It is not easy to deal with the important reasonings flowing from those facts, and influencing the form and economy of the steam-engine, nor to show if Trevithick was right in discrediting the laws laid down by Watt. Newcomen's engine had the interior, as well as the exterior of the steam-cylinder exposed to the cooling atmosphere. Watt, by putting a cover on the cylinder, reduced the loss of the heat from the interior, and by his steam-case hoped to reduce the loss from the exterior, though by it he increased the amount of surface exposed to the cold. In Trevithick's early engines the boiler alone exposed heat-losing surface, and this was further reduced by its own comparatively small size, the engine and boiler complete not exposing one-quarter of the surface of a Watt low-pressure engine of equal power. One object of the Binner Downs experiment was to further curtail this loss of power by increasing the heat of the steam while in operation in the cylinder, since called superheating steam. This principle of giving increased heat to steam, after it had left its state as water, was made practical by Trevithick's boiler at Wheal Prosper in 1810, where the flues having first been carried around the water portion of the boiler, then passed over the steam portion;[166] and again in the upright boiler of 1815, having the upper end of the fire-tube surrounded by steam above the water line.[167] Those early beginnings of superheating steam and surface condensation culminated in the Binner Downs experiments of 1828, one immediate practical result of which was the tubular surface condenser, enabling steamboat boilers to avoid, in a great measure, the use of salt water, facilitating in a marked degree the application of marine boilers and engines with steam of an increased pressure. [Footnote 166: See vol. ii., p. 71.] [Footnote 167: See vol. i., p 354.] The Binner Downs engine, with a cylinder of 70 inches in diameter, and a stroke of 10 feet when working with steam in the boilers of 45 lbs. to the square inch above the atmosphere, and using the heating flues around the cylinder, required 13 gallons of injection-water at each stroke, and consumed at the rate of 3 bushels of coal an hour, to produce a duty equal to eighteen millions; by removing the cylinder superheating flues, the quantity of injection-water for the same amount of work increased to 15-1/2 gallons, and the coal to 4-1/4 bushels. Watt's rule for his low-pressure steam vacuum engine doing a duty of eighteen millions, gave 57 gallons of injection-water, and 11-1/4 bushels of coal. On the question of coal, this statement agrees very nearly with Trevithick's letters of sixteen years before, when he used the high-pressure boilers in the Dolcoath pumping engine,[168] promising that his high-pressure expansive engine would do the work with one-third of the coal required in the low-pressure vacuum engine. [Footnote 168: See vol. ii., p. 171.]