The first really successful screw-propelled boats were those of Colonel John Stevens, which were in operation on the Hudson River from the years 1802 to 1806, and were the first to be used for the effective navigation of the waters of any country. References have already been made to Stevens’ experiment with paddle propulsion in 1796. When he, Chancellor Livingston, Nicholas J. Roosevelt, and Isambard Brunel were making experiments in steam propulsion on the Passaic River, New Jersey, they tried a horizontal centrifugal wheel in a boat of 30 tons, drawing water from the bottom of the boat and discharging it at the stern. This is in its general principles similar to the plan that Mr. Ruthven tried in England on the _Waterwitch_ more than half a century afterwards. They also, unsuccessfully, attempted to use elliptical paddle-wheels. Probably the best description of Colonel Stevens’ propeller is that which he himself contributed to the _Medical and Philosophical Journal_ of New York in January 1812. He refers to the “mischievous effects necessarily resulting from the alternating stroke of the engine of the ordinary construction” which induced him to turn his attention to the rotary principle of steam-engine construction. “For simplicity, lightness, and compactness the engine far exceeded any I have yet seen. A cylinder of brass, about eight inches in diameter and four inches long, was placed horizontally on the bottom of the boat: and by the alternate pressure of the steam on two sliding wings, an axis passing through its centre was made to revolve. On one end of this axis, which passed through the stern of the boat, wings like those on the arms of a windmill were fixed, adjusted to the most advantageous angle for operating on the water. This constituted the whole of the machinery. Working with the elasticity of the steam merely, no condenser, no air-pump was necessary; and as there were no valves, no apparatus was required for opening and shutting them. This simple little steam-engine was, in the summer of 1802, placed on board a flat-bottomed boat I had built for the purpose. This boat was 25 feet long, and about 5 or 6 feet wide. She was occasionally kept going until the cold weather stopped us. When the engine was in the best order, her velocity was about four miles an hour. I found it, however, impracticable, on so contracted a scale, to preserve due tightness in the packing of the wings in the cylinder for any length of time. This defect determined me to revert again to the reciprocating engine.” [Illustration: STEVENS’ 1804 ENGINE, FITTED INTO OPEN BOAT WITH TWIN-SCREW PROPELLERS.]