Fulton believed that previous failures were due not so much to a defective steam engine, as to the wrong methods employed in applying the steam power thus generated. He criticised Rumsey’s method of propelling a ship by forcing water through the stern (in a manner similar to that which John Allen and Fitch had suggested) as the worst method of all. Ten years before his Seine success Fulton had been in communication with the Earl of Stanhope, who in 1790 had patented a means of propelling a ship in a strange way. This consisted in using a gigantic arrangement resembling a duck’s foot, placed on either side. These feet opened and shut like umbrellas and could send the ship along at three miles an hour. Fulton, then staying at Torquay, wrote to Lord Stanhope and proposed the use of paddle-wheels, but the noble earl would not listen to the suggestion. A similar freak idea was also put into practice in North America in 1792 by one Elijah Ormsbee. [Illustration: FULTON’S FIRST PLANS FOR STEAM NAVIGATION _From the Drawings in possession of the Rt. Hon. the Earl of Stanhope._] The illustrations on this page represent Fulton’s first plans for steam navigation. They were sent by him to Lord Stanhope in the year 1793 and are here reproduced from a copy, by kind permission of the present earl. In his letter descriptive of these ideas Fulton shows the upper part of this illustration, marked No. 1, to be an attempt to imitate the spring in the tail of a salmon. Amidships will be noticed an object resembling a bow such as one usually associates with arrows. This bow was to be wound up by the steam engine, and the collected force attached to the end of the paddle, shown in the stern of the boat, would urge the ship ahead. But the sketch of a ship in the lower part of the picture marked No. 2 represents the model at which he was then working. It will be noticed that she has something of the characteristic stern which was so marked a feature of the sailing ships of this period and had been inherited from the Dutch of the seventeenth century, and is still traceable in the design of the modern royal steam yachts in England, as will be seen by a comparison with the illustration of the _Alexandra_. In referring to this No. 2, Fulton points out that he had found that three or six paddles answered better than any other number, since they do not counteract each other. By being hung a little above the water there is allowed a short space from the delivery of one paddle to the entrance of the other, and, also, the paddle enters the water more perpendicularly; the dotted lines show its situation when it enters and when it is covered. In the smaller illustration, No. 3, he emphasises the importance of arranging the paddle-blades still further. Thus the paddles _A_, _B_, _C_, and _D_ strike the water almost flat and rise in the same situation, whilst that paddle marked _E_ is the only one that pulls, the others acting against it.