[31]: In an experimental trip with a heavy train at 12 miles an hour, I found the consumption of coke to be only four ounces per ton per hour. I believe, however, the practical consumption in ordinary work to be very nearly eight ounces.

[32]: Some of the preceding observations appeared in an article contributed by me to the British and Foreign Review.

[33]: In boilers constructed for stationary purposes, or for steam navigation, the steam-pipe, after it has passed through the blowing box, is continued and made to form a series of returned flues over the boiler, so as to take up the waste heat after it has passed the boiler, and before it reaches the chimney. But in locomotive engines for common roads, it has been found by experience, that the power gained by the waste heat is not sufficient to propel the weight of the material necessary for taking it up.

[34]: See Report of the Commons.

[35]: Ten thousand grains of pure sea water contain muriate of soda 220 grs., sulphate of soda 33 grs., muriate of magnesia 42 grs., and muriate of lime 8 grs.

[36]: This order has, as Mr. Hall informs me, been given without requiring any guarantee as to the performance of the engines.

[37]: The fuel used in this form of engine is coke, and not coal. A ton of coke occupies the same space as two tons of coal; the saving of tonnage, therefore, by the increased economy of fuel will not be in so great a proportion as the saving of fuel. A quantity of fuel of equivalent power will occupy about half the present space, but the displacement or immersion which it produces will be only one fourth of its present effect.

[38]: This splendid ship is 860 tons burden, with engines of 220 horse power.

[39]: See Report quoted in Mechanic's Magazine, vol. xxii. p. 275. This saving cannot amount to less than 40 per cent. upon the whole consumption of fuel; it certainly is considerably beyond what I should have conceived to be possible; but I have no reason to doubt the accuracy of the Report. I estimate the former consumption at 10 pounds per horse power per hour, which on 220 horse-power would be 2200 pounds, of which 840 pounds = 10 bushels, would be about 4-10ths.

[40]: See a more detailed account of these reports in the Mechanic's Magazine, vol. xxii. p. 274., from which I have taken the drawing of this paddle-wheel; and also see Report of the Committee of the House of Commons on Steam Navigation to India, Evidence of William Morgan, p. 95.