Just as the new building was nearing completion, the Kitty Hawk region had one of the worst storms in years. It came without warning, soon blowing forty miles an hour, and increased during the night until the next day the wind was more than seventy-five miles an hour. Orville risked climbing to the roof to nail down some of the more exposed parts. But by the time he got to the roof edge, the wind had blown his coat about him in a manner to pinion his arms and leave him helpless. Wilbur rushed to his assistance and held down his coat, but the wind was so strong that it was almost impossible to swing a hammer accurately enough to hit a nail. Three weeks were needed for assembling the new machine. From time to time, also, they took out the 1902 glider, still in fairly good condition in the shed where they had left it, and got practice. After the first few trials each brother was able to make a new world’s record by gliding for more than a minute. It was hoped to have the power machine ready for its first trial early in November. But at the first run of the motor on the completed machine, an unexpected strain from back-firing twisted one of the propeller shafts and tore loose the cross-arm to which the propeller was fastened. Both shafts were then sent back to the bicycle shop at Dayton to be made stronger. Dr. Spratt had arrived on October 23 to witness tests of the new machine, but the weather had become so wintry that he started home on November 5, taking with him as far as Norfolk the shafts for shipment to Dayton. Octave Chanute came, on invitation, the next day, but he too found it difficult to be comfortable with the weather increasingly wintry and he stayed less than a week. Before leaving camp, Chanute had unintentionally given them something else to worry about. He had remarked that at least twenty per cent usually must be allowed in chain transmission for loss in power. As the Wrights had allowed only five per cent, they felt considerable alarm. Since Chanute was a capable and famous engineer, it seemed prudent to find out whose estimates were more nearly correct. After Chanute had gone, the brothers suspended one of the drive chains over a sprocket and hung a bag of sand at each end of the chain. By measuring the amount of weight on one side needed to lift that on the other, they calculated the loss in transmission. As nearly as they could tell, this loss was even less than the five per cent they had estimated.