be the furnace heating the boiler b; c a pipe conveying the steam to the interior of a cylinder d, the upper end of which is open to the air, in which a piston e works; this is attached to a lever ƒ oscillating in the centre g, and having at its other extremity h a weight, as a pump-rod weighted, which, acting as a counterpoise, pulls the piston up to the top of the cylinder. Now supposing steam to be introduced to the interior of the cylinder, and thereafter to be condensed by throwing cold water over the exterior surface, a vacuum will be produced in the cylinder; and the atmosphere pressing on the upper side of the piston will force it downwards; and pulling down the end ƒ of the great beam, will raise the end h, and along with it the counterpoise-weight or pump-rod there attached. The power of this engine will obviously depend on the surface of piston, the atmosphere exerting a pressure of about 14.75 lbs. on each square inch. Supposing the piston to be 50 inches in area, the weight which the engine would be able to lift would be nearly 800 lbs. This is the theoretical view of the case; the practical one widely differs. The effect obtained, as above noticed, depends on the absence of friction and the perfect formation of a vacuum. None of these desiderata, in the earlier stages of the engine, could be obtained. To reduce the friction of the piston moving in the cylinder, the piston was provided with " packing" placed round its edges, and made of hemp or leather well-lubricated. The friction, however, of the piston and working-beam was of so considerable an amount, that it detracted much from the working capabilities of the engine. The bad formation of a vacuum also reduced its working-power. When the void was imperfect, the vapour remaining in the cylinder resisted the atmospheric pressure in proportion to its temperature; this being tantamount to reducing the weight which could be lifted at the end of the beam. The fall of the cylinder was also prevented to a certain extent by air, which entered into the cylinder along with the steam from the boiler; this air not being condensed by the cold water, remained in the interior of the cylinder, and operated as an opposing power to the descent of the piston in proportion to the amount. This defect, which was termed wind-logging, continued to increase in power with the operation of the engine; and unless means had been afterwards adopted, the air would have so increased in volume as to stop the motion of the engine. The form of engine introduced at first by Newcomen resembled in some points that of Papin; an essential difference between them will, however, be obvious on consideration. In Papin's engine the piston was raised by the force of the steam; it required therefore, to be of a pressure considerably above that of the atmosphere. In Newcomen's engine, however, the steam was used at the ordinary pressure, or 212°, and only as a means of producing a vacuum beneath the piston, and thus aiding the atmospheric pressure by the improved mechanical means. The substitution, moreover, of the beam and connecting-rods opened the way to a great number of applications. Hence it will be observed that the name, atmospheric engine," by which Newcomen's contrivance was known, was not so inappropriate.

66

We now proceed to detail the various improvements introduced from time to time, which resulted in bringing the atmospheric engine to comparative perfection. And first, as to the method of forming a vacuum in the cylinder. As in the engine already described, the cold water, to produce the vacuum, was originally thrown over the exterior of the cylinder. This

plan was productive of much loss of working effect, from the circumstance that the boiler, being placed immediately beneath the cylinder, received the plashings of the cold water on its surface; this happening at every stroke of the engine, tended to condense the steam in the boiler. To obviate this, the cylinder was surrounded with an outer case, leaving a space between it and the interior cylinder. Into the space thus formed the cold water was introduced. A practical inconvenience was soon discovered to attend this plan, as the water soon became heated, and therefore comparatively useless in producing a vacuum in the interior of the cylinder. Again, it was necessary that, during the time when the cylinder was being filled with steam, the water in the space should be nearly of the same temperature as the steam. Means, therefore, were adopted by which this hot water was quickly withdrawn from the space as soon as the cylinder was filled with steam, and also for refilling the space as quickly with cold water, to produce the vacuum. These desiderata were effected by supplying the cold water from a cistern placed immediately above the cylinder, and by leading a pipe from the space between the cylinder and its casing to a small reservoir. The heated water thus obtained was conveyed to the boiler, compensating in some measure for the loss of power from causes already named. It is evident that if no means were taken to prevent it, the successive condensations of the steam in the interior of the cylinder would produce water, which in time would fill it. To remove this water, a pipe was inserted in the bottom of the cylinder, and conducted to a distance of at least thirty feet below it. It was necessary to take this so far down to counteract the force of the atmospheric pressure; as the pipe communicated with a vacuum, the water in the cistern, with which the eduction-pipe was connected, would be forced into the cylinder, unless the column of water in the tube or eduction-pipe was of sufficient length.

