The ballonet became particularly valuable as soon as airship construction became general, and it was in the course of advance in Astra Torres design that the project was introduced of using the ballonets in order to give inclination from the horizontal. In the earlier Astra Torres, trimming was accomplished by moving the car fore and aft--this in itself was an advance on the separate 'sliding weigh' principle--and this was the method followed in the Astra Torres bought by the British Government from France in 1912 for training airship pilots. Subsequently, the two ballonets fitted inside the envelope were made to serve for trimming by the extent of their inflation, and this method of securing inclination proved the best until exterior rudders, and greater engine power, supplanted it, as in the Zeppelin and, in fact, all rigid types. In the kite balloon, the ballonet serves the purpose of a rudder, filling itself through the opening being kept pointed toward the wind--there is an ingenious type of air scoop with non-return valve which assures perfect inflation. In the S.S. type of airship, two ballonets are provided, the supply of air being taken from the propeller draught by a slanting aluminium tube to the underside of the envelope, where it meets a longitudinal fabric hose which connects the two ballonet air inlets. In this hose the non-return air valves, known as 'crab-pots,' are fitted, on either side of the junction with the air-scoop. Two automatic air valves, one for each ballonet, are fitted in the underside of the envelope, and, as the air pressure tends to open these instead of keeping them shut, the spring of the valve is set inside the envelope. Each spring is set to open at a pressure of 25 to 28 mm. II. THE FIRST DIRIGIBLES Having got off the earth, the very early balloonists set about the task of finding a means of navigating the air but, lacking steam or other accessory power to human muscle, they failed to solve the problem. Joseph Montgolfier speedily exploded the idea of propelling a balloon either by means of oars or sails, pointing out that even in a dead calm a speed of five miles an hour would be the limit achieved. Still, sailing balloons were constructed, even up to the time of Andree, the explorer, who proposed to retard the speed of the balloon by ropes dragging on the ground, and then to spread a sail which should catch the wind and permit of deviation of the course. It has been proved that slight divergences from the course of the wind can be obtained by this means, but no real navigation of the air could be thus accomplished.