After spending some time in making these experiments and realising the enormous amount of muscular power which was needed, it was suggested to Miller by James Taylor, who was tutor to his children and a personal friend of William Symington, of Wanlockhead, that it would be far preferable to employ steam power to drive the paddle-wheels; and the upshot was that Symington was commissioned to design a suitable engine, which in October of 1788 was placed on one deck of a double-hulled pleasure craft 25 feet long and 7 feet wide, whilst the boiler was placed on the other deck. Thus fitted, the strange little ship was tried on Dalswinton Loch, Dumfriesshire, when she exhibited a speed of five knots per hour, and afterwards seven knots. At the first attempt the boards of the paddle-wheels were broken by concussion. Symington’s engine, however, was really of the atmospheric pattern, with the addition of a separate condenser, and was an infringement of Watt’s patent. After but a few trials the experiments accordingly had to be abandoned, although Miller afterwards got into communication with Boulton and Watt, whom he endeavoured to interest in steam navigation, but they declined. Miller next bought one of the boats used on the Forth and Clyde Canal, and gave an order to the Carron Iron Works to make a steam engine in accordance with Symington’s plan. On December 26, 1789, this vessel towed a heavy load seven miles an hour, but was afterwards dismantled. Symington’s first engine is shown in the illustration facing page 42, which is taken from a model in the South Kensington Museum, the original being in the Andersonian Museum, Glasgow, and it will be useful for reference in case our description of Newcomen’s engine was lacking in clearness. As will be noticed, there are two cylinders, each being open at the top, and a piston working up and down inside. It will be seen, too, that there are two paddle-wheels; these were placed in the ship fore and aft between the two hulls, and not on either side as in our modern paddle-wheel steamers. There were eight floats in each wheel, which were not feathering, but fixed. Each piston was connected with a drum by means of chains, the latter turning the drums alternately in opposite directions, and power was obtained both from the upward and downward strokes. By means of a ratchet arrangement, alternately engaging with pawls, the paddle-wheel was made always to revolve in one direction. The engine was fitted with air pumps for the purpose of which we have already dealt. In many ways it will be seen that Symington’s engine and gear resembled the method proposed by Hulls.