and screw-holes round its upper edge; the interior is supplied with some soft substance (hemp or cotton) to surround the piston rod; and upon the cover of the box is formed a ring or gland which fits into the box, and being screwed down upon the stuffing, presses it closely round the piston rod, and thereby prevents the escape of the steam from the cylinder. d is the working beam, resting upon its centre e, and connected at one end to the piston rod, at the other to the bar f. To the cylinder a (which, in this modification, has no external jacket) is attached a tube, through which the steam is allowed to pass above and below the piston, through the pipe Z, connected with the boiler; in this tube are placed valves, one above and one below the point of junction with Z, which are moved by external levers. Now, supposing the blowing-valve to have been opened, and the vacuum formed in the condenser in the manner before described, the steam rushing through the upper elbow of the tube upon the piston, forces it to descend to the bottom of the cylinder; by this action the lever 1 is turned downwards, by means of tappets placed on the pump rod 4, and shuts that valve; whilst 2, on a pipe behind that which we see, opens a passage to the condenser. The lever 3 is at the same time opened, admitting steam under the piston, which consequently ascends. 5 is a rod connected with the discharging-pump attached to the condenser; 6 a small pump, which supplies the boiler with heated water from the condenser.

The vertical descent of the piston rod (through the stuffing box) by its attachments to the beam, is explained in the article PARALLEL MOTION, and need not here be repeated.

Mr. Watt had employed the fly-wheel, in order to equalize the motion of the piston in the cylinder (see article FLY); but as it became an important object to convert the alternate motion into a rotary one, he applied himself to the discovery of a ready means of effecting it. The crank (see article CRANK) could have been appropriated to this purpose; but a patent was at that time in force for the exclusive adaptation of it by another; and as some dispute had arisen upon the subject, Watt was compelled to resort to some other mode, by which a similar effect might be achieved, without the invasion of another's privilege. This ended in the construction of what is now known by the name of the Susand-Planet wheels, the action of which is as follows :-the bar ƒ (an inflexible rod) is attached at one end to the working beam, and at the other to h, a toothed wheel that can revolve upon its axis; o is likewise a toothed wheel fixed to the fly. As the beam rises the planet wheel h is drawn up on the circumference of the sun wheel, and turns it round, causing the sun wheel to make two revolutions while the planet wheel travels once over its circumference; the momentum of the fly being sufficiently powerful to preserve the tendency of the machinery to revolve in the same direction during the change of motion in the piston, and to urge the planet wheel over the inactive points in its circuit, the continued rotary motion becomes at once effected, and with this advantage, that as the fly makes twice the number of revolutions it would make by the common crank, a lighter body of material composing the fly is required. There are, however, several disadvantages attending this mode of converting an alternate into a rotary motion, such as being more complex, expensive, and liable to derangement.

One of the last improvements made by Watt upon his engines was the application of a fine piece of mechanism which had been previously used for other objects-the governor or regulator. The invention has been ascribed, but improperly, to Watt; and although it is said that the notion of applying it to steam engines was first suggested by a Mr. Clarke of Manchester, it does not appear that he ever carried it into practice. The governor s is composed of two balls, fixed each to a lever attached to other and shorter levers, u u, above the point of junction, the latter being fixed by a moveable joint to a slider, w, moving freely on the vertical rod s. The horizontal lever w H has a fulcrum, and raises or lowers another lever H Z, which is attached to a valve inside the steam-pipe at Z. On the pulley is a cord q proceeding from the fly-wheel; by this means a rotary motion is given to the vertical rod, and the balls by their centrifugal force (vide article) rising outwards, draw downwards the slider; which movement

641

raises the opposite end of the horizontal lever H, which acting on the lever connected to it, opens or shuts (as it may be adjusted) the valve Z inside the

steam pipe, and diminishes or enlarges the area by which the steam flows into the cylinder. The fall of the balls when the motion decreases, reverses all

VOL. II.

4 M

these movements, of course; and by thus enlarging or contracting the steamway, and admitting more or less steam into the cylinder, the impulse of the piston is rendered much more uniform. The valve in this part of the steampipe is now called the throttle valve, and the regulating pendulum the governor An important improvement of Mr. Watt's was carried into practice in 1778. It consists in shutting off the steam from the cylinder, some time before the piston has completed its stroke, so that the remainder may be performed by the expansion of the steam already contained in the cylinder. This serves as a method of regulating the acting force of the engine, because, as the steam can be shut off at any part of the ascending or descending stroke, so much steam may be admitted as barely to carry the piston through its required motion, and by the adjustment of the valve gear, the quantity of steam admitted may at all times be varied in an instant. If this were the only advantage, it is a great one; but it will be seen that a great saving of fuel will likewise be effected by this method.

h

We shall endeavour to explain this more clearly by the aid of a diagram. The pressure of steam, as ascertained by numerous and carefully conducted experiments, is so nearly in direct proportion to its density, that for practical purposes it may be assumed to be in that ratio; and its density is of course inversely as the space occupied by it. Having premised thus much, let a 1 ch represent a cylinder into which the steam flows during the whole period of the stroke; at the end of which, or when the piston arrives at 1 c, the steam is discharged from the cylinder; then the effect being as the pressure of the steam multiplied by the space through which it acts, which pressure is in this case the same throughout the stroke, if the line a h represents the pressure, and a 1 the space through which it acts, the area of the rectangle a 1 ch will represent the effect.

