part of each of the curved ends of the beam, and to the pump and piston rods, as represented in fig. 170, which answered the required purpose; but in double-acting engines, where the piston-rod forces upwards as well as pulls, some other mode of converting the action is required. The most perfect yet introduced is the parallel motion, the principle of which may be comprehended by referring to fig. 191.

Suppose A B to represent one end of a beam, vibrating on its centre at A, the other end B will perform the arc C C', and carry whatever is attached to it in the direction of that arc; then suppose another rod, G H, of equal length to A B, vibrating on its centre G, the point in the connecting piece, L H, will, by the vibrations of A B and H G, move upwards and downwards in a perpendicular line. For so much as the curve from the radius A B draws it towards A, so much will the curve of the radius H G draw it towards G, and the movements, correcting each other, will cause I to rise and fall in a perpendicular line.

A more simple mode, which answers very well in small engines, is represented in fig. 192. The sling piece attached to the end of the beam has a pair of rollers, one on each side, which press on each side of the guiding bars D D D D, and carry the piston, attached at C, in a perpendicular

line.

All steam-engines are proportioned to go at a settled rate, and to make a certain number of strokes of the piston per minute, which, reckoning the number of feet the piston travels in its upward and downward motion in the cylinder, has been by general agreement settled at 200 to 220 feet per minute.

To obtain this regularity of action, it is manifest that a regularity must be obtained in the amount of power that is to produce the movement; or, in other words, a regular amount of elastic force must be exerted at each stroke of the piston. This is somewhat difficult to accomplish, and depends much upon keeping the fire, which generates the steam, of a uniform heat; consequently the care of the fire should only be intrusted to one who is well skilled in his business. There are, however, contrivances, called governors, which greatly assist in maintaining a regular action, and where very great nicety is not required, they produce it to a sufficient extent.

The governor, (which we have already described under the article "On the Equalization of Motion," in Mill-geering,) acts upon the principle of centrifugal force, and is applicable only to engines that have a rotative motion.

The mode of applying the governor is by connecting it by means of levers to a throttle valve, shown in fig. 193. A B represents a section of the pipe that conveys the steam from the boiler, having a small circular fan of iron, capable of being

turned so as to be parallel with the line of the pipe, or of lying across it in such a manner as to stop the communication.

The governor is placed at any convenient part of the engine, upon an upright spindle, and is driven either by a band or by bevil-wheels, from the fly-wheel shaft. When the fly-wheel increases or decreases in velocity it transmits the same to the governor, and causes the balls to fly from, or approach nearer to, each other; and the lever affixed to the collar of the governor turns the handle C of the throttle valve, which regulates the supply of steam, and produces regular motion.

In single-acting engines, which do not create rotative motion, this sort of governor cannot be applied, nor is it decidedly adapted to the sort of regulation they want; for as they act only one way, and the steam being entirely shut off at the end of every stroke, it is a regulation of the amount of power to be exerted at every interval which is sought, rather than a continuous equable supply of steam.

This sort is most wanted in engines that supply towns with water, as when the different districts are being supplied, the diversity of their magnitude and situation causes a considerable variation in the burden thrown upon the engine. In engines used for this purpose there are two pieces of wood, called spring-beams, placed across each end of the beam, which, when the burden is lightened, and the engine makes too long a stroke, strike, as they descend, the upper floor of the engine-house and ring a bell, to warn the attendant that steam must be shut off earlier, or that the upper tapit of the plug-tree must be moved downwards. In engines of recent construction, the spring-beams are made to strike a lever, which either shuts off the steam entirely, or opens a cock that admits it into the cylinder: in either case the movement of the engine ceases.

The regulation of the plug-tree tapit is obtained by introducing a pipe from the water in the air-vessel to a small cylinder, having a working piston. The water from the air-vessel forces against the under side of the piston and raises it and its rod, and by that motion the tapit of the plug-tree is raised or lowered as is required. This mode, from error in construction, or otherwise, has mostly proved inefficient when the burden of the engine is liable to great and rapid fluctuations.

Such is the general construction of engines now in use: but there are many ingenious contrivances in boilers which require special notice.

Fig. 194 represents a boiler fitted up with all the appendages now generally applied, and set in a furnace of a proper construction. Part of the furnace is shown in a sectional view. BBBB the boiler, the operation of which has been described. C, the steam-gauge, represented at large in fig. 195. Its object is to ascertain the pressure acting in the boiler. It is formed of a bent iron tube, at one end A communicating with the boiler, and at the other end B open to the atmosphere. The tube is filled up to C and D with mercury, and has a thin piece of stick, E, placed in the leg B, which floats perpendicularly in the mercury at D. To the leg B is appended a flat piece of brass, divided into inches, and numbered upwards, to form a scale. The stick is made of such length that the top of it shall be even with the first mark on the scale.

