pressure acting against a vacuum produced by the condensation of steam, he stated that his invention, besides being applicable to pumping water, could be available for rowing vessels against wind and tide, which he proposed to accomplish in the following manner.

Paddle-wheels, such as have since been brought into general use, were to be placed at the sides, and attached to a shaft extending across the vessel. Within the vessel, and under this shaft, he proposed to place several cylinders supplied with pistons, to be worked by the atmospheric pressure. On the piston-rods were to be constructed racks furnished with teeth: these teeth were to work in the teeth of wheels or pinions, placed on the shaft of the paddlewheels. These pinions were not to be fixed on the shaft, but to be connected with it by a ratchet; so that when they turned in one direction, they would revolve without causing the shaft to revolve; but when driven in the other direction, the catch of the ratchet-wheel would act upon the shaft so as to compel the shaft and paddle-wheels to revolve with the motion of the pinion or wheel upon it. By this arrangement, whenever the piston of any cylinder was forced down by the atmospheric pressure, the rack descending would cause the corresponding pinion of the paddle-shaft to revolve; and the catch of the ratchet wheel, being thus in operation, would cause the paddle-shaft and paddle-wheels also to revolve; but whenever the piston would rise, the rack driving the pinion in the opposite direction, the catch of the ratchet wheel would merely fall from tooth to tooth, without driving the paddle-shaft.

It is evident that by such an arrangement a single cylinder and piston would give an intermitting motion to the paddleshaft, the motion of the wheel being continued only during the descent of the piston; but if several cylinders were provided, then their motion might be so managed, that when one would be performing its ascending stroke, and therefore giving no motion to the paddle-shaft, another should be performing its descending stroke, and therefore driving the paddle-shaft. As the interval between the arrival of the piston at the bottom of the cylinder and the commencement

of its next descent would have been, in the imperfect machine conceived by Papin, much longer than the time of the descent, it was evident that more than two cylinders would be necessary to insure a constantly acting force on the paddle-shaft, and, accordingly, Papin proposed to use several cylinders.

In addition to this, Papin proposed to construct a boiler having a fireplace surrounded on every side by water, so that the heat might be imparted to the water with such increased rapidity as to enable the piston to make four strokes per minute. These projects were promulged in 1690, but it does not appear that they were ever reduced to experiment.

(106.) Savery proposed, in his original patent, in 1698, to apply his steam engine as a general prime mover for all sorts of machinery, by causing it to raise water to make an artificial fall, by which overshot water-wheels might be driven. This proposal was not acted on during the lifetime of Savery, but it was at a subsequent period partially carried into effect. Mr. Joshua Rigley erected several steam engines on this principle at Manchester, and other parts of Lancashire, to impel the machinery of some of the earliest manufactories and cotton mills in that district. The engines usually raised the water from sixteen to twenty feet high, from whence it was conveyed to an overshot wheel, to which it gave motion. The same water was repeatedly elevated by the engine, so that no other supply was necessary, save what was sufficient to make good the waste. These engines continued in use for some years, until superseded by improved machines.*

(107.) In 1736, Jonathan Hulls obtained a patent for a method of towing ships into or out of harbour against wind and tide. This method was little more than a revival of that proposed by Papin in 1690. The motion, however, was to be communicated to the paddle-shaft by a rope passing over a pulley fixed on an axis, and was to be maintained during the returning stroke of the piston by the descent of a weight which was elevated during the descending stroke. There is no record, however, of this plan, any more than that of Papin, ever having been reduced to experiment.

(108.) During the early part of the last century the manu

*Farey, Treatise on the Steam Engine, p. 122.

factures of this country had not attained to such an extent as to render the moving power supplied by water insufficient or uncertain to any inconvenient degree; and accordingly mills, and other works in which machinery required to be driven by a moving power, were usually built along the streams of rivers. About the year 1750 the general extension of manufactures, and their establishment in localities where water power was not accessible, called the steam engine into more extensive operation. In the year 1752, Mr. Champion, of Bristol, applied the atmospheric engine to raise water, by which a number of overshot wheels were driven. These were applied to move extensive brass-works in that neighbourhood, and this application was continued for about twenty years, but ultimately given up on account of the expense of fuel and the improved applications of the steam engine.

