Chapter I (9)

If such an arrangement of mechanism as has been here described can be connected with the beam of the steam engine, so that while the point _x′_ is attached to the top of the steam piston, and the space through which it ascends and descends shall be equal to the length of the stroke of that piston, the point _x_ shall be attached to the rod of the air-pump piston, the stroke of the latter being half that of the steam piston, then the points _x′_ and _x_ will guide the motion of the two pistons so as to preserve them in true vertical straight lines.

The manner in which these ideas are reduced to practice admits of easy explanation: let the point O be the centre of the great working beam, and let O P′ be the arm of the beam on the side of the steam cylinder. Let P be a pivot upon the beam, at the middle point between its centre O and its extremity P′; and let the links P p, P′ _x′_, and P _p_ be jointed together, as already described. Let the point or pivot _o_ be attached to some part of the fixed framing of the engine or engine house, and let the rod _o p_, equal to half the arm of the beam, be attached by a pivot to the corner of the parallelogram at [Pg201] _p_. Let the end of the steam piston-rod be attached to the corner of the parallelogram _x′_, and let the end of the air-pump be attached to the middle point _x_ of the link P _p_; by which arrangement it is evident that the rectilinear motion of the two piston-rods will be rendered compatible with the alternate circular motions of the points P′ and P on the beam.

Among the many mechanical inventions produced by the fertile genius of Watt, there is none which has excited such universal, such unqualified, and such merited admiration as that of the parallel motion. It is indeed impossible, even for an eye unaccustomed to view mechanical combinations, to behold the beam of a steam engine moving the pistons, through the instrumentality of the parallel motion, without an instinctive feeling of pleasure at the unexpected fulfilment of an end by means having so little apparent connection with it. When this feeling was expressed to Watt himself, by those who first beheld the performance of this exquisite mechanism, he exclaimed with his usual vivacity, that he himself, when he first beheld his own contrivance in action, was affected by the same sense of pleasure and surprise at its regularity and precision. He said, that he received from it the same species of enjoyment that usually accompanies the first view of the successful invention of another person.

"Among the parts composing the steam engine, you have doubtless," says M. Arago, "observed a certain articulated parallelogram. At each ascent and descent of the piston, its angles open and close with the sweetness—I had almost said with the grace—which charms you in the gestures of a consummate actor. Follow with your eye alternately the progress of its successive changes, and you will find them subject to the most curious geometrical conditions. You will see, that of the four angles of the jointed parallelogram, three describe circular arches, but the fourth which holds the piston-rod is moved nearly in a straight line. The immense utility of this result strikes mechanicians with even less force than the simplicity of the means by which Watt has attained it."

The parallel motion, of which there are several other varieties, depending, however, generally upon the same [Pg202] principle, formed part of a patent which Mr. Watt obtained in the year 1784, another part of which patent was for a locomotive engine, by which a carriage was to be propelled on a road. In a letter to Mr. Smeaton dated 22d October, in the same year, Watt says,—

"I have lately contrived several methods of getting entirely rid of all the chains and circular arches about the great levers of steam engines, and nevertheless making the piston-rods ascend and descend perpendicularly, without any sliding motions or right-lined guides, merely by combinations of motions about centres; and with this further advantage, that they answer equally well to push upwards as to pull downwards, so that this method is applicable to our double engines which act both in the ascent and descent of their pistons.

"A rotative engine of this species with the new motion which is now at work in our manufactory (but must be sent away very soon) answers admirably. It has cost much brain work to contrive proper working gear for these double engines, but I have at last done it tolerably well, by means of the circular valves, placed in an inverted position, so as to be opened by the force of the steam; and they are kept shut by the working gear. We have erected an engine at Messrs. Goodwyne and Co.'s brewery, East Smithfield, London."

(121.) By the contrivance which has been explained above, the force of the piston in ascending and descending would be conveyed to the working end of the beam; and the next problem which Watt had to solve was, to produce by the force exerted by the working end of the beam in ascending and descending a continuous motion of rotation. In the first instance he proposed to accomplish this by a crank placed upon the axle to which rotation was to be imparted, and driven by a rod connecting it with the working end of the beam. Let K (_fig._ 38.) be the centre, to which motion is to be imparted by the working end H of the beam. On the axle K suppose a short lever K I to be fixed so that when K I is turned round the centre K, the axle must turn with it. Let an iron rod, the weight of which shall balance the piston and piston-rod at the other end of the beam, be connected by joints with the working end H of the beam, and the extremity I of the [Pg203] lever K I. As the end H of the beam is moved upwards and downwards, the lever K I will be turned round the centre K, taking successively the positions represented by faint lines in the figure; and thus a motion of continued rotation will be imparted to the axle K.

