His second machine, built in the early spring of 1899, held over 7,000 cubic feet of gas and gave a further 44 lbs. of ascensional force. The balloon envelope was very long and very narrow; the first attempt at flight was made in wind and rain, and the weather caused sufficient contraction of the hydrogen for a wind gust to double the machine up and toss it into the trees near its starting-point. The inventor immediately set about the construction of 'Santos-Dumont No. 3,' on which he made a number of successful flights, beginning on November 13th, 1899. On the last of his flights, he lost the rudder of the machine and made a fortunate landing at Ivry. He did not repair the balloon, considering it too clumsy in form and its motor too small. Consequently No. 4 was constructed, being finished on the 1st, August, 1900. It had a cubic capacity of 14,800 feet, a length of 129 feet and greatest diameter of 16.7 feet, the power plant being a 7 horse-power Buchet motor. Santos-Dumont sat on a bicycle saddle fixed to the long bar suspended under the machine, which also supported motor propeller, ballast; and fuel. The experiment of placing the propeller at the stem instead of at the stern was tried, and the motor gave it a speed of 100 revolutions per minute. Professor Langley witnessed the trials of the machine, which proved before the members of the International Congress of Aeronautics, on September 19th, that it was capable of holding its own against a strong wind.