The air which found its way to the interior of the cylinder was ejected into the atmosphere through a pipe furnished with what was called a snifting-valve, opening upwards into a kind of cup containing water; the piston descending forced the air through this, the water surrounding it keeping it tight after the air escaped. The piston was rendered tight, and air prevented from finding its way to the interior of the cylinder, by a small quantity of water placed on its upper side, this supplied by a small pipe leading from the cistern above the cylinder. The pipe leading from the boiler to the cylinder was furnished with a cock, by which the supply of steam was regulated.

The operation of the engine, as thus constructed, is easily understood. On the steam being raised at a temperature of 212°, and the cock on the steam-pipe opened, the counterpoise-suppose this to be the weighted rod of a pump for withdrawing water from a mine-draws the piston to the top of the cylinder; the supply of steam is then shut off by turning the cock on the steam-pipe to its original position. The cold water is now admitted to the space between the cylinder and its casing, the steam under the piston being condensed; the atmospheric pressure on the upper side of the piston forces it downwards, drawing along with it the end of the beam, and raising the other, and also the water in the pump; the cold-water cock is then shut, and the steam opened. An equilibrium being thus restored on both sides of the piston, the counterpoise draws it up to the top of the cylinder; the cock of the pipe for withdrawing the hot water from

the space around the cylinder is opened, and the water descends to the cistern; the cold-water cock is immediately opened, and the condensation being effected, the piston descends as before.

In the spring of 17.12 Newcomen succeeded in obtaining a contract for drawing water from a mine at Wolverhampton. The account given by Desaguliers of the difficulties encountered in bringing this engine to work is very curious: "After a great many laborious attempts having been made, he at last made the engine work; but not being philosophers enough to understand the reasons, or mathematicians enough to calculate the powers and proportions of the parts, they very luckily found by accident what they sought for. They were at a loss for the pumps; but being so near Birmingham, and having the assistance of so many admirable and ingenious workmen, they soon came to the method of making the pumps, valves, clacks, and buckets, whereas they had but an imperfect notion of them before." The erection of this engine was the occasion of an improvement of great importance being accidentally discovered. The improvement consisted in a quicker means of obtaining a vacuum. As before described, the piston was kept tight by water playing on its upper surface; on the first trial of this engine, it made several strokes in quick succession. "After a search, they found a hole in the piston, which let the cold water in to condense the steam in the inside of the cylinder. The method of effecting the condensation was therefore changed; and effected henceforth by injecting cold water into the interior of the cylinder. The diagram in fig. 14 shows this arrangement: a a the cylinder, b the piston, c a pipe leading from the coldwater cistern, and provided with a cock to regulate the supply; the lower extremity of this pipe is inserted in the bottom of the cylinder, and the water is delivered in the form of a jet d, and, diffusing itself among the steam, the condensation is quickly effected; e the snifting-valve and pipe; f the pipe 30 feet long, for taking away the hot-water from the cylinder. This plan of obtaining a quick condensation suggested also a means of regulating the speed of the engine in cases where the weight to be lifted was variable, by throwing in a greater or less quantity of injection-water, thus producing a vacuum more or less perfect. Notwithstanding the very great improvements thus effected from time to time, the atmospheric engine at this stage of its progress was much restricted in its usefulness, and this from the unremitting attention which its operation demanded from the attendant. “When, for instance, the attendant opened the steam-cock, he was obliged to watch the descent of the piston, and at the instant when it was elevated to the proper height, it was to be again quickly shut, and at the same moment the injection-cock was to be opened. If the one did not follow the other, there resulted a great loss of vapour, or of effect; and this difficulty was further increased by the irregular production of the steam itself, from the varying intensity of the heat of the furnace.