This simple and effectual expedient of producing a continued rotatory motion by a crank was abandoned by Watt, as already explained, by reason of a patent having been obtained upon information of his experiments surreptitiously procured. To avoid litigation, he therefore substituted for the crank the sun and planet wheel already described; but at the expiration of the patent, which restricted the use of the crank, the sun and planet wheel was discontinued in Watt's engine, and the crank restored.

(122.) Whether the crank or the sun and planet wheel be used, there is still a difficulty in the maintenance of a regular motion of rotation. In the various positions which the crank and connecting rod assume throughout a complete revolution, there are two in which the moving power loses all influence in impelling the crank. These positions are those which the crank assumes when the piston is at the top and bottom of the [Pg204] cylinder, and is just about to change the direction of its motion. When the piston is at the bottom of the cylinder, the pivot I (_fig._ 38.), by which the connecting rod H I is attached to the end of the crank, is immediately over the axle K of the crank, and under the pivot H, which joins the upper end of the connecting rod with the beam. In fact, in this position the connecting rod and crank are in the same straight line, extending from the end of the beam to the axle of the crank. The steam, on entering the cylinder below the piston, and pressing it upwards, would produce a corresponding downward force on the connecting rod at H, which would be continued along the connecting rod and crank to the axle K. It is evident that such a force could have no tendency to turn the crank round, but would expend its whole energy in pressing the axle K downwards.

The other position in which the power loses its effect upon the crank is when the piston is at the top of the cylinder. In this case, the working end of the beam will be at the lowest point of its play, and the crank-pin I will be immediately below the axle K; so that K will be placed immediately between H and I. When the steam presses on the top of the piston, it will expend its force in drawing the end H of the connecting rod upwards, by which the crank-pin I will likewise be drawn upwards. It is evident that this force can have no effect in turning the crank round, but will expend its whole energy in producing an upward strain on the axle K.

If the crank were absolutely at rest in either of the positions above described, it is apparent that the engine could not be put in motion by the steam; but if the engine has been previously in motion, then the mass of matter forming the crank, and the axle on which the crank is formed, having already had a motion of rotation, will have a tendency to preserve the momentum it has received, and this tendency will be sufficient to throw the crank K I out of either of those critical positions which have been described. Having once escaped these dead points, then the connecting rod forming an angle, however obtuse or acute, with the crank, the pressure or pull upon the former will have a tendency to produce rotation in the latter. As the crank revolves, however, the influence [Pg205] of the connecting rod upon it will vary according to the angle formed by the connecting rod and crank. When that angle is a right angle, then the effect of the connecting rod on the crank is greatest, since the force upon it has the advantage of the whole leverage of the crank; but according as the angle formed by the crank and connecting rod becomes more or less acute or obtuse in the successive attitudes which they assume in the revolution of the crank, the influence of the connecting rod over the crank varies, changing from nothing at the two dead points already described, to the full effect produced in the two positions where they are at right angles. In consequence of this varying leverage, by which the force with which the connecting rod is driven by the steam is transmitted to the axle on which the crank revolves, a corresponding variation of speed would necessarily be produced in the motion imparted to the crank. The speed at the dead points would be least, being due altogether to the momentum already imparted to the revolving mass of the crank and axle; and it would gradually increase and be greatest at the points where the effect of the crank on the connecting rod is greatest. Although this change of speed would not affect the actual mechanical efficacy of the machine, and although the same quantity of steam would perform the same work at the varying velocity as it would do if the velocity were regulated, yet this variation of speed would be incompatible with the purposes to which it was now proposed that the steam engine should be applied in manufactures. In these a regular uniform motion should be imparted to the main axle.

(123.) One of the expedients which Watt proposed for the attainment of this end was, by placing two cranks on the same axle, in different positions, to be worked by different cylinders, so that while one crank should be at its dead points, the other should be in the attitude most favourable for its action. This expedient has since, as we shall see, been carried into effect in steam vessels; but one more simple and efficient presented itself in the use of a _fly-wheel_.

On the main axle driven by the crank Watt placed a large wheel of metal, as represented in _fig._ 43., called a _fly-wheel_. This wheel being well constructed, and nicely balanced on its [Pg206] axle, was subject to very little resistance from friction; any moving force which it would receive it would therefore retain, and would be ready to impart such moving force to the main axle whenever that axle ceased to be driven by the power. When the crank, therefore, is in those positions in which the action of the power upon it is most efficient, a portion of the energy of the power is expended in increasing the velocity of the mass of matter composing the fly-wheel. As the crank approaches the dead points, the effect of the moving power upon the axle and upon the crank is gradually enfeebled, and at these points vanishes altogether. The momentum which has been imparted to the fly-wheel then comes into play, and carries forward the axle and crank out of the dead points with a velocity very little less than that which it had when the crank was in the most favourable position for receiving the action of the moving power.

