11

the invention. A partnership was then concluded between Boulton and Watt, and a part of Soho works was handed over to Watt for the manufacture of his engines. He now made rapid progress, and by 1778 had erected several excellent engines in the neighbouring counties, and one at Stratford near London. In these early engines he appears to have employed the method of condensing by external cold, and he persisted in the use of that species of condenser until the expense of its construction and the inconvenience of its bulk in the larger class of engines more than balanced the expenditure of power requisite to extract the injection water from a vacuum. The preceding diagram shews the manner in which Watt's improvements at this period of their history were applied to the ordinary pumping engines of the time, the substitution being restricted to the cylinder and its appendages, and the great lever, chains, pumps, and other parts remaining unaltered. E, E is the cylinder surrounded by a steam case, into which the steam is delivered by the steam-pipe a; J is the piston, loaded with a pressure of between 10 and 11 lbs. on the square inch, instead of 7 as formerly, in consequence of the superior vacuum. n the piston-rod passing through a short tube affixed to the cylinder-cover, called the stuffing-box, and now cast in the same piece with the cover; f, f, the cylinder port; e, the steam valve; i, the eduction valve; g, g, the eduction pipe; F, the condenser, consisting of a number of copper cylinders of small diameter, arranged vertically in the cold water cistern G, G; H, the pump, by which the condensed steam is extracted from the condenser, and returned to the boiler; 1, the plug-beam, in which are inserted two plugs, or tappets, which strike at the proper times the handles D and r, and thereby give the requisite movement to the steam and eduction valves, by the rods 4 and 6, with which those rods communicate; 16 is a weight which serves the same end as the tumbling bob in the old engines, in confirming the inclination given to the valve gear by the tappet. In the engine as thus constructed, the steam-casing is obviously an integral part of the machine; but this plan was found to be attended with some inconveniences, and in 1778 Watt contrived a way of removing this objection by the application of an upper port to the cylinder as shown in fig. 10. where E and J are the cylinder and piston as before; a, the steam-pipe ; b, the regulating or throttling valve; c, the upper port; f, the under port; Fig. 10.

E E is the cylinder and J the piston as before, a the steam pipe, b the regulating valve, e the eduction and equilibrium valve, for in this variety of construction a single valve performs the functions of both; c the upper and f the under port; d, j, g, the eduction pipe by which the steam passes from above the piston during every returning stroke to the condenser, a perpetual exhaustion being maintained beneath it. There was no advantage found to arise in practice from this ingenious transformation, but rather the reverse; for, although a longer time was thus obtained for the accomplishment of the condensation, yet the process of condensation is so rapid that this additional time was found to be of very little advantage; and whatever advantage was thus realised was to some extent only transferred from the working to the returning stroke, a heavier counterweight being necessary to redress any difference in the vacuum above and beneath the piston which might arise from a want of rapidity in the condensation. There was also a greater leakage of air at the stuffing-box and around the cylinder cover in this species of engine; and the rare steam or vapour, remaining in the cylinder after the act of exhaustion, had its elasticity increased by the heat transmitted from the steam jacket, thereby opposing more resistance to the piston, and robbing the effective steam of a portion of its heat. These disadvantages, taken singly, are all trivial enough, and indeed the sum of them is of no very serious import; yet, upon the whole, this species of engine appears to be somewhat inferior to engines of the ordinary kind, and it never, therefore, met with an extensive adoption. Boulton and Watt at this time charged their profits in proportion to the saving of fuel their engines effected, one third part of which was to be payable to them annually during the term of their patent, or the payment might be redeemed at a ten years' purchase. Their proposal was to raise 500,000 cubic feet of water one foot high by the consumption of 112 lbs. of Wednesbury coals, which is equivalent to about 23 millions of pounds raised one foot high by a bushel or 84 lbs. Their early engines, however, scarcely accomplished this; for two of them experimented on by Smeaton only lifted 18 and 18 millions of pounds one foot high with a bushel; but these were small engines. The larger engines appear to have realised a performance of about 20 millions; so that the following rule, laid down by Mr. Boulton in 1778, pretty nearly expresses the performance of the larger engines constructed by Boulton and Watt at that time. Compute the area of the space described by the piston per stroke in cubic feet, and multiply this by the load

upon each square inch of the piston in pounds; the product will be the weight of coals in pounds required to work the engine 1800 strokes. This is equivalent to about 21 millions raised one foot high by a bushel of coal.

