The wings of this new power machine had a total span of a few inches more than forty feet, and the upper and lower wing surfaces were six feet apart. To reduce the danger of the engine ever falling on the pilot, it was placed on the lower wing a little to right of center. The pilot would ride lying flat, as on the glider, but to the left of center, to balance the weight. To guard against the machine rolling over in landing, the sled-like runners were extended farther out in front of the main surfaces than on the glider. These two runners were four feet, eight inches apart. The tail of the machine had twin movable vanes instead of a single vane as in the 1902 glider. The Wrights left the designing of the propellers until the last, because they felt sure that part of the job would be easy enough. Their tables of air pressures, derived from wind-tunnel experiments, would enable them, they thought, to calculate exactly the thrust necessary to sustain the machine in flight. But to design a propeller that would give the needed amount of thrust, with the power at their command, was a problem they had not yet considered. No data on air propellers were available, but the Wrights had always understood that it was not difficult to obtain an efficiency of fifty per cent with marine propellers. All that should be necessary would be to learn the theory of the operation of propellers from books on marine engineering and then substitute air pressures for water pressures. What could be simpler or easier? Accordingly, the brothers got several such books from the Dayton Public Library. But when they began to read those books, they discovered to their surprise that much less was known about propellers than they had supposed. All the formulae on propellers in the books were found to be based on experiment and observation rather than on theory. The marine engineers, when they saw that a propeller would not move a boat fast enough, had then tried one larger, or of a different pitch, until they got one that would serve their purpose. But they could not design a propeller on paper and foresee exactly what its performance on a certain type of motor-boat would be. Exact knowledge of the action of the screw propeller, though it had been in use for a century, was still lacking.