By this expedient, the motion of revolution received by the axle from the steam piston is subject to no other variation than just the amount of change of momentum in the great mass of the fly-wheel, which is sufficient to extricate the crank twice in every revolution from the mechanical dilemma to which its peculiar form exposes it; and this change of velocity may be reduced to as small an amount as can be requisite by giving the necessary weight and magnitude to the fly-wheel.

(124.) By such arrangements the motion imparted to the main axle K would be uniform, provided that the moving power of the engine be always proportionate to the load which it drives. But in the general application of the steam engine to manufactures it was evident that the amount of the resistance to which any given machine would be subject must be liable to variation. If, for example, the engine drive a cotton-mill, it will have to impart motion to all the spinning frames in that mill. The operation of one or more of these may from time to time be suspended, and the moving power would be relieved from a corresponding amount of resistance. If, under such circumstances, the energy of the moving power remained the same, the velocity with which the machines would be driven would be subject to variation, being increased whenever the operation of any portion of the machines usually [Pg207] driven by it is suspended; and, on the other hand, diminished when any increased number of machines are brought into operation. In fine, the speed would vary nearly in the inverse proportion of the load driven, increasing as the load is diminished, and _vice versâ_.

On the other hand, supposing that no change took place in the amount of the load driven by the engine, and that the same number of machines of whatever kind would have to be continually driven, the motion imparted to the main axle would still be subject to variation by the changes inevitable to the moving power. The piston of the engine being subject to an unvaried resistance, a uniform motion could only be imparted to it, by maintaining a corresponding uniformity in the impelling power. This would require a uniform supply of steam from the boiler, which would further imply a uniform rate of evaporation in the boiler, unless means were provided in the admission of steam from the boiler to the cylinder to prevent any excess of steam which might be produced in the boiler from reaching the cylinder.

This end was attained by a contrivance afterwards called the _throttle-valve_. An axis A B (_figs._ 39, 40.) was placed across the steam pipe in a ring of cast-iron D E, of proper thickness. On this axis was fastened a thin circular plate T, of nearly the same diameter as the steam pipe. On the outer end B of this axle was placed a short lever or handle B C, by which it could be turned. When the circular plate T was turned into such a position as to be at right angles to the length of the tube, it stopped the passage within the tube altogether, so that no steam could pass from the boiler to the engine. On the other hand, when the handle was turned through a fourth of a revolution from this position, then the circular plate T had its plane in the direction of the length of the tube, so that its edge would be presented towards the current of steam flowing from the boiler to the cylinder. In that position the passage within the tube [Pg208] would be necessarily unobstructed by the throttle-valve. In intermediate positions of the valve, as that represented in _figs._ 39, 40., the passage might be left more or less opened, so that steam from the boiler might be admitted to the cylinder in any regulated quantity according to the position given to the lever B C.

A view of the throttle-valve taken by a section across the steam pipe is exhibited in _fig._ 40., and a section of it through the axis of the steam pipe is represented in _fig._ 39. The form of the valve is such, that, if accurately constructed, the steam in passing from the boiler would have no effect by its pressure to alter any position which might be given to the valve; and any slight inaccuracy of form which might give a tendency to the steam to alter the position would be easily counteracted by the friction of the valve upon its axle. The latter might be regulated at pleasure.

By this expedient, however the evaporation of water in the boiler might vary within practical limits, the supply of steam to the cylinder would be rendered regular and uniform. If the boiler became too active, and produced more steam than was necessary to move the engine with its load at the requisite speed, then the throttle-valve was shifted so as to contract the passage and limit the supply of steam. If, on the other hand, the process of evaporation in the boiler was relaxed, then the throttle-valve was placed with its edge more directed towards the steam. Independently of the boiler, if the load on the engine was lightened, then the same supply of steam to the cylinder would unduly accelerate the motion. In this case, likewise, the partial closing of the throttle-valve would limit the supply of steam and regulate the motion; and if, on the other hand, the increase of load upon the engine rendered necessary an increased supply of steam, then the opening of the throttle-valve would accomplish the purpose. By these means, therefore, a uniform motion might be maintained, provided the vigilance of the engine-man was sufficient for the due management of the lever B C, and provided that the furnace under the boiler was kept in sufficient activity to supply the greatest amount of steam which would be necessary [Pg209] for the maintenance of a uniform motion with the throttle-valve fully opened.

(125.) Watt, however, soon perceived that the proper manipulation of the lever B C would be impracticable with any degree of vigilance and skill which could be obtained from the persons employed to attend the engine. He, therefore, adapted to this purpose a beautiful application of a piece of mechanism, which had been previously used in the regulation of mill-work, and which has since been well known by the name of the _Governor_, and has always been deservedly a subject of much admiration.

