He hoped at first to obtain from manufacturers a gasoline engine sufficiently light and sufficiently powerful for a man-carrying machine. After several disappointments, the automobile industry being then in its infancy, he succeeded in constructing a five-cylinder gasoline motor of fifty-two horse-power and weighing only about a hundred and twenty pounds. He also constructed new launching apparatus. After tests with superposed sustaining surfaces, he adhered to the "single-tier plan." There is interesting evidence that in 1900 Langley renewed his study of the flight of soaring birds, the area of their extended wing surface in relation to weight, and the vertical distance between the center of pressure and the center of gravity in gulls and different species of buzzards. He noted among other things that the tilting of a wing was sufficient to bring about a complete change of direction. By the summer of 1903 two new machines were ready for field trials, which were undertaken from a large house-boat, especially constructed for the purpose and then moored in the mid-stream of the Potomac about forty miles below Washington. The larger of these two machines weighed seven hundred and five pounds and was designed to carry an engineer to control the motor and direct the flight. The motive power was supplied by the light and powerful gasoline engine already referred to. The smaller aeroplane was a quarter-size model of the larger one. It weighed fifty-eight pounds, had an engine of between two and a half and three horse-power, and a sustaining surface of sixty-six square feet. This smaller machine was tested August 8, 1903, the same launching apparatus being employed as with the steam-driven models of 1896. In spite of the fact that one of the mechanics failed to withdraw a certain pin at the moment of launching, and that some breakage of the apparatus consequently occurred, the aeroplane made a good start, and fulfilled the main purpose of the test by maintaining a perfect equilibrium. After moving about three hundred and fifty feet in a straight course it wheeled a quarter-circle to the right, at the same time descending slightly, the engine slowing down. Then it began to rise, moving straight ahead again for three or four hundred feet, the propellers picking up their former rate. Once more the engine slackened, but, before the aeroplane reached the water, seemed to regain its normal speed. For a third time the engine slowed down, and, before it recovered, the aeroplane had touched the water. It had traversed a distance of one thousand feet in twenty-seven seconds. One of the workmen confessed that he had poured into the tank too much gasoline. This had caused an overflow into the intake pipe, which in turn interfered with the action of a valve.