Trevithick saw without apparent reasoning, while his friend's reasonings failed to make plain the full bearing of the questions, and so cramped the position as to make a change of front difficult--an operation in which Trevithick excelled. We learn, however, that in 1801 he suggested a blast in copper-ore furnaces, and in 1811 was on the verge of a discovery that has since revolutionized the iron-smelter's art by the use of hot blast. Wasted heat from a blast-furnace 10 feet high led him to the conclusion that by doubling the height of the furnace, enabling the cold mineral thrown in at the top to take up the heat wasted through the top of the low furnace, seven-eighths of the coal would be saved. His idea of sending blast through the furnace of his steam-boiler to economize heat could have been readily applied to the iron furnace, and we should have had the modern hot-blast iron furnaces. [Rough draft.] "CAMBORNE, _March 5th, 1812_. "GENTLEMEN, "Your favour of the 15th February, with a sketch of your brewery, I have received; and from which I find the head of water is 30 feet above the brewery, which makes it difficult to erect the chain and buckets so as to take advantage of the whole height of water; and as the stream is so very small, it will not admit of losing any part of the power. "To erect a machine so high, to engage the whole fall, would be, I fear, more expense than the power you would get would warrant; therefore I would recommend it to be made use of in a cylinder, in the same way as we use falls of water of 200 feet in our Cornish copper mines. We allow one-third loss for friction and leakage in those machines; but your machine being so very small, the loss will exceed that proportion; therefore I cannot promise you above one-half of the real weight and fall to be performed on your machinery, and that must be by a well-executed machine, for a small defect would destroy the value of so trifling a power. "As there is no expansion in water, it will be somewhat difficult to make the machine turn the centres with a fly-wheel, for if the valve shuts a little too early or too late for the turn of the crank over the centre, the fly-wheel's velocity must break something by confining the water between the piston and the bottom of the cylinder, which, after the valve is shut, cannot make its escape, and not having an elastic principle, the piston will strike as dead on the water as on a piece of iron, because, unless the valve is shut by the engine before the stroke is finished, it cannot shut at all. "I know persons who have attempted to put fly-wheels on pressure-engines of this kind, but never yet has one been made to work rotative. I do not see much difficulty in making an engine of this kind to work a crank and fly-wheel, by connecting an air-vessel with the cylinder to receive the pressure and contract and expand and shut the valves, the same as in steam-engines.