The governor is an apparatus by which the axle of the fly-wheel is made to regulate the throttle-valve, so that the moment that the axle begins to increase its velocity, it shifts the position of the throttle-valve, so as to limit the supply of steam from the boiler, and thereby to check the increase of speed. And on the other hand, whenever the velocity of the axle is diminished, the lever B C is moved in the contrary direction, so as to open more fully the passage for the steam, and accelerate the motion of the engine.

A small grooved wheel A B (_fig._ 41.) is attached to a vertical spindle supported in pivots or sockets C and D, in which it is capable of revolving. An endless cord works in the groove A B, and is carried over proper pulleys to the axle of the fly-wheel, where it likewise works in a groove. When this cord is properly tightened the motion of the fly-wheel will give motion to the wheel A B, so that the velocity of the one will be subject to all the changes incidental to the velocity of the other. By this means the speed of the grooved wheel A B may be considered as representing the speed of the fly-wheel, and of the machinery which the axle of the fly-wheel drives.

It is evident that the same end might be attained by substituting for the grooved wheel A B a toothed wheel, which might be connected by other toothed wheels, and proper shafts, and axles with the axle of the fly-wheel.

A ring or collar E is placed on the upright spindle, so as to be capable of moving freely upwards and downwards. To this ring are attached by pivots two short levers, E F, the [Pg210] pivots or joints at E allowing these levers to play upon them. At F these levers are joined by pivots to other levers F G, which cross each other at H, where an axle or pin passes through them, and attaches them to the upright spindle C D. These intersecting levers are capable, however, of playing on this axle or pin H. To the ends G of these levers are attached two heavy balls of metal I. The levers F G pass through slits in a metallic arch attached to the upright spindle, so as to be capable of revolving upon it. If the balls I are drawn outwards from the vertical axis, it is evident that the ends F of the levers will be drawn down, and therefore the pivots E likewise drawn down. In fact, the angles E F H will become more acute, and the angle F E F more obtuse. By these means the sliding ring E will be drawn down. To this sliding ring E, and immediately above it, is attached a grooved collar, which slides on the vertical spindle upwards and downwards with the ring E. In the grooved collar are inserted the prongs of a fork K, formed at the end of the lever K L, the fulcrum or pivot of the lever being at L. By this arrangement, when the divergence of the balls I causes the collar E to be drawn down, the fork K, whose prongs are inserted in the groove of that collar, is likewise drawn down; and, on the other hand, when, by reason of the balls I falling towards the [Pg211] vertical spindle, the collar E is raised, the fork K is likewise raised.

The ascent and descent of the fork K necessarily produce a contrary motion in the other end N of the lever. This end is connected by a rod, or system of rods, with the end M of the short lever which works the throttle-valve T. By such means the motion of the balls I, towards or from the vertical spindle, produces in the throttle-valve a corresponding motion; and they are so connected that the divergence of the balls I will cause the throttle-valve to close, while their descent towards the vertical spindle will cause it to open.

These arrangements being comprehended, let us suppose that, either by reason of a diminished load upon the engine or an increased activity of the boiler, the speed has a tendency to increase. This would impart increased velocity to the grooved wheel A B, which would cause the balls I to revolve with an accelerated speed. The centrifugal force which attends their motion would therefore give them a tendency to move from the axle, or to diverge. This would cause, by the means already explained, the throttle-valve T to be partially closed, by which the supply of steam from the boiler to the cylinder would be diminished, and the energy of the moving power, therefore, mitigated. The undue increase of speed would thereby be prevented.

If, on the other hand, either by an increase of the load, or a diminished activity in the boiler, the speed of the machine was lessened, a corresponding diminution of velocity would take place in the grooved wheel A B. This would cause the balls I to revolve with less speed, and the centrifugal force produced by their circular motion would be diminished. This force being thus no longer able fully to counteract their gravity, they would fall towards the spindle, which would cause, as already explained, the throttle-valve to be more fully opened. This would produce a more ample supply of steam to the cylinder, by which the velocity of the machine would be restored to its proper amount.

(126.) The principle which renders the governor so perfect a regulator of the velocity of the machine is difficult to be [Pg212] explained without having recourse to the aid of the technical language of mathematical physics. As, however, this instrument is of such great practical importance, and has attracted such general admiration, it may be worth while here to attempt to render intelligible the mechanical principles which govern its operation. Let S (_fig._ 42.) be the point of suspension of a common pendulum S P, and let P O P′ be the arch of its vibration, so that the ball P shall swing or vibrate alternately to the east and to the west of the lowest point O, through the arches O P′ and O P. It is a property of such an instrument that, provided the arch in which it vibrates be not considerable in magnitude, the time of its vibration will be the same whether the arch be long or short. Thus, for example, if the pendulum, instead of vibrating in the arch P P′, vibrated in the arch _p p′_, the time which it would take to perform its vibrations would be the same. If, however, the magnitude of the arch of vibration be increased, then a variation will take place in the time of vibration; but unless the arch of vibration be considerably increased, this variation will not be great.