About this time Smeaton applied himself with great activity and success to the improvement of wind and water mills, and succeeded in augmenting their useful effect in a twofold proportion with the same supply of water. From the year 1750 until the year 1780 he was engaged in the construction of his improved water mills, which he erected in various parts of the country, and which were imitated so extensively that the improvement of such mills became general. In cases where a summer drought suspended the supply of water, horse machinery was provided, either to work the mill or to throw back the water. These improvements necessarily obstructed for a time the extension of steam power to mill work; but the increase of manufactures soon created a demand for power greatly exceeding what could be supplied by such limited means.

In the manufacture of iron, it is of great importance to keep the furnaces continually blown, so that the heat may never be abated by day or night. In the extensive ironworks at Colebrook Dale, several water-wheels were used in the different operations of the manufacture of iron, especially in driving the blowers of the iron furnaces. These wheels were usually driven by the water of a river, but in the summer months the supply became so short that it was insufficient to work them all. Steam engines were accordingly erected to

return the water for driving these wheels. This application of the engine as an occasional power for the supply of waterwheels having been found so effectual, returning engines were soon adopted as the permanent and regular means of supplying water-wheels. The first attempt of this kind is recorded to have been made by Mr. Oxley, in 1762, who constructed a machine to draw coals out of a pit at Hartley colliery, in Northumberland. It was originally intended to turn the machine by a continuous circular motion received from the beam of the engine; but that method not being successful, the engine was applied to raise water for a wheel by which the machine was worked. This engine was continued in use for several years, and though it was at length abandoned, on account of its defective construction, it nevertheless established the practicability of using steam power as a means of driving water wheels.*

(109.) In the year 1777, Mr. John Stewart read a paper before the Royal Society, describing a method for obtaining a continued circular motion for turning all kinds of mills from the reciprocating motion of a steam engine. He proposed to accomplish this by means of two endless chains passing over pulleys, which should be moved upwards and downwards by the motion of the engine, in the manner of a window sash. The joint pins of the links of the two chains worked in teeth at the opposite sides of a cog wheel, to which they imparted a circular motion, first by one chain, and then by the other, acting alternately on opposite sides of the wheel. One chain impelled it during the descent of the piston, and the other during the ascent; but one of these chains always passed over its pulleys so as to produce no effect on one side of the cog wheel, whilst the other chain worked on the opposite side to turn it round. For this purpose each chain was provided with a catch, to prevent its circulating over its pulleys in one direction, but to allow it free motion in the other. The cog wheel thus kept in revolution might be applied to the axis of any mill which the engine was required to work. Thus, if it were applied to a flour-mill, the millstone itself would perform the office of a fly-wheel to regulate the intermission of

* Farey on the Steam Engine, p. 297.

the power, and in other mills a fly-wheel might be added for this purpose.

The hints obtained by Mr. Stewart from Papin's contrivance, before mentioned, will not fail to be perceived. In Mr. Stewart's paper he notices indirectly the method of obtaining a continued circular motion from a reciprocating motion by means of a crank or winch, which, he says, occurs naturally in theory, but in practice would be impossible, from the nature of the motion of the engine, which depends on the force of the steam, and cannot be ascertained in its length. Therefore, on the first variation, the machine would be either broken in pieces or turned back. Such an opinion, pronounced by a man of considerable mechanical knowledge and ingenuity, against a contrivance which, as will presently appear, proved in practice, not less than in theory, to be the most effectual means of accomplishing the end here pronounced to be impossible, is sufficiently remarkable. It might cast some doubt on the extent of Mr. Stewart's practical knowledge, if it did not happen to be in accordance with a judgment so generally unimpeachable as that of Mr. Smeaton. This paper of Mr. Stewart's was referred by the council of the Royal Society to Mr. Smeaton, who remarked upon the difficulty arising from the absolute stopping of the whole mass of moving power, whenever the direction of the motion is changed; and observed, that although a fly-wheel might be applied to regulate the motion, it must be such a large one as would not be readily controlled by the engine itself; and he considered that the use of such a fly-wheel would be a greater incumbrance to a mill than a water-wheel to be supplied by water pumped up by the engine. This engineer, illustrious as he was, not only fell into the error of Mr. Stewart in respect of the crank, but committed the further blunder of condemning the very expedient which has since rendered the crank effectual. It will presently appear that the combination of the crank and fly-wheel have been the chief means of establishing the dominion of the steam engine over manufactures.

(110.) In 1779, Mr. Matthew Wasbrough, an engineer at Bristol, took out a patent for the application of a steam engine