If the steam in the boiler presses against the mercury at C, and raises the surface, D, one inch, (which will be indicated by the end of the stick rising to 1 upon the scale,) it proves that there is one half pound pressing per square inch against the internal surface of the boiler, tending to burst it; for if the section of the bore of the pipe was just one superficial inch the pressure would be supporting one cubic inch of mercury, which will be found to weigh near half a pound; therefore for every two inches rise, one pound pressure may be reckoned, and as condensing engines seldom work with more than three or four pounds pressure upon the inch, the scale need not be longer than eight or nine inches.

C is a strong iron plate, covering a circular or oval hole of about 18 inches diameter, to admit a man into the boiler with a view to clean or examine it.

D is the steam pipe, containing the throttle valve E, to which the rod from the governor is connected. FF are gauge cocks. II is a feed pipe which passes into the boiler and reaches very near to the bottom. HHHH the cistern, on the top of the feed pipe; ii is a float, formed of stone, and balan ced so as to remain always on the surface of the water in the boiler. By the raising and falling of the water the float acts upon the lever K K1 by the wire 13, which passes through a steam-tight joint at I4, and as the water sinks, draws down the end K, which raises K', and the valve M attached to it. By this contrivance when the boiler requires a fresh supply of water the valve M opens and supplies it from the cistern H H H H.

The feed pipe II is made to contain a column of water equal to the amount of pressure exerted by the steam in the boiler, which we have already stated should not exceed the supporting of eight inches of mercury. As one inch of mercury being equal in weight to about 134 inches of water, the pipe should be about nine feet high from the surface of the water when the boiler is supplied; and the water in the feed pipe should stand about three feet when the pressure is six inches of mercury, or three pounds to the square inch of surface.

The feed pipe contains likewise an iron bucket weight O, hung by a chain that passes over two pullies, P P; to the other end of the chain is attached a sheet of iron called a damper. When the steam in the boiler is urged to too great an extent, it forces the water in the feed pipe upwards, and raises the

iron bucket weight O, which lowers the damper into the flue of the chimney, and checks the force of the fire.

S is a safety valve, loaded with a determinate weight, and of such a dimension in the bore, as will relieve the boiler of its pressure should it arrive be yond a certain temperature. It is enclosed in a case, to prevent the engine tender having access to it, as some engine tenders have been known to load the safety valves, to save themselves the trouble of attending to the fire with that diligence which is necessary, and by which they have endangered their own lives as well as the lives of others.

A pipe proceeds from this case to the chimney, to carry whatever steam may escape into the flue of the chimney.

There is very frequently another safety valve, which is open to the view of the engine tender, to indicate when the fire is too high.

TT is a flue formed of sheet iron, and passing lengthwise though the centre of the boiler so near to the bottom that it is always covered with water.

The flame and smoke from the fire at n n passes first under the boiler, and then immediately returns back through this flue, then dividing itself passes through flues which lead it on both sides the boiler to the chimney.

V is a cock for the purpose of emptying the boiler when required to be cleaned or repaired.

Such is the construction and general arrangement of the parts of Messrs. Boulton and Watt's engines both single and reciprocating. We shall now proceed in the examination of some other forms of engines which likewise condense their

steam.

Mr. Hornblower, conceiving he could obtain greater power from the complicated force of steam acting in two cylinders, obtained a patent, in 1781, for that object. His own account, taken from the specification, we here transcribe.

"First," says Mr. H. "I use two vessels, in which the steam is to act, and which in other engines are called cylinders. Secondly, I employ the steam after it has acted in the first vessel to operate a second time in the other, by permitting it to expand itself, which I do by connecting the vessels together, and forming proper channels and apertures, whereby the steam shall occasionally go in and out of the said vessels. Thirdly, I condense the steam, by causing it to pass in contact with metalline surfaces, while water is applied to the opposite side. Fourthly, to discharge the engine of the water used to condense the steam, I suspend a column of water in a tube or vessel constructed for that purpose, on the principles of the barometer, the upper end having open communication with the steam vessels, and the lower end being immersed in a vessel of water. Fifthly, to discharge the air which enters the steam vessels with the condensing water or otherwise, I introduce it into a separate vessel, whence it is protruded by the admission of steam. Sixthly, that the condensed vapour shall not remain in the steam