Now let it be supposed that while the pendulum P P′ continues to vibrate east and west through the arch P P′, it shall receive such an impulse from north and south as would, if it were not in a state of previous vibration, cause it to vibrate between north and south, in an arch similar to the arch P P′. This second vibration between north and south [Pg213] would not prevent the continuance of the other vibration between east and west; but the ball P would be at the same time affected by both vibrations. While, in virtue of the vibration from east to west, the ball would swing from P to P′, it would, in virtue of the other vibration, extend its motion towards the north to a distance from the line W E equal to half a vibration, and will return from that distance again to the position P′. While returning from P′ to P, its second vibration will carry it towards the south to an equal distance on the southern side of W E, and it will return again to the position P. If the combination of these two motions or vibrations be attentively considered, it will be perceived that the effect on the ball will be a circular motion, precisely similar to the circular motion of the balls of the governor already described.

Now the time of vibration of the pendulum S P between east and west will not in any way be affected by the second vibration, which it is supposed to receive between north and south, and therefore the time the pendulum takes in moving from P to P′ and back again from P′ to P will be the same whether it shall have simultaneously or not the other vibration between north and south. Hence it follows that the time of revolution of the circular pendulum will be equal to the time of similar vibrations of the same pendulum, if, instead of having a circular motion, it were allowed to vibrate in the manner of a common pendulum.

If this point be understood, and if it also be remembered that the time of vibration of a common pendulum is necessarily the same whether the arch of vibration be small or great, it will be easily perceived that the revolving pendulum or governor will have nearly the same time of revolution whether it revolve in a large circle or a small one: in other words, whether the balls revolve at a greater or a less distance from the central spindle or axis. This, however, is to be understood only approximately. When the angle of divergence of the balls is as considerable as it usually is in governors, the time of revolution at different distances from the axis will therefore be subject to some variation, but to a very small one. [Pg214]

The centrifugal force (which is the name given in mechanics to that influence which makes a body revolving in a circle fly from the centre) depends conjointly on the velocity of revolution, and on the distance of the revolving body from the centre of the circle. If the velocity of revolution be the same, then the centrifugal force will increase in the same proportion as the distance of the revolving body from the centre. If, on the other hand, the distance of the revolving body from the centre remain the same, the centrifugal force will increase in the same proportion as the square of the time of vibration diminishes, or, in other words, it will increase in the same proportion as the square of the number of revolutions per minute. It follows from this, therefore, that the greater is the divergence of the balls of the governor, and the more rapidly they revolve, the greater will be their centrifugal force. Now this centrifugal force, if it were not counterbalanced, would give the balls a constant tendency to recede from the centre; but from the construction of the apparatus, the further they are removed from the centre the greater will be the effect of their gravitation in resisting the centrifugal force.

It is evident that the ball at P will have a greater tendency to fall by gravitation towards O than it would have at _p_, because the acclivity of the arch descending towards O at P is greater than its acclivity at _p_. The gravitation, therefore, or tendency of the ball to fall towards the central axis being greater at P than at _p_ it will be able to resist a greater centrifugal force. This increased centrifugal force, which the ball would have revolving at the distance P above what it would have at the distance _p_, is produced partly by the greater distance of the ball from the central axis, and partly by the greater velocity of its motion. But it will be evident that the time of its revolution may nevertheless be the same, or nearly the same, at both distances. If it should appear that the actual velocity of its motion of revolution at P be greater than its velocity at _p_, in the same proportion as the circles in which they revolve, then it is evident that the time of revolution would be as much increased by the greater space which P will have to travel over, as it will have to be [Pg215] diminished by the greater speed with which that space is traversed. The time of revolution, therefore, may be the same, or nearly the same, in both cases.

If this explanation be comprehended, it will not be difficult to apply it to the actual case of the governor. If a sudden increase of the energy of the moving power, or a diminution of the load, should give the machine an increased velocity, then the increased speed of the balls of the governor will give them an increased centrifugal force, which for the moment will be greater than the tendency of their gravitation to make them fall towards the vertical axis. This centrifugal force, therefore, prevailing, the balls will recede from the axis; but as they recede, their gravitation towards the vertical axis will, as has been already explained, be increased, and will become equal to the centrifugal force produced by the increased velocity, provided that velocity do not exceed a certain limit. When the balls, by diverging, get such increased gravitation as to balance the centrifugal force, then they will continue to revolve at a fixed distance from the vertical axis. When this happens, the time of the revolution must be nearly the same as it was before their increased divergence; in other words, the proportion of the moving power to the load will be so restored by the action of the levers of the governor on the throttle-valve that the machine will move at its former velocity, or nearly so.