The next improvement of Watt's that we have to mention, is his plan of working steam expansively. This method consists in arresting the flow of steam into the cylinder at a certain part of the stroke, leaving the remainder of the stroke to be accomplished by the effort of the steam shut within the cylinder to occupy a larger volume. The power of the engine is of course diminished by this procedure; for the piston will descend with less force when urged merely by the diminishing effort of the expanding steam, than if pressed upon by steam of the full pressure entering to the end of the stroke from the boiler. But steam, or, in other words, fuel, is saved in a greater proportion than the power of the engine is diminished; so that the expansive principle augments, and very materially, too, the motive efficacy of the fuel. In fact, whatever power is obtained from the steam during the act of expanding is obtained without any expense; for if the steam valve of the cylinder be closed when half the stroke of the piston is performed, there will only be half the steam expended; but the steam shut within the cylinder will press with a varying force on the piston up to the end of the stroke, and the power thus realised is evidently got without any expenditure. We think it very probable that this improvement originated in the desire merely to moderate the force of the single acting engine towards the end of the stroke; and, indeed, Professor Robison, who, from his intimacy with Watt, was probably well acquainted with the circumstances of the discovery, virtually says that such was the case. In Newcomen's engine the momentum of the piston could easily be lessened by shutting the injection cock earlier or by opening it less, and dangerous shocks might thus be averted; but, in engines provided with a condenser, the impulse of the piston would be best checked by shutting the steam valve at such a period of the stroke as would prevent the catch-pins from striking. The realisation of an increased power, therefore, from a given quantity of steam by this expedient, was an unexpected result; but Watt immediately saw the importance of the principle, which in a letter from Glasgow in 1769 to Dr. Small, of Birmingham, he describes as capable of doubling the effect of the steam. The distractions, however, incidental to his other pursuits appear to have prevented him from carrying the principle of expansion into practice until 1776, when it was tried upon the engine at Soho, and in 1778 it was applied to an engine for raising water, erected at Shadwell. In 1782 Watt took out a patent for improvements in the steam engine, in which the principle of expansion formed a prominent feature, having been, probably, instigated to that act by a patent taken out by Hornblower in 1781 for a method of using the steam in an engine twice over; first impelling a small piston by the method of high pressure, propounded by Leupold, and then catching the steam thus used, and making it instrumental in giving motion to a larger piston by the method of condensation. This scheme is identical in principle with the plan of using the steam expansively, for it is obvious that the same power will be given out by a cylinder, whether it be tall and narrow, or short and wide, provided its capacity remains unaltered; and if we suppose Hornblower's high pressure cylinder to be shortened and widened, and set upon the top of his condensing cylinder, we shall then have a single cylinder operating on the plan of expansion, while the power is obviously the same as before this transformation. The subject of expansion, however, is so important, and at the same time so mysterious to tyros in steam science, that it would be unpardonable to pass it over without offering such further familiar elucidations as may make the principle intelligible to unskilful persons, and we believe these explanations may be introduced here with greater propriety than at a more advanced stage of our progress.

It is a well-known law of pneumatics that the pressure of elastic fluids varies inversely, as the spaces into which they are compressed. For example, if a cubic foot of air of the atmospheric density be compressed into the compass of half a cubic foot, its elasticity will be increased from 15 lbs. on the square inch to 30 lbs. on the square inch, whereas if its volume be enlarged to two cubic feet, its elasticity will be diminished to 7 Ibs. on the square inch, being just the half of its original pressure. The same law holds in all other proportions, and with all other gases and vapours, provided their temperature remain unchanged; and if the steam valve of an engine be closed when the piston has descended through onefourth of the stroke, the steam within the cylinder will, at the end of the stroke, just exert one-fourth of its initial pressure. Thus let E (fig. 12.) be a cylinder; J the piston, a the steam pipe, b the steam valve, c the upper port, f the lower port, d the steam pipe prolonged to e the equilibrium valve, g the eduction valve, M the steam casing, N top of cylinder, O stuffing box, n piston rod, P cylinder bottom: let the cylinder be supposed to be divided in the direction of its length into any number of equal parts, say twenty, an I let the diameter of the cylinder represent the pressure of the steam, which for the sake of simplicity we may take at 10 lbs., so that we may divide the cylinder in the direction of its diameter into ten equal parts. If, now, the piston he supposed to descend through five of the divisions, and the steam valve then be shut, the pressure represented at each subsequent position of the piston will be represented by a series computed according to the laws of pneumatics, and which, if the initial pressure be represented by 1 will give a pressure of 5 at the middle of the stroke, and of 25 at the end of it. If this series be set off on the horizontal lines they will give a hyperbolic curve, the area of the part

