It will be remembered that in the Newcomen engine the steam was only employed for the purpose of producing a vacuum, and that its working power was in the down stroke, which was effected by the pressure of the air upon the piston; hence it is now usual to call it the atmospheric engine. Watt perceived that the air which followed the piston down the cylinder would cool the latter, and that steam would be wasted in re-heating it. In order, therefore, to avoid this loss of heat, he resolved to put an air-tight cover upon the cylinder, with a hole and stuffing-box for the piston-rod to slide through, and to admit steam above the piston, to act upon it instead of the atmosphere. When the steam had done its duty in driving down the piston, a communication was opened between the upper and lower part of the cylinder, and the same steam, distributing itself equally in both compartments, sufficed to restore equilibrium. The piston was now drawn up by the weight of the pump-gear; the steam beneath it was then condensed in the separate vessel so as to produce a vacuum, and a fresh jet of steam from the boiler was let in above the piston, which forced it again to the bottom of the cylinder. From an atmospheric it had thus become a true steam-engine, and with a much greater economy of steam than when the air did half the duty. But it was not only important to keep the air from flowing down the inside of the cylinder: the air which circulated within cooled the metal and condensed a portion of the steam within; and this Watt proposed to remedy by a second cylinder, surrounding the first with an interval between the two which was to be kept full of steam. One by one these various contrivances were struck out, modified, settled, and reduced to definite plans; the separate condenser, the air and water pumps, the use of fat and oil (instead of water as in the Newcomen engine) to keep the piston working in the cylinder air-tight, and the enclosing of the cylinder itself within another to prevent the loss of heat. They were all but emanations from the first idea of inventing an engine working by a piston, in which the cylinder should be kept continually hot and perfectly dry. “When once,” says Watt, “the idea of separate condensation was started, all these improvements followed as corollaries in quick succession; so that in the course of one or two days the invention was thus far complete in my mind.”