The inventor finally succeeded in making a motor--remarkable at that time--which, weighing only 66 pounds, would produce 3-1/2 horse-power. It is easy to understand why a petroleum motor is such a power-producer for its size. The greater part of its fuel is in the air itself, and the air is all around the balloon, ready for use. The aëronaut does not have to take it up with him. That proportion of his fuel that he must carry, the petroleum, is comparatively insignificant in weight. A few figures will prove interesting. Two and one-half gallons of gasoline, weighing 15 pounds, will drive a 2-1/2 horse-power autocycle 94 miles in four hours. Santos-Dumont's balloon needs less than 5-1/3 gallons for a three hours' trip. This weighs but 37 pounds, and occupies a small cigar-shaped brass reservoir near the motor of his machine. An electric battery of the same horse-power would weigh 2,695 pounds. [Illustration: "Santos-Dumont No. 1."] Santos-Dumont tested his new motor very thoroughly by attaching it to a tricycle with which he made some record runs in and around Paris. Having satisfied himself that it was thoroughly serviceable he set about making the balloon, cigar-shaped, 82 feet long. "To keep within the limit of weight," he says, "I first gave up the network and the outer cover of the ordinary balloon. I considered this sort of second envelope, holding the first within it, to be superfluous, and even harmful, if not dangerous. To the envelope proper I attached the suspension-cords of my basket directly, by means of small wooden rods introduced into horizontal hems, sewed on both sides along the stuff of the balloon for a great part of its length. Again, in order not to pass the 66 pounds weight, including varnish, I was obliged to choose Japan silk that was extremely fine, but fairly resisting. Up to this time no one had ever thought of using this for balloons intended to carry up an aëronaut, but only for little balloons carrying light registering apparatus for investigations in the upper air. [Illustration: Basket of "Santos-Dumont No. 1." _Showing propeller and motor._] "I gave the order for this balloon to M. Lachambre. At first he refused to take it, saying that such a thing had never been made, and that he would not be responsible for my rashness. I answered that I would not change a thing in the plan of the balloon, if I had to sew it with my own hands. At last he agreed to sew and varnish the balloon as I desired." After repeated trials of his motor in the basket--which he suspended in his workshop--and the making of a rudder of silk he was able, in September, 1898, to attempt real flying. But, after rising successfully in the air, the weight of the machinery and his own body swung beneath the fragile balloon was so great that while descending from a considerable height the balloon suddenly sagged down in the middle and began to shut up like a portfolio.