One day, as he rode around Paris on a motorized De Dion tricycle, it occurred to him that he may have prematurely dismissed the petroleum motor. The single-cylinder tricycle engine, he realized, “happened to be very much perfected at the moment,” compared with the troublesome higher-powered petroleum engines in four-wheeled automobiles. Pound for pound the 1.75-horsepower motor in his De Dion was relatively powerful, although not strong enough to guide an airship. To increase the power, he planned to combine two of them. Usually he was cocky about his inventions, but this time he was not confident enough to experiment in public.
“I looked for the workshop of some little mechanic … in the central quarter of Paris,” recalled Santos-Dumont. “There I could have my plans executed under my own eyes and apply my own hands to the work. I found such a workshop in the Rue du Colisée. There I worked out a tandem of two cylinders of a petroleum motor, that is, their prolongation, one after the other, to work the same connecting rod, while fed by a single carburetor. To bring everything down to the minimum of weight, I cut out from each part what was not strictly necessary to solidity. In this way I realized something which was remarkable at the time—a 3½ horse-power motor weighing only sixty-six pounds.”
He was pleased with his handiwork and set out to test the reconstructed engine in his tricycle. The Paris – Amsterdam automobile race was approaching, and he could not think of a better way to put the engine through its paces than to enter the competition. He was disappointed to learn that his souped-up vehicle did not meet the eligibility requirements but made the best of the situation by driving the tricycle alongside the race until he convinced himself that he could keep pace with the leaders. “I might have had one of the first places at the finish (the average speed was only 40 kilometers, or 25 miles per hour),” wrote Santos-Dumont, “had I not begun to fear that the jarring of my motor in so long and strenuous an effort might at last derange it and delay the more important work on my air-ship. I, therefore, fell out of the race while still at the head of the procession.”
The shaking of the motor reminded him of how the machines on the coffee plantation had fallen apart from their own vibrations. To ensure that his balloon engine would not have a similar fate, he drove his tricycle to the Bois in the middle of the night when the park was abandoned. He had hired two burly workmen to meet him there with heavy-duty ropes and paid them generously so that they would tell no one about the nocturnal experiments. He selected an ample tree with a thick branch just above his head. The workmen tossed the ropes over the branch and tied them securely to both ends of the tricycle. He mounted the vehicle and gave the order to hoist him five feet into the air. With the engine going full throttle, he sat there feeling the vibrations; they were noticeable but were much less than they were on the ground, where the engine had something to vibrate against. He pronounced the test a success, swore the workmen to secrecy once more, and sneaked out of the park before he could be arrested for violating the curfew.
When dawn broke, he told friends about his plan. “From the beginning everybody was against the idea,” he recalled. “I was told that an explosive gas engine would ignite the hydrogen in the balloon above it, and that the resulting explosion would end the experiment with my life.” He reminded his doubters that half a century earlier Henri Giffard had gone up in a hydrogen balloon powered by a fiery steam engine and although the flight was only a qualified success (because the engine was not powerful enough to work against the wind), Giffard made it down unscathed. The tricycle engine, Santos-Dumont insisted, would spit far fewer sparks and less smoke.
He wrote down his plans for a cigar-shaped airship and returned to the balloon makers in Vaugirard. When he tried to place an order for the balloon, Lachambre at first refused to take it, Santos-Dumont recalled, “saying that such a thing had never been made, and that he would not be responsible for my rashness.” Santos-Dumont reminded him that he had voiced similar doubts before building Brazil. He also promised to indemnify Lachambre against any explosion or damages and agreed to work on the engine himself far away from the workshop. Worn down by Santos-Dumont’s persistence, Lachambre “went to work without enthusiasm.”
Santos-Dumont’s guiding principle in designing the dirigible, which he called Santos-Dumont No. 1 in anticipation of building a series of airships, was to make it the smallest elongated balloon that could carry aloft the reconfigured engine, a propeller, a rudder, the balloon basket, the rigging, a minimum of ballast, the guide rope, and of course his own weight, which fluctuated between a hundred and a hundred ten pounds depending on how well he had been eating. The sketch he gave Lachambre called for “a cylinder terminating at each end by a cone.” It would be 82.5 feet long by 11.5 feet in diameter, with a gas capacity of 6,454 cubic feet, which would give it a lifting power of 450 pounds. Knowing the weight of everything the balloon would have to take aloft, Santos-Dumont computed that only sixty-six pounds remained for the balloon material, the varnish, and the chemise (the outer cover, or netting, by which the balloon was attached to the basket). The use of Japanese silk, which had proved so staunch in Brazil, would not by itself keep the weight within sixty-six pounds. He needed another innovation. First he considered alternatives to the varnish, but he could not find a lighter liquid that would sufficiently seal the silk. Then he focused on the chemise, and ended up doing away with it entirely. The rigging lines from the basket would now be attached directly to the balloon envelope, the lines joined to long wooden rods housed in horizontal hems stitched into the balloon fabric. Santos-Dumont was proud of this simple idea, and instructed a reluctant Lachambre to sew the balloon accordingly. The veteran balloon maker was worried that the stitches might rip, releasing the basket for a fatal free-fall. As with the engine, Santos-Dumont absolved him of all responsibility.
While Lachambre went to work, Santos-Dumont busied himself in the rue du Colisée workshop getting the motor ready. He switched the motor from his tricycle to the rear of the balloon basket and attached an aluminum propeller directly to the motor shaft. By suspending the basket, the motor, and the 6.6-foot propeller from the rafters of the workshop, he got an idea of how the machinery would perform in the air. With the motor full throttle, the basket shot violently forward. Pulling the basket back with a horizontal rope attached to a dynamometer, Santos-Dumont measured the traction power of the propeller to be as high as twenty-five pounds, “promising good speed for a cylindrical balloon of my dimensions, whose length was equal to seven times its diameter.” He repeated the trials daily to be sure of the results. If all went well, he concluded, the airship would cruise along at eighteen miles an hour.
Having introduced the propeller (and a rudder made from silk stretched over a triangular steel frame) to wrest control of the balloon’s horizontal motion from the vagaries of the wind, he turned his attention to the question of vertical equilibrium, which was uneasily maintained in spherical balloons by the jettisoning of ballast or the venting of gas. “Suppose you are in equilibrium at five hundred meters height,” he wrote.
All at once a little cloud, almost imperceptible, masks the sun for a few seconds. The temperature of the gas in your balloon cools down a little; and if, at the very moment, you do not throw out enough ballast to correspond to the ascensional force lost by the condensation of the gas, you will begin descending. Imagine that you have thrown out the ballast—just enough, for if you throw too much, you will become too light and go too high. The little cloud ceases to mask the sun. Your gas heats up again to its first temperature and regains its old lifting-power. But, having less to lift by the amount of ballast thrown out, it now shoots higher into the air, and the gas in the balloon dilates still more, and either escapes through the safety-valve or has to be deliberately sacrificed, and the trouble recommences. These montagnes-russes, or shoot-the-chutes,