The weight and momentum of the water forced a volume of air with such shock against a discharge valve that it was opened and the air was discharged into the tank; the valve was then closed, the water checked; a portion of it was allowed to discharge and the space was filled with air, which was in turn forced into the tank. The efficiency of this compressor was about 50 per cent. At the St. Gothard tunnel, begun in 1872, Prof. Colladon first introduced the injection of water in the form of spray into the compressor cylinder to absorb the heat of compression. [Illustration: FIG. 3.] Fig. 3 illustrates the air cylinder of the Dubois-Francois type of compressor, which was the best in use about the year 1876. This compressor was exhibited at the Centennial Exposition and was adopted by Mr. Sutro in the construction of the Sutro tunnel. A characteristic feature seems to be to get as much water into the cylinder as possible. The water which flooded the bottom of the cylinder arose from the voluminous injection; this water was pushed into the end of the cylinder and some of it escaped with the air through the discharge valve. An improved pattern of this compressor is shown in Fig. 4. [Illustration: FIG. 4.]