In the year 1690, Papin took the first step towards constructing the steam-engine by inventing a cylinder in which the piston, instead of being driven by the explosion of gunpowder, was forced upwards by boiling some water under it. This plan made the plug unnecessary. The piston A was driven up by the steam, and fixed at the top by a rod E until it could be taken off the fire and made to do some work. There was one difficulty about this cylinder. If the piston fitted tight, it was not able to sink down to the level of the water at the bottom of the cylinder, because in doing so the air in the cylinder would be compressed, and would resist with a great force. This difficulty was got rid of by Papin by means of another rod M, which fitted in a hole in the piston. This rod was pulled out so as to let the air escape. When the piston was down far enough, the rod was pushed into the hole again so as to stop it up. The great fault of this machine was that the cylinder had first to be heated in order to drive out the piston and afterwards cooled, both of which processes took time. About the year 1704, a German prince gave Papin the order to build a steam-engine for pumping up water. [Illustration: Fig. 45.--Papin’s Steam-Engine.] Just at about that time Papin had seen a design of a machine by a man of the name of Savery, which was not, however, a success. His own machine consisted of a boiler which was to be put in the fire and kept there. The steam from this boiler was conducted into a cylinder half-filled with water. On the top of this water there was a wooden float, so that the steam hardly came into contact with the water at all. The float contained an iron box, into which a red hot iron ball was placed. The steam, therefore, was not cooled on entering the cylinder, but drove the float to the bottom of the cylinder and forced the water into an upright pipe which was surrounded by another cylinder. This water compressed the air in the second cylinder, and valves were so arranged that the air could not turn back. It was therefore used for forcing water up a pipe. Water was then allowed to return into the first cylinder, and the operation was repeated. The first cylinder afterwards always remained sufficiently hot to preserve the power of the steam. In this respect Papin’s machine was a great improvement on Newcomen’s machine set up at Wolverhampton in the year 1711. Papin tried his machine in the presence of the German prince in the year 1706. He did succeed in pumping water seventy feet high, but the long pipes began to leak, and the German prince was so disgusted that he would not give the money necessary to carry out any further experiments. Papin, therefore, went back to London, and in the next year he published a book called _New Art of Lifting Water with the Help of Fire in a most Successful Manner_. He had great plans about his machine. It was not only to pump water out of mines, but also to fire guns, and to drive ships and carriages.