Well, he did get a large portion of the water out of the cask by working the pump with all his might, but as soon as the water flowed out, he heard a gurgling, or bubbling, sound made by the air which was pressed in through all the joints which were not quite air-tight. Then he put one cask inside another, but he did not succeed much better, because again there was a noise resembling hissing and the twittering of birds. This experiment continued for several days, and in the end the cask was found full of air. Guericke therefore was convinced that he could not get an empty space while using a wooden cask. He then tried a ball made of beaten copper. Three men pumped the air out of this ball, and no sooner had most of the air been pumped out than the ball collapsed with a loud report, crushed by the pressure of the outside air. He then had another copper globe made. It was divided into halves, and these were joined by means of a ring of leather which was well soaked in a mixture of wax and turpentine. When that globe was emptied of air by means of the pump, the halves stuck so well together, that no human power was able to separate them; but when air was let in by turning a cock, they immediately fell apart. This experiment was performed again and again, and on one famous occasion it was performed before the German Emperor and his Parliament. On this occasion the two half-globes were fixed so firmly together by the pressure of the air merely, that it took sixteen horses to pull them apart. The total force with which they stuck together was about equal to the weight of a ton. When Boyle heard of these experiments, he made an air-pump of his own. He arranged it so that he could easily pump the air out of a glass globe which had a valve. [Illustration: Fig. 38.--Old Print showing horses trying to pull apart two Magdeburg Hemispheres from which the air had been pumped.] He could put in various objects from above by opening the valve, and then pump out the air and see what happened. The whole globe could be taken off the pump very easily, since it had a brass nozzle which fitted into the pump. The brass nozzle was fixed to the glass by means of a cement consisting of pitch, resin, and potash.