Пе

и

b

fig. 14.

d

a

After the injection had condensed the steam, and the piston was at liberty to descend, if the communication between the boiler and the cylinder was not opened at the precise instant when it had reached the limit of its downward movement, the immense weight on the piston, falling into the vacuum with a great velocity, would shake the apparatus to pieces. All this precision was required too from a mercenary attendant fourteen times every minute, at the hazard of the total destruction of the apparatus." It is obvious, then, that the further and more extended practical introduction of the engine depended on some method being discovered of making its movements self-acting. According to Desaguliers the honour of the invention of the self-acting movements is due to an idle boy of the name of Humphrey Potter, an attendant on the engine. The following is the statement: "It was usual to work with a buoy in the cylinder, enclosed in a pipe, which buoy rose when the steam was strong, and opened the injection-pipe, and made a stroke, whereby they were only able, from this imperfect mechanism, to make six or eight strokes in a minute, till a boy named Humphrey Potter, who attended the engine, added what is called a scoggan, a catch that the beam or lever always opened; and then it would go fifteen or sixteen strokes in a minute." "To scog," says a writer, in explanation of the term, “is a verb found in certain vocabularies throughout the north of England, implying to sculk; and this young gentleman, impelled by a love of idleness or play common to boyhood, and having his wits about him, after some meditation, devised this contrivance, by which so important an improvement was effected, and himself allowed the means of scogging for his own diversion." Whether this is the correct account of the origin of this improvement, it is difficult to ascertain; certainly the statement of Desaguliers has never been proved to be wrong. In the year 1718 an engine was erected having self-acting movements, termed "hand-gear," the invention of Mr. Henry Beighton, an engineer of Newcastle, and which consisted of a series of tappets operated on by the beam, and by which the various cocks were opened and shut as required. The following is a description of the means employed for this purpose: The entrance to the steam-pipe was covered with a sliding-valve placed inside the boiler; this valve was worked by a lever attached to the spindle of the valve which projected through the top of the boiler, as in fig. 15, where a is the steam-pipe, b the sliding-valve, the spindle of which passes through c, and is worked by the bent lever d. To this bent lever a horizontal one e, fig. 16, is connected; the other extremity is formed into a fork at f; a spindle joins the two extremities of the fork; and two stirrups, as at g, connected each side of the fork with an axle rotating on the centre h. This axle was made to move by the pins in the beam n n striking the ends of the spanners km, which were firmly fixed in the axle n. On the axle, at a position between the two prongs, a lever called a tumbling-bob was fixed, having a Y-end, or two projecting arms ii, and a weight at the other. The injection-cock was opened and shut by the mechanism shown in fig. 17. Let a a be the beam corresponding to nn, in fig. 16, having projecting-pins which strike the end of the lever d, terminated with a toothed quadrant taking into a second quadrant, fixed on the spindle of the injection-cock. The following is the operation of the apparatus: on the beam n n falling, a tappet, or projectingpin, strikes one of the spanners k; this turns the axle vibrating at n, and causes the tumbling-bob to fall over with considerable force, one of the

arms i striking the bar which joins the two prongs of the fork eƒ; this pulls forward the lever e towards the beam, and opens the steam-cock b,

fig. 16. Simultaneously with the striking of the tappet on the spanner k, another tappet strikes the end of the lever d, fig. 15, and operating on the quadrants fg, shuts the injection-cock h. The piston is now drawn upwards by the counterpoise. On arriving within a short distance of the top of the cylinder, a tappet strikes the spanner m, fig. 16, and causes the tumbling-bob to fall over, moving, as before, the lever e from the beam; one steam-valve is thus shut, and by means of the tappets, lever, and quadrants, the injection-cock h, fig. 17, is opened. By means of this contrivance the atmospheric engine was rendered self-acting. Thus improved, the machine remained for a considerable period in statu quo. Minor improvements were from time to time introduced; but it was reserved for the celebrated engineer John Smeaton to bring it to as great a state of perfection as was possibly allowed by the nature of its principle. In fig. 18 we give a diagram illustrative of the general arrangements of an atmospheric engine, after the introduction of the self-acting movements of Beighton, and anterior to the improvements of Smeaton.

In 1767, Smeaton was employed to construct an engine for the New River Company, and he availed himself of an opportunity of introducing several improvements. In calculating the proportions, on considering that the stoppage of water occurring at every stroke, and putting the piston, beam, and other appliances, from a state of rest to that of motion twice every stroke, resulted in a great loss of power, he determined to work the engine slower, putting on the piston all the load it would bear, working with larger pumps. In order still further to reduce the velocity of the column of water in the pump-barrel, he made the beam oscillate on a centre out of its true centre; the stroke of the piston being then nine feet, whilst that of the pump, which lifted thirty feet, was only six feet. This arrangement necessitated the employment of a long narrow cylinder, eighteen inches in diameter: with these arrangements he increased the load of the piston from seven pounds to ten pounds and a half on the inch. From