is a pear-shaped bag--usually made of silk--filled with some gas lighter than air. The tendency of a heavier medium to displace a lighter drives the gas upwards, and with it the bag and the wicker-work car attached to a network encasing the bag. The tapering neck at the lower end is open, to permit the free escape of gas as the atmospheric pressure outside diminishes with increasing elevation. At the top of the bag is a wooden valve opening inwards, which can be drawn down by a rope passing up to it through the neck whenever the aeronaut wishes to let gas escape for a descent. He is able to cause a very rapid escape by pulling another cord depending from a "ripping piece" near the top of the bag. In case of emergency this is torn away bodily, leaving a large hole. The ballast (usually sand) carried enables him to maintain a state of equilibrium between the upward pull of the gas and the downward pull of gravity. To sink he lets out gas, to rise he throws out ballast; and this process can be repeated until the ballast is exhausted. The greatest height ever attained by aeronauts is the 7-1/4 miles, or 37,000 feet, of Messrs. Glaisher and Coxwell on September 5, 1862. The ascent nearly cost them their lives, for at an elevation of about 30,000 feet they were partly paralyzed by the rarefaction of the air, and had not Mr. Coxwell been able to pull the valve rope with his teeth and cause a descent, both would have died from want of air. [Illustration: FIG. 171.] The _flying-machine_, which scientific engineers have so long been trying to produce, will probably be quite independent of balloons, and will depend for its ascensive powers on the action of air on oblique surfaces. Sir Hiram Maxim's experimental air-ship embodied the principles shown by Fig. 171. On a deck was mounted an engine, E, extremely powerful for its weight. This drove large propellers, S S. Large aeroplanes, of canvas stretched over light frameworks, were set up overhead, the forward end somewhat higher than the rear. The machine was run on rails so arranged as to prevent it rising. Unfortunately an accident happened at the first trial and destroyed the machine. In actual flight it would be necessary to have a vertical rudder for altering the horizontal direction, and a horizontal "tail" for steering up or down. The principle of an aeroplane is that of the kite, with this difference, that, instead of moving air striking a captive body, a moving body is propelled against more or less stationary air. The resolution of forces is shown by the arrows as before. Up to the present time no practical flying-machine has appeared. But experimenters are hard at work examining the conditions which must be fulfilled to enable man to claim the "dominion of the air." [34] The "Romance of Modern Mechanism," p. 243 Chapter XVIII. HYDRAULIC MACHINERY.