The heat of the cylinder itself assists this process; so that in order to produce a tolerably perfect vacuum, it was found necessary to introduce a quantity of condensing water, sufficient to reduce the temperature of the water in the cylinder lower than 100°, and consequently to cool the cylinder itself to that temperature. Under these circumstances, the descent of the piston was found to suffer very little resistance from any vapour within the cylinder: but then on the subsequent ascent, an immense waste of steam ensued; for the steam, on being admitted under the piston, was immediately condensed by the cold cylinder and water of condensation, and this continued until the cylinder became again heated up to 212°, to which point the whole cylinder should be heated before the ascent could be completed. Here, then, was an obvious and an extensive cause of the waste of heat. At every descent of the piston, the cylinder should be cooled below 100°; and at every ascent it should be again heated to 212°. It, therefore, became a question whether the force gained by the increased perfection of the vacuum was adequate to the waste of fuel in producing the vacuum; and it was found, on the whole, more profitable not to cool the cylinder to so low a temperature, and consequently to work with a very imperfect vacuum, and a diminished power. Watt, therefore, found the engine involved in this dilemma: either much or little condensation-water must be used. If much were used, the vacuum would be perfect; but then the cylinder would be cooled, and would entail an extensive waste of fuel in heating it. If little were used, a vapour would remain, which would resist the descent of the piston, and rob the atmosphere of a part of its power. The great problem then pressed itself on his attention, _to condense the steam without cooling the cylinder_. From the small quantity of water in the form of steam which filled the cylinder, and the large quantity of injected water to which this communicated heat, Watt was led to inquire what proportion the bulk of water in the liquid state bore to its bulk in the vaporous state; and also what proportion subsisted between the heat which it contained in these two states. He found by experiment that a cubic inch of water formed about a cubic foot of steam; and that the cubic foot of steam contained as much heat as would raise a cubic inch of water to about 1000°. (15.) This gave him some surprise, as the thermometer indicated the same temperature, 212°, for both the steam and the water from which it was raised. What then became of all the additional heat which was contained in the steam, and not indicated by the thermometer? Watt concluded that this heat must be in some way engaged in maintaining the water in its new form.