The earlier aviators had found that two planes, or “double-deckers,” gave the best results. The Wrights adopted this type, believing that it was the strongest form, and could be made more compact and be more easily managed than the single plane, or the many-winged type. They built their gliding-machine of cloth and spruce and steel wire. But instead of the aviator hanging below the wings, as in the other planes, he lay flat across the centre of the lower wing. A horizontal rudder extended in front of the plane instead of behind it. This not only guided the flight of the machine, but counterbalanced the changes of the centre of air-pressure. To steer, the wings were moved by cords controlled by the aviator’s body. They considered that the shiftings of the air were too rapid to be followed by conscious thought, and so their plan was to have a plane that would balance automatically, or by reflex action, as a bicycle is balanced. Langley had adopted wings that slanted upward from the point at which they joined, copying the wings of a soaring buzzard. The Wrights doubted whether this was the best form for shifting weather, and built theirs more on the pattern of the gull’s wings, curving slightly at the tips. They were made of cloth, arched over ribs to imitate the curved surfaces of bird’s wings, and were fastened to two rectangular wooden frames, fixed one above the other by braces of wood and wire. Their next step was to try to find some method by which they might keep their gliding-machine continuously in the air, so that they might gain an automatic balance. The old method of launching the plane from a hill gave little chance for a real test. Study taught them that birds are really aeroplanes, and that buzzards and hawks and gulls stay in the air by balancing on or sliding down rising currents of air. They looked for a place where there should be winds of proper strength to balance their machine for a considerable time as it slid downward, and decided to make their experiments at Kitty Hawk, North Carolina, on the stretch of sand-dunes that divided Albemarle Sound from the Atlantic Ocean. They calculated that their gliding-machine, with 165 square feet of surface, should be held up by a wind blowing twenty-one miles an hour. The machine was to be raised like a kite, with men holding ropes fastened to the end of each wing. When the ropes were freed the aviator would glide slowly to the ground, having time to test the principle of equilibrium. This plan would also do away with the former need of carrying the plane up to the top of a hill before each flight. They found in practice that their plan of raising the plane like a kite was impracticable, and that the wind was not strong enough to support it at a proper angle. They had to glide from hills as others had done, but they discovered that their theory of steering and balancing by automatically shifting surfaces worked very much better than the old method of shifting the aviator’s weight.