The young Brazilian ascended thirty times in spherical balloons before he attempted any work on an elongated shape.  He realised that many things must be learned before he could handle successfully the much more delicate and sensitive elongated gas-bag.

In general, Santos-Dumont worked on the theory of the dirigible balloon—­that is, one that might be controlled and made to go in any direction desired, by means of a motor and propeller carried by a buoyant gas-bag.  His plan was to build a balloon, cigar-shaped, of sufficient capacity to a little more than lift his machinery and himself, this extra lifting power to be balanced by ballast, so that the balloon and the weight it carried would practically equal the weight of air it displaced.  The push of the revolving propeller would be depended upon to move the whole air-ship up or down or forward, just as the motion of a fish’s fins and tail move it up, down, forward, or back, its weight being nearly the same as the water it displaces.

The theory seems so simple that it strikes one as strange that the problem of aerial navigation was not solved long ago.  The story of Santos-Dumont’s experiments, however, his adventures and his successes, will show that the problem was not so simple as it seemed.

Santos-Dumont was built to jockey a Pegasus or guide an air-ship, for he weighed but a hundred pounds when he made his first ascensions, and added very little live ballast as he grew older.

Weight, of course, was the great bugbear of every air-ship inventor, and the chief problem was to provide a motor light enough to furnish sufficient power for driving a balloon that had sufficient lifting capacity to support it and the aeronaut in the air.  Steam-engines had been tried, but found too heavy for the power generated; electric motors had been tested, and proved entirely out of the question for the same reason.

Santos-Dumont has been very fortunate in this respect, his success, indeed, being largely due to the compact and powerful gasoline motors that have been developed for use on automobiles.

Even before the balloon for the first air-ship was ordered the young Brazilian experimented with his three-and-one-half horse-power gasoline motor in every possible way, adding to its power, and reducing its weight until he had cut it down to sixty-six pounds, or a little less than twenty pounds to a horse-power.  Putting the little motor on a tricycle, he led the procession of powerful automobiles in the Paris-Amsterdam race for some distance, proving its power and speed.  The motor tested to his satisfaction, Santos-Dumont ordered his balloon of the famous maker, Lachambre, and while it was building he experimented still further with his little engine.  To the horizontal shaft of his motor he attached a propeller made of silk stretched tightly over a light wooden framework.  The motor was secured to the aeronaut’s basket behind, and the reservoir of gasoline hung to the basket in front.  All this was done and tested before the balloon was finished—­in fact, the aeronaut hung himself up in his basket from the roof of his workshop and started his motor to find out how much pushing power it exerted and if everything worked satisfactorily.