The principle on which the governor acts, as just explained, necessarily supposes temporary disarrangements of the speed. In fact, the governor, strictly speaking, does not maintain a uniform velocity, but restores it after it has been disturbed. When a sudden change of motion of the engine takes place, the governor being immediately affected will cause a corresponding alteration in the throttle-valve; and this will not merely correct the change of motion, but it will, as it were, overdo it, and will cause a derangement of speed of the opposite kind. Thus if the speed be suddenly increased to an undue amount, then the governor being affected will first close the throttle-valve too much, so as to reduce the speed below the proper limit. This second error will again affect the governor in the contrary way, and the speed [Pg216] will again be increased rather too much. In this way a succession of alterations of effect will ensue until the governor settles down into that position in which it will maintain the engine at the proper speed.

To prevent the inconvenience which would attend any excess of such variations, the governor is made to act with great delicacy on the throttle-valve, so that even a considerable change in the divergence of the balls shall not produce too much alteration in the opening of that valve: the steam in the boiler should have at least 2 lbs. per square inch pressure more than is generally required in the cylinder. This excess is necessary to afford scope for that extent of variation of the power which it is the duty of the throttle-valve to regulate.

The governor is usually so adjusted as to make thirty-six revolutions per minute, when in uniform motion; but if the motion is increased to the rate of thirty-nine revolutions, the balls will fly to the utmost extent allowed them, being the limitation of the grooves in which their rods move; and if, on the other hand, the speed be diminished to thirty-four revolutions per minute, they will collapse to the lowest extent of their play. The duty of the governor, therefore, is to correct smaller casual derangements of the velocity; but if any permanent change to a considerable extent be made either in the load driven by the machine or in the moving power supplied to it from the boiler, then a permanent change is necessary to be made in the connection between the governor and the throttle-valve, so as to render the governor capable of regulating those smaller changes to which the speed of the machine is liable.

(127.) Having thus explained the principal mechanical contrivances provided by Watt for the maintenance and regulation of the rotatory motion to be produced by his double-acting steam engine, let us now consider the machine as a whole, and investigate the process of its operation. A section of this engine is represented in _fig._ 43.

Steam is supplied from the boiler to the cylinder by the steam pipe S. The throttle-valve T in that pipe, near the cylinder, is regulated by a system of levers connected with [Pg217] the governor. The piston P is accurately fitted in the steam cylinder C by packing, as already described in the single-acting engine. This piston, as it moves, divides the cylinder into two compartments, between which there is no communication by which steam or any other elastic fluid can pass. The upper steam box B is divided into three compartments by the two valves. Above the upper steam valve V is a compartment communicating with the steam pipe; below the upper exhausting valve E is another compartment communicating with the eduction pipe which leads to the condenser. By the valves V and E a communication may be opened or closed between the boiler on the one hand, or the condenser on the other, and the top of the cylinder. The continuation S′ of the steam pipe leads to the lower box B′, which, like the upper, is divided into three compartments by two valves V′ and E′. The upper compartment communicates with the steam pipe, and thereby with the boiler; and the lower compartment communicates with the eduction pipe, and thereby with the condenser. By means of the two valves V′ and E′, a communication may be opened or closed between the steam pipe on the one hand, or the exhausting pipe on the other, and the lower part of the cylinder. The four valves V, E, V′, and E′ are connected by a system of levers with a handle or spanner _m_, which, being driven downwards or upwards, is capable of opening or closing the valves in pairs, in the manner already described (116.). The condensers, the air-pump, and the hot-water pump, are in all respects similar to those already described in the single-acting engine, except that the condensing jet is governed by a lever I, by which it is allowed to play continually in the condenser, and by which the quantity of water admitted through it is regulated. The cold-water pump N is worked by the engine as already described in the single-acting engine, and supplies the cistern in which the air-pump and condenser are submerged, so as to keep down its temperature to the proper limit. On the air-pump rod R are two pins properly placed, so as to strike the spanner _m_, upwards and downwards, at the proper times, when the piston approaches the termination of the stroke at the top or bottom of the cylinder. The pump L [Pg218] conducts the warm water drawn by the air-pump from the condenser to a proper reservoir for feeding the boiler. The vertical motion of the piston-rod in a straight line is rendered compatible with the circular motion of the end of the beam by the parallel motion already described. The point _b_, on the beam, moves upwards and downwards in a circular arch, of which the axis of the beam is the centre. In like manner the point _d_ of the rod _d c_ moves upwards and downwards, in a similar arch of which the fixed pivot _c_ is the centre. The joint or bar _d b_, which joins these two pivots, will be moved so that its middle point _e_ will ascend and descend nearly in a straight line, as has been already explained (120.); [Pg219] opposite this point _e_ is attached the piston-rod of the air-pump, which is accordingly guided upwards and downwards by this means. The jointed parallelogram _b d g f_ is attached to the beam by pivots; and, as has been explained (120.), the point _g_ will be moved upwards and downwards in a straight line, through twice the space through which the point _e_ is moved. To the point _g_ the rod of the steam piston is attached. Thus, the rods of the steam piston and air-pump are moved by the same system of jointed bars, and moved through spaces which are in the proportion of two to one.