g

Now let a bfh represent another cylinder of the same diameter as the former, but four times as long, and let it be supposed that only the same quantity of steam is admitted as before, the steam being cut off when the piston has reached to the position c 1, which is only of the stroke in the cylinder ab; the piston will in this case continue to descend, but the pressure upon it will gradually decrease; and when the piston has made half its stroke, as the original volume of steam a 1 will then be expanded into double its bulk, or occupy the space a 2; its density, and consequently its pressure, will only be half as great as at the moment at which it was cut off, or when the piston was in the position c 1; therefore if the line c 1 represent the full pressure of the steam, the line d 2 will represent its pressure at the position 2. At the end of the stroke, or when the piston arrives at fb, the steam will be expanded into four times its original volume; consequently its pressure will then be only one fourth of its original pressure, and will be represented by the line be.

In this case the total effect of the steam will be represented by the area heebla which exceeds the effect of the steam in the first supposed case, by the area c deb 1, which therefore represents the increase of power due to the expansive action of the steam.

The curve e de represents the ratio in which the pressure of the steam decreases as the piston descends in the cylinder, and if we deduct the area included by the points c d e f c, from the area of the rectangle c 1 bf, we shall obtain the area of the figure c e b 1, which represents the effect of the expansive action. Now, as the curve c d e differs but little from a parabola, the mean pressure may be readily computed by the following rule. It is only an approximation, but it is sufficiently near the truth for all practical purposes, so long as the steam does not expand to more than four times its original volume, and within those limits gives a result rather below the true one.

RULE To the pressure of the steam, above the pressure of the atmosphere, add 15 lbs per square inch for the atmospheric pressure, and call the sum the

total pressure of the steam; square the fraction of the stroke during which the steam acts expansively, and deduct of the product from unity, and the remainder, multiplied by the total pressure, will be the mean pressure, after deducting 15 lbs for the pressure of the atmosphere in the case of non-condensing engines.

Example in a non-condensing engine; suppose the pressure of the steam, at the commencement of the stroke, to be 45lbs per square inch above the atmosphere, and that the steam is cut off at of the stroke: required, the mean pressure of the steam.

The mechanical effect being as the pressure of the steam, and the space through which such pressure is exerted, the effect of the steam whilst acting at full pressure in the above example, will be as 45lbs through a space 145, but the total effect will be as 21-375lbs through a space 4, which is equal to 85-5, or nearly double the former, and shows the great advantage to be derived from using steam expansively. There are, however, in practice certain limitations to the extent to which steam may be allowed to expand, for independently of the inconvenience of the great size of the cylinder when the expansion is carried to an extreme, the pressure of the steam upon the piston should never be less than the resistance from the friction, &c. added to the pressure of the atmosphere in non-condensing engines, or the pressure of the non-condensed steam in condensing engines; but in the preceding example, the pressure at the end of the stroke is merely equal to that of the atmosphere. Mr. Tredgold gives the following rule for ascertaining the point at which the steam should be cut off, so as to produce the greatest effect:

Divide the amount of the friction, &c. added to the pressure of the atmosphere, in non-condensing engines, or of the non-condensed steam of condensing engines, by the pressure of the steam in the boiler, and the quotient will give the proportion of the stroke at which the steam should be cut off.

Example. The pressure of steam in the boiler being equal to 120 inches of mercury, the loss from friction, &c. (reckoned as of the whole) is 48, which added to 30 inches for the pressure of the atmosphere is 78, and this divided

by 120 gives as the part of the stroke at which the steam should be cut off.

The principle of expansion was subsequently adopted by Hornblower, who in 1781 obtained a patent for an expansion engine, arranged as exhibited in the accompanying cut, of which the following is a description extracted from the Encyclopædia Britannica.

Let A and B represent two cylinders, of which A is the largest ; a piston moves in each, having their rods C and D moving through collars at E and F. These cylinders may be supplied with steam from the boiler by means of the square pipe G, which has a flange to connect it with the rest of the steam-pipe. This square part is represented as branching off to both cylinders; c and d are two cocks which have handles and tumblers as usual, worked by the plug-beam W. On the fore side of the cylinders is represented another communicating pipe, whose section is also square, or rectangular, having also two cocks, a b. The pipe Y immediately under the cock b establishes a communication between the upper and lower parts of the cylinder B, by opening the cock b. There is a similar pipe on the other side of the cylinder A, immediately under the cock a When the cocks c and a are open, and the cocks b and d are shut, the steam from the boiler has free admission into the upper part of the small cylinder B, and the steam from the lower part of B has free admission into the upper part of the great cylinder A; but the upper part of each cylinder has no communication with its lower part. From the bottom of the great cylinder proceeds the eduction pipe K, having a valve at its opening into the cylinder; it then bends