exterior to which represents the total efficacy of the stroke, and the interior area therefore represents the diminution in the efficacy of a stroke when the steam is cut off at one fourth of the descent. If the squares above the point where the steam is cut off be counted, they will be found to amount to 50, and if those beneath that point be counted or estimated they will be found to amount to about 68, and these squares are representative of the power exerted, so that while an amount of power, represented by 50, has been obtained by the expenditure of a quarter of a cylinder full of steam we get an amount of power represented by 68 without any expenditure of steam at all, merely by permitting the steam first used to expand into four times its original volume. The efficacy of a given quantity of steam is therefore more than doubled by expanding the steam four times, while the efficacy of each stroke is made nearly one half less; and therefore to carry out the expansive system in practice the cylinders require to be larger than usual in the proportion to which the expansion is carried. Every one who is acquainted with simple arithmetic can compute the terminal pressure of the steam in a cylinder when he knows the initial pressure and the point at which the steam is cut off, and he can also find by the same process any pressure intermediate between the first and the last. By setting down these pressures in a table, and taking their mean, he can easily determine the effect with tolerable accuracy of any particular measure of expansion, of which the mean pressure thus determined will be the representative. We shall, however, at the proper place give practical rules for computing the effect of expansion at a single step: at present we aspire to accomplish nothing more than to convey a few general ideas on the subject to those who are not familiar with such inquiries; and as a summary of the ascertained effects of expansion will probably induce a more careful examination of the principle at a future stage of our progress, we may here set down some of the most notorious.

A forty-horse engine, constructed by Mr. Watt about the time of the introduction of his expansive principle, was found to require about 8 lbs. of coal per horse-power per hour when working without expansion, and about 6 lbs. when the expansion was 1.518 times. The water evaporated from the boiler was, without expansion, 674 cubic feet per minute; and, with the amount of expansion already mentioned, 501 cubic feet per minute and the quantity of injection water was, without expansion, 19.3 cubic feet per minute, at a temperature of 52 degrees; and, with the expansion aforesaid, 14:35 cubic feet.

Previous to the introduction of Boulton and Watt's engine into Cornwall, Mr. Jonathan Hornblower had been one of the principal makers of engines for that district; and he appears to have made very strenuous efforts to prevent himself from being driven out of so profitable a market. In 1781 he took out a patent, as we believe we have already mentioned, for a double cylinder engine, which is said to have been invented in 1776. The form of this engine will be understood at once by a glance at the accompanying sketch (fig. 13.), where A is the larger cylinder; B the smaller one; C and D their respective piston rods; G a square pipe for supplying the small cylinder with steam from the boiler; K is the eduction pipe leading to the

There is thus less tendency for the motion to become irregular, and less risk of fracture from the extra pressure of steam requisite to maintain the power of an engine when working expansively. But there is no theoretical gain by the use of two cylinders, and there is certainly greater complexity, while experience shows that an engine constructed on the common proportions is strong enough to be capable of working up to its power at all ordinary degrees of expansion, and the motion may be rendered uniform enough for most purposes by the use of a short stroke and a swift and heavy fly wheel. There are cases, nevertheless, in practice in which we think two cylinders are to be preferred to one, but their number is by no means considerable, and they lie very little among the class of pumping engines. Several excellent pumping engines, indeed, of the double cylinder description have been erected in Cornwall; and some, we believe, are being constructed there at the present moment, with the view, we suppose, of carrying the principle of expansion very far, without involving the necessity of making the parts of the engine inconveniently strong. But the same end might, we conceive, be attained as effectually, and with greater ease, by increasing the length of the stroke; and in a pumping engine uniformity of motion is an object of but little importance.

The form Mr. Watt's engine had attained about the period at which we have now arrived is shown in the following figure, where A is the cylinder, B the piston, C the piston rod, E the cylinder steam casing, F the steam

Fig. 14.

