To make it air-tight in the cylinder, it had been found necessary to keep a quantity of water supplied above the piston. In the present case, any of this water which might escape through the piston, or between it and the cylinder, would boil, the cylinder being kept at 212°; and would thus, by the steam it would produce, vitiate the vacuum. To avoid this inconvenience, Watt proposed to lubricate the piston, and keep it air-tight, by employing melted wax and tallow. Another inconvenience was still to be removed. On the descent of the piston, the air which must then enter the cylinder would lower its temperature; so that upon the next ascent, some of the steam which would enter it would be condensed, and hence would arise a source of waste. To remove this difficulty, Watt proposed to close the top of the cylinder altogether by an air-tight and steam-tight cover, allowing the piston-rod to play through a hole furnished with a stuffing-box, and _to press down the piston by steam instead of the atmosphere_. This was the third step in this great invention, and one which totally changed the character of the machine. It now became really a _steam engine_ in every sense; for the pressure above the piston was the elastic force of steam, and the vacuum below it was produced by the condensation of steam; so that steam was used both directly and indirectly as a moving power; whereas, in the atmospheric engine, the indirect force of steam only was used, being adopted merely as an easy method of producing a vacuum. The last difficulty respecting the economy of heat which remained to be removed, was the circumstance of the cylinder being liable to be cooled on the external surface by the atmosphere. To obviate this, he first proposed casing the cylinder in wood, that being a substance which conducted heat slowly. He subsequently, however, adopted a different method, and inclosed one cylinder within another, leaving a space between them, which he kept constantly supplied with steam. Thus the inner cylinder was kept continually at the temperature of the steam which surrounded it. The outer cylinder was called the _jacket_.[15] [Footnote 15: It is a remarkable circumstance, that Watt used the same means for keeping the cylinder hot as Newcomen used in his earlier engines to cool it. (38.)] (48.) Watt computed that in the atmospheric engine three times as much heat was wasted in heating the cylinder, &c. as was spent in useful effect. And, as by the improvements proposed by him nearly all this waste was removed, he contemplated, and afterwards actually effected, a saving of three fourths of the fuel.