A french writer about the year 1630 describes a method of raising water to the upper part of a house by filling a chamber with steam, and suffering it to condense of itself, but it seems to have been mere theory, as his method was scarcely practicable as he describes it. In 1655 the Marquis of Worcester mentions a method of raising water by fire in his Century of Inventions, but he seems only to have availed himself of the expansive force and not to have known the advantages arising from condensing the steam by an injection of cold water. This latter and most important improvement seems to have been made by Capt. Savery sometime prior to 1698, for in that year his patent for the use of that invention was confirmed by act of parliament. This gentleman appears to have been the first who reduced the machine to practice and exhibited it in an useful form. This method consisted only in expelling the air from a vessel by steam and condensing the steam by an injection of cold water, which making a vacuum, the pressure of the atmosphere forced the water to ascend into the steam-vessel through a pipe of 24 to 26 feet high, and by the admission of dense steam from the boiler, forcing the water in the steam-vessel to ascend to the height desired. This construction was defective because it required very strong vessels to resist the force of the steam, and because an enormous quantity of steam was condensed by coming in contact with the cold water in the steam-vessel. About or soon after that time M. Papin attempted a steam-engine on similar principles but rather more defective in its construction. The next improvement was made very soon afterwards by Messrs. Newcomen and Cawley of Dartmouth, it consisted in employing for the steam-vessel a hollow cylinder, shut at bottom and open at top, furnished with a piston sliding easily up and down in it, and made tight by oakum or hemp, and covered with water. This piston is suspended by chains from one end of a beam, moveable upon an axis in the middle of its length, to the other end of this beam are suspended the pump-rods. The danger of bursting the vessels was avoided in this machine, as however high the water was to be raised it was not necessary to increase the density of the steam but only to enlarge the diameter of the cylinder. Another advantage was, that the cylinder not being made so cold as in Savary's method, much less steam was lost in filling it after each condensation.