HORNBLOWER'S ENGINE.

condenser, L, which stands on the box, M, and to which are fixed the pumps O and N, for extracting the air and water; P is a cold water cistern in which the condenser stands; U is the hot well to which the water is conveyed from the air pumps by the trough, T; and into the hot well the feed pump V dips to obtain water for the boiler; W is the plug rod; and X the pump rod of the mine. At S a valve is situated to admit cold water into the condenser; and p is a pipe leading from the condenser, closed on the top with a valve which serves as a blow-through valve. a b c d are cocks, two of which are the reciprocal of the other two, and which establish the requisite communications between the boiler and the cylinders. When d and a are open, and b and c are shut, the steam from the boiler enters through d above the piston of B, and the steam from beneath the piston of B enters, above the piston of A, through a. At the same time the valve situated in the eduction pipe, where it joins the cylinder, is open, and the engine makes a stroke. Again, when b and c are open, and a and d shut, a communication is established between each side of the respective pistons; and, as an equilibrium of pressure is thus produced, the pistons rise by virtue of the preponderance of the pump end of the beam. There is a great deal of manifest plagiarism in this engine: the condenser and air pump, the cylinder covers with their stuffing boxes, and indeed every thing good about the engine, is evidently borrowed from Mr. Watt, with the single exception of the two cylinders; and by them nothing was accomplished that had not been already attained by Mr. Watt quite as effectually with a single cylinder only. The plan of two cylinders is still sometimes used in cases requiring a great equability of motion, or where the principle of expansion is carried to an unusual extent; for the initial impulse is less vehement on the double cylinder plan than where the expansion is accomplished in a single cylinder, while the mean impelling force remains the same.

WATT'S SINGLE-ACTING ENGINE.

pipe, G the steam valve, by which the admission of steam to the cylinder is restrained or prevented, H the boiler, I the prolongation of the steam pipe to the equilibrium valve K, L the eduction valve, M the condenser, J the eduction pipe, N the injection cock, O the foot valve, P the air-pump, Q the air-pump bucket, S the delivery valve, communicating with the hot well, T the suction pipe of the feed pump, U the feed pump, V the cold water pump for lifting the injection water into the cold water cistern, Y the valve handles, which are moved by tappets on the plug-tree, Z, suspended from the beam; a is the body of the beam, b the main centre; c, c, truss rods, proceeding from the central upright, d, to each end of the beam, to obviate deflection; e, f arch heads, to which the chains of the piston rod and pump rod accommodate themselves, g a small arch head for giving motion to the air-pump bucket, on the rod of which the plug tree is affixed; h, j counterbalance weights; k, suction valve of the pump; o, o, flues encircling the boiler; p furnace; r damper in the flue leading to the chimney. The manner in which this engine operates is easily comprehended. In this as in the atmospheric engine, the piston necessarily stands at the top of the cylinder when the engine is at rest, in consequence of the preponderance given to the pump end of the beam, and the first step in starting the engine is to expel the air with which the space beneath the piston, and the several pipes and passages have become filled. This is accomplished by relieving the handles of the steam, equilibrium, and eduction valves, of their several catches, when the valves immediately fall open by the gravitation of weights with which they are connected, and the steam finds an access to the whole internal part of the engine. At first it is rapidly condensed by coming into contact with the cold metal; but as the iron becomes hot the steam flows onward, expelling the air before it, and finally issues at the snifting or blow valve, situated on the lower part of either the air-pump or condenser. The D

engine is not represented in the specification of his patent, but the sketch given in fig. 15. will explain in what manner the crank was applicable to the common atmospheric engine.

