The free balloon was, however, entirely at the mercy of the winds, and the captive balloon could not be moved about readily, so that it was thus limited in its sphere of observation, except when attached to some movable conveyance. This showed the necessity of inventing some means of propulsion and steering. The first experiments were attempts to row ordinary spherical balloons, as you would a boat, but the earliest record of any definite progress being achieved in forcing a lighter-than-air craft through the air was the experiment in France of two brothers named Robert in 1784. They constructed a melon-shaped balloon, 52 feet long and 32 feet in diameter, made of proofed silk. The gas employed was pure hydrogen. Underneath this envelope was suspended a long, narrow car, in general idea not unlike that used on some modern airships; and three pairs of oars with blades made like racquet-frames covered with silk, and a rudder of similar material, were the only implements for navigation. The two brothers and their brother-in-law went up in the apparatus and succeeded in describing a curve of one kilometre radius, which showed, at any rate, that they could deviate slightly from the direction of the feeble wind then prevailing. The development of the steam-engine was potent with suggestions for aerial navigation of a dirigible. Thus, on December 24, 1852, Henry Gifford, another Frenchman, first ascended in a dirigible balloon. It was spindle-shaped, 143 feet long and 39 feet in diameter. It was driven by a 3 horse-power steam-engine and an 11-foot screw propeller. He went out from the Hippodrome in Paris and made six miles per hour relative to the air and several successful landings. This was the first recorded dirigible flight. A decade later, Tissandier, with a spindle-shaped balloon, much on the lines of those of his predecessors, succeeded in reaching a speed of eight miles an hour with the aid of an electric motor and a bichromate-of-potash battery. Captain Charles Renard brought the airship another stage toward realization by building an envelope with a true stream-line. The method of suspending the car was of the type adopted by later builders, namely, to place an enormous sheet over the back of the airship and to attach suspensory cords to its edges. This airship had a cubic capacity of 66,000 feet, and was kept rigid by means of an internal air balloonet or interior gas-bag which was confined to a definite shape by an outer framework or cover. This balloonet was kept full by a fan-blower coupled to the motor. The car was 108 feet long, and really served as a spar employed in later airships of what became known as the semirigid type.