Although this system of jointed rods forming the parallel motion, appears in the figure to consist only of one parallelogram _b d g f_, and one rod _c d_, called the _radius rod_, it is, in fact, double, a similar parallelogram and radius rod being attached to corresponding points, and in the same manner on the other side of the beam; but from the view given in the cut, the one set of rods hides the other. The two systems of rods thus attached to opposite sides of the beam at several inches asunder, are connected by cross rods, the ends of which form the pivots or joints, and extend between the parallelograms. The ends of these rods are only visible in the figure. It is to the middle of one of these rods, the end of which is represented at _e_, that the air-pump piston-rod is attached; and it is to the middle of another, the end of which is represented at _g_, that the steam piston-rod is attached. These two piston-rods, therefore, are driven, not immediately by either of the parallelograms forming the parallel motion, but by the bars extending between them.

To the working end of the beam H is attached a rod of cast-iron O, called the _connecting rod_, the lower end of which is attached to the crank by a pivot. The weight of the connecting rod is so made, that it shall balance the weight of the piston-rods of the air-pump and cylinder on the other side of the beam; and the weight of the piston-rod of the cold-water pump N nearly balances the weight of the piston-rod of the hot-water pump L. Thus, so far as the weights of the machinery are concerned, the engine is in equilibrium, and the piston would rest in any position indifferently in the cylinder.

The axis of the fly-wheel on which the crank is formed is [Pg220] square in the middle part, where the fly-wheel is attached to it, but has cylindrical necks at each end, which rest in sockets or bearings supported by the framing of the machine, in which sockets the axis revolves freely. On the axle of the crank is placed the fly-wheel, and connected with its axle is the governor Q, which regulates the throttle-valve T in the manner already described.

Let us now suppose the engine to be in full operation. The piston being at the top of the cylinder, the spanner _m_ will be raised by the lower pin on the air-pump rod, and the upper steam valve V, and the lower exhausting valve E′, will be opened, while the upper exhausting valve E and the lower steam valve V′ are closed. Steam will, therefore, be admitted above the piston, and the steam which filled the cylinder below it will be drawn off to the condenser, where it will be converted into water. The piston will, therefore, be urged by the pressure of the steam above it to the bottom of the cylinder. As it approaches that limit, the spanner _m_ will be struck downwards by the upper pin on the air-pump rod, and the valves V and E′ will be closed, and at the same time the lower steam valve V′ and the upper exhausting valve E will be opened. Steam will, therefore, be admitted below the piston, while the steam above it will be drawn off into the condenser, and converted into water. The pressure of the steam, therefore, below the piston will urge it upwards, and in the same manner the motion will be continued.

While this process is going on in the cylinder and the condenser, the water formed in the condenser will be gradually drawn off by the operation of the air-pump piston, in the same manner as explained in the single-acting engine; and at the same time the hot water thrown into the hot well by the air-pump piston will be carried off by the hot-water pump L.

Such are the chief circumstances attending the continuance of the operation of the double-acting engine. It is only necessary here to recall what has been already explained respecting the operation of the fly-wheel. The commencement of the motion of the piston from the top and bottom of the cylinder is produced, not by the pressure of the steam upon it upwards or downwards, which must, for the reasons [Pg221] already explained, be entirely inefficient; but by the momentum of the fly-wheel, which extricates the crank from those positions in which the moving power cannot affect it.

The manner in which the motion of the crank affects the connecting rod at the dead points produces an effect of great importance in the operation of the engine. When the crank-pin is approaching the lowest point of its play, and therefore the piston approaching the top of the cylinder, the motion of the crank-pin becomes nearly horizontal, and consequently its effect in drawing the connecting rod and the working end of the beam downwards and the piston upwards, is extremely small. The consequence of this is, that as the piston approaches the top of the cylinder, its motion becomes very rapidly retarded; and as the motion of the crank-pin at its lowest point is actually horizontal, the piston is brought to a state of rest by this gradually retarded motion at the top of the cylinder. In like manner, when the crank-pin moves from its dead point upwards, its motion at first is very nearly horizontal, and consequently its effect in driving the working end of the beam upwards, and the piston downwards, is at first very small, but gradually accelerated. The effect of this upon the piston is, that it arrives at and departs from the top of the stroke with a very slow motion, being absolutely brought to rest at that point.

