When Watt found that water, in its conversion into vapour, became such a reservoir of heat, he was more than ever bent on economising it; for the great waste of heat involving so heavy a consumption of fuel, was felt to be the principal obstacle to the extended employment of steam as a motive power. He accordingly endeavoured, with the same quantity of fuel, at once to increase the production of steam, and to diminish its waste. He increased the heating surface of the boiler, by making flues through it; he even made his boiler of wood, as being a worse conductor of heat than the brickwork which surrounds common furnaces; and he cased the cylinders and all the conducting-pipes in materials which conducted heat very slowly. But none of these contrivances were effectual; for it turned out that the chief expenditure of steam, and consequently of fuel, in the Newcomen engine, was occasioned by the reheating of the cylinder after the steam had been condensed, and the cylinder was consequently cooled by the injection into it of the cold water. Nearly four-fifths of the whole steam employed was condensed on its first admission, before the surplus could act upon the piston. Watt therefore came to the conclusion, that to make a perfect steam-engine, it was necessary that _the cylinder should be always as hot as the steam that entered it_; but it was equally necessary that the steam should be condensed when the piston descended,--nay, that it should be cooled down below 100°, or a considerable amount of vapour would be given off, which would resist the descent of the piston, and diminish the power of the engine. Thus the cylinder was never to be at a less temperature than 212°, and yet at each descent of the piston it was to be less than 100°; conditions which, on the very face of them, seemed to be wholly incompatible.