A is the furnace, B the ash-pit, C the boiler, E the cylinder, L the great lever, G the injection cistern, M the connecting rod, N the crank, Q the fly-wheel, I the plug-tree which gives motion to the valves. It is obvious that this engine can only give an impulse to the fly-wheel during its downward stroke, but the energy then imparted to the fly-wheel will continue the motion with considerable uniformity during the ascent of the piston. To add to the regularity of the motion, in the case of a crank applied to an engine on the atmospheric plan, Mr. Francis Thompson of Ashover, in Derbyshire, contrived about 1793 a plan of combining two cylinders so as to produce a double action. The foregoing representation shows the nature of his arrangements:- E (fig. 16.) is the ordinary cylinder, and F an inverted cylinder situated above it with a space, n, between the cylinders to permit the access of the atmosphere to press upon the pistons. LL is the main lever, and K the arch head which is connected with the piston rod by double-acting chains through the medium of a bridle, so that the piston rod may communicate an upward as well as a downward motion to the lever. There are three chains in all-two, for pulling the lever down, and one for pulling it up. G is the cold water or injection cistern, which receives its water from the pump R, through the pipe S; V is the lever wall, U the spring beams, and X a solid pier of masonry on which the lower cylinder rests. M is the connecting rod which turns the wheel, N, working into the pinion O, on the shaft of the fly-wheel Q, thereby giving an increased energy to the fly-wheel by its augmented velocity. a is the steam pipe, b and c the steam valves, f a passage leading from the steam valve into the lower cylinder, g g eduction pipes, h, i the injection cocks, j the injection pipe leading from the elevated cold water cistern, k a snifting valve to the lower cylinder, and there is a similar valve attached to the upper cylinder, but it is not shown; the plug rod which gives motion to the valve handles, r and s, by which the ingress of the steam and injection water into the cylinder is duly regulated. A large engine on this plan was set up by Mr. Thompson at Arnold near Nottingham, to drive a spinning mill, and it accomplished its task successfully for many years, though with a large consumption of fuel. The cylinders were 40 inches in diameter, the length of the stroke 6 feet, and the number of strokes per minute 18. Very few similar engines, however, have since been constructed, as they were found quite as difficult to keep in good order as the more refined engines of Boulton and Watt, and far more extravagant in fuel.

Another double-acting atmospheric engine, of a different construction, but also employed for spinning cotton, was erected in Manchester by Messrs. Sherratts about 1794. The idea of the double-acting mechanism of this engine is evidently taken from the double-acting air pump, as constracted by philosophical instrument makers; and of the rest of the plan

the greater part is pirated from Mr. Watt. EE (fig. 17.) are the cylinders open at the top, JJ the pistons, K K the piston rods, which are toothed, and work into the wheel LL, which answers the purpose of the great lever. I are racks for giving motion to the air pumps, and which work into the smaller wheel 7, 7, situated on the same axis as L. P is a lever also fixed on the same axis p, and which, by means of the connecting rod M, imparts motion to the crank, N, fixed on the shaft of the fly-wheel Q. a a is the steam pipe with a middle branch, d, descending in an inclined direction to the box holding the steam valves ff, and beneath which are two eduction valves communicating with the eduction pipe g. F is the condenser, k the foot valve, HH the air pumps, II the hot well, G G the cold water cistern; UU are strong beams for supporting the axis, d, of the wheel L, XX strong beams for supporting the cylinders, and Z a throttle valve in the descending branch, d, of the steam pipe, to regulate the flow of the steam. The cylinders of this engine were 36 inches in diameter, and the length of the stroke 4 feet. At first the engine made 40 strokes per minute, but its speed was afterwards reduced to 30 strokes per minute. It continued to work constantly for upwards of 30 years, but its proprietors had to pay Messrs. Boulton and Watt for a licence to use it, in consequence of the large adoption of their improvements. The use of the rack work is not so objectionable as might at first sight appear, as the pressure on the teeth is always in the same direction, provided steam of no higher tension than the atmosphere be used, and there is therefore no hack-lash between the teeth.

The only other engine we shall notice before proceeding to describe Mr. Watt's rotative engine is the engine of Mr. Edmund Cartwright, patented in 1797. The following are the improvements enumerated in his specification:

"1st.

The water or other liquid which is used to produce the steam to work the engine is to circulate continually through the engine, without having any communication with the external air, and without mixing with any other fluid.

2nd. The lower part of the cylinder is to be always open to the condenser, and the condensation is to be performed during the returning stroke of the piston, which has only a single action. There are to be only two valves, one at the top of the cylinder, and the other in the piston; and these valves are to be opened and shut at the proper intervals, by the motion of the piston itself.

"3rd. The condensation is to be performed in a close vessel by the application of cold water to the external surface of that vessel on the common principle of distillation, without any injection of cold water into it. The condenser is to be composed of two cylinders of very thin metal, one fixed within the other, so as to leave a very narrow space between them for the steam. This condenser is to be immersed in cold water, which is to surround the outside cylinder, and to fill the internal cylinder, whereby a very thin stratum of steam will be exposed to a large surface of cold metal. The condensed liquid which drains to the bottom of the condenser is to be drawn