The same effect is produced when the piston arrives at the bottom of the cylinder. This retardation and suspension of the motion of the piston at the termination of the stroke affords time for the process of condensation to be effected, so that when the moving power of the steam upon the piston can come into action, the condensation shall be sufficiently complete. As the piston approaches the top of the cylinder, and its motion becomes slow, the working gear is made to open the lower exhausting valve; the steam enclosed in the cylinder below the piston, and which has just driven the piston upwards, presses with an elastic force of 17 lbs. per square inch on every part of the interior of the cylinder, while the uncondensed vapour in the condenser presses with a force of about 2 lbs. per square inch. The steam, therefore, will have a tendency to rush from the cylinder to the [Pg222] condenser through the open exhausting valve, with an excess of pressure amounting to 15 lbs. per square inch, while the piston pauses at the top of the cylinder. This process goes on, and when the piston has descended by the motion of the fly-wheel, a sufficient distance from the top of the cylinder to call the moving force of the steam into action, the exhaustion will be complete, and the pressure of the uncondensed vapour in the cylinder will become the same as in the condenser.

The pressure of steam in the cylinder, and of uncondensed vapour in the condenser, varies, within certain limits, in different engines, and therefore the amount here assigned to them must be taken merely as an example.

The size of the valves by which the steam is allowed to pass from the cylinder to the condenser should be such as to cause the condensation to take place in a sufficiently short time, to be completed when the steam impelling the piston is called into action.

Watt, in the construction of his engines, made the exhaustion-valves with a diameter which was one fifth of the diameter of the cylinder, and therefore the actual magnitude of the aperture for the escape of the steam was one twenty-fifth of the magnitude of the cylinder; but the spindle of the valve diminished this so that the available space for the escape of steam did not exceed one twenty-seventh of the magnitude of the cylinder. This was found to produce a sufficiently rapid condensation.

It was usual to make the steam valves of the same magnitude as the exhausting valves, but the flow of steam through the former was resisted by the throttle-valve, while no obstruction was opposed to its passage through the latter.

The rapidity with which the cylinder must be exhausted by the condenser will, however, depend upon the velocity with which the piston is moved in it. The magnitude, therefore, of the exhausting valves which would be sufficient for an engine which acts with a slow motion would be too small where a rapid motion is required.

In the single-acting steam engine, where the moving force always acted downwards on the piston, the pressure upon [Pg223] all the joints of the machinery by which the force of the piston was conveyed to the working parts, always took place in the same direction, and consequently whatever might be the mechanical connection by which the several joints were formed, the pins by which they were connected, must always come to a bearing in their respective sockets, however loosely they may have been fitted. For the same reason, however, that the arch head and chain were abandoned as a means of connecting the steam piston with the beam, and the parallel motion substituted, it was also necessary in the double-acting engine, where all joints whatever were driven alternately in opposite directions, to fit the connecting pins with the greatest accuracy in their sockets, and to abandon all connection of the parts by chains. If any sensible looseness was left in the joints, a violent jerk would be produced every time the motion of the piston was reversed. Any looseness either in the pivots or joints of the parallel motion of the working beam, the connecting rod, or crank, would, at every change of stroke, be so accumulated as to produce upon the machinery the effects of percussion, and would consequently be attended with the danger of straining and breaking the moveable parts of the mechanism.

To secure, therefore, the necessary accuracy of the joints, Watt contrived that every joint in the engine should admit of the size of the socket being exactly adapted to the size of the pin, so as always to make a good fitting by closing the socket upon the pin, when any looseness would be produced by wear. With this view, all the joints were fitted with sockets made of brass or gun-metal, capable of adjustment. Each socket was composed of two pieces, accurately fitted into a cell or groove, in which one of the brasses can be moved towards the other by means of a wedge or screw. Each brass has in it a semi-cylindrical cavity, and the two cavities being opposed to each other, form a socket for the joint-pin. One of the two brasses can always be tightened round that pin, so as to enclose it tight between the two semi-cylindrical cavities, and to prevent any looseness taking place. The brasses, and other parts of such a joint, are represented [Pg224] in _fig._ 44. These joints still continue to be used in the engines as now constructed.

The motion of the working beam, and the pump-rods which it drives, and of the connecting rod, ought, if the whole were constructed with perfect precision, to take place in the same or parallel vertical planes; but this supposes a perfection of execution which could hardly have been expected in the early manufacture of such engines, whatever may have been attained by improvements which have been since made. In the details of construction, Watt saw that there would be a liability to lateral strain, owing to the planes of the different motions not being truly vertical and truly parallel, and that if a provision were not made for such lateral motion, the machinery would be subject to constant strain in its joints and rapid wear. He provided against this by constructing the main joints by which the great working lever was connected with the pistons and connecting rod, so as to form universal joints, giving freedom of motion laterally as well as vertically.

The great lever, or working beam, was so called from being originally made from a beam of oak. It is now, however, universally constructed of cast-iron. The connecting rod is also made of cast-iron, and attached to the beam and to the crank by axles or pivots.