the inside of the cylinder, their upper and under sides are also ground perfectly flat, and applied one upon the other. On the upper ring is placed a plate of metal, rather smaller in diameter than the cylinder, while a similar flat plate is placed below the under ring, both of which, with the rings between, are attached loosely to each other by means of the piston-rod passing through them.

A shell being thus formed, the rings are each of them cut into three pieces, and in cutting them, such a portion of the metal is taken away as to leave room to introduce between two of the pieces, a spring in form of the letter V, the open end of which is placed outwards, almost close to the circumference; by which means the two pieces against which the two sides of the spring act, are pressed in the direction of the circumference, against the ends of the third piece, so that the three pieces are thus kept so uniformly in contact with the cylinder, that the longer the machine is worked, the better the rings must fit. To prevent steam passing through the cuts in the lower rings, the solid parts in those upon the upper side, are made to fall upon the divisions and springs of the under ones, thus interrupting the communication that would otherwise remain open, and forming a perfect break-joint.

.

A diagram of Mr. Barton's piston is annexed: and it

will be seen that the flexible springs t t t operate on the

wedges ccc and as such must expand the remaining portions, rp, so as to ensure the accurate fitting of the piston. Mr. Perkin's piston will be examined in a future page. The interior surface of the cylinder in which the piston works, requires to be bored with the greatest exactness, though this was but little attended to in the early atmospheric engines, some of them being composed of timber hooped together in the same manner as barrels are constructed. Mr. Watt, in his first attempts at improving the steam engine, employed this material in the construction of his cylinders, though he afterwards abandoned it for those of bored metal; the operation of boring being performed with the greatest precision, by an apparatus invented by Mr. Wilkinson.*

Mr. Murray has also effected considerable improvements in this part of the engine, and the boring machines employed in his manufactory are of considerable value. They are worked by a separate steam engine, which is never stopped during the operation, as in that case a shoulder or ring would be formed, running completely round the cylinder.

In small engines, it is common to place the cylinder within the boiler, in which case no artificial mode of retaining the heat is required; but to this arrangement in those of larger dimensions there are several objections, not the least of which is the frequent repairs that are necessary in the boiler; and a similar effect has been produced by the use of a double cylinder. This was first adopted by Messrs. Boulton and Watt, the outer cylinder or steamjacket keeping the inner cylinder at the temperature of boiling water, by the action of a partition of steam made to pass between the jacket and the working cylinder.

* For a description of Mr. Wilkinson's patent cylinder apparatus, see Appendix, A.

We have already stated, that Mr. Watt's great improvement consisted in condensing the steam in a separate vessel where a vacuum was formed by the continued application of cold water. A metal box constructed for this purpose, and furnished with a pump for drawing off the water and air, is called a condenser. It is necessary that the parts appropriated to this purpose should be kept as cold as possible; and upon this account the air-pump and condenser are placed in a cistern of cold water, which is kept full by the continued action of a pump, also worked by the engine, and called the cold water pump, a little being allowed to pass off continually to preserve the water at an equable temperature.

The air-pump and condenser are usually of the same size; if of one-eighth, the capacity of the working cylinder, it will be found sufficient to keep the condenser empty in Mr. Watt's single engine. The best proportion for a double-action engine is about two-thirds the diameter of the cylinder and half the length of stroke, the condenser, as in the single engine, being of similar capacity.

In Mr. Maudslay's portable engine the condenser is a hollow cylinder, and the air-pump is placed within it, so that there is no necessity for a pipe of communication from the air-pump to the condenser; and in this case a small cistern is fixed over the pump to contain the hot water, the discharge-valves being placed in the lid, which thus forms the bottom of the cistern or hot well.

In the early engines, on Messrs. Boulton and Watt's construction, the air-pump and condensing-cistern were placed at the outer end of the beam; in which case the pump-bucket being drawn up by the descent of the piston, the engine required less counter-weight than in the present form, in which the air-pump must be wholly worked by the counter-weight. It was necessary also, that the

parts appropriated to the condensation of steam should be kept as cold as possible; on which account, the air-pump and condenser were placed in a cistern of cold water, which being continually on the overflow, carried off the excess of heat in the manner already described.

The mode of condensing by outward cold, was not however found sufficient; and Mr. Watt afterwards introduced a small jet of water, the dimensions of the air-pump being so far increased, as to extract the injection-water as well as the air.

To shew the degree of vacuum in the condenser, and consequently the amount of pressure on the piston, a barometer-gauge has been employed. This is justly considered as a most important instrument, though unfortunately for the profit of steam-engine proprietors but little attended to. This gauge is in fact a common barometer tube, of thirty inches in length, with a graduated scale, and connected with the condenser by a small tube furnished with a stop-cock. When the air is expelled from the cylinder this must be closed, otherwise the steam entering the tube would blow the mercury from the cup. On the cock being turned, and the communication opened with the condenser, the exact degree of vacuum will be shewn by the height of the mercurial column, which, if the condensation be not complete, or air be admitted, will descend, and on the contrary, if perfect, it will ascend, as in the Torricellian tube.

The steam-gauge employed by Mr. Watt, consists of an inverted syphon or bent tube of glass or iron, one leg of which is jointed to the steam-pipe, while the other is open to the atmosphere. A quantity of mercury being poured into the tube, it will occupy the lower or bent part, and the surface of the fluid metal in one leg being exposed to the pressure of the steam, while the external air acts upon the other, it is evident that the difference of level of the

two surfaces will express the pressure of the steam in the height of the mercurial column passing up the graduated tube.

This gauge is just the reverse of the preceding; the barometer shewing the pressure of the atmosphere on a given space of the piston, while the steam-gauge indicates the force of elastic vapour entering from the boiler. It is the duty of the fireman frequently to look at this gauge, that he may know when to increase the fire in the furnace, and thus a sufficient supply of steam will always be secured to the engine.

In the early atmospheric engines, the working-beam was composed of a large and almost unhewn tree; but Mr. Smeaton employed a framing of wood for this purpose, which was afterwards much simplified and improved by Mr. Hornblower.

In double-acting engines it is usual to have the beam cast in one piece, the extremities being turned in a lathe to form cylindrical pins, upon which are fitted sockets or pieces, having other pins projecting from them to form the points of the parallel motion and connecting rod. Thus, there is one pin on each side of the socket, the two links of the parallel motion being fitted to the two projecting pins at one end, while the double joint of the connecting rod is fitted on the two pins at the other end of the beam. The advantage of this construction is, that the joints at the ends of the beam become universal joints, having liberty of motion in all directions; and in some of Mr. Murray's best engines, the same contrivance is applied to the crank pin and connecting rod.

The annexed figure represents one of Boulton and Watt's parallel-motions, and is that now most commonly used by themselves, and other engine-makers. This has