through, and to admit steam above the piston to act upon it instead of the atmosphere." To prevent cooling of the cylinder by the external air, he would surround this by a larger one, the "jacket," the interspace to be kept filled with steam, and would cover or clothe the whole with wood or other substance conducting heat poorly. Thus, we must credit to the marquis of Worcester the first successful application of steam pressure to use; to Savery the application of the vacuum due to condensation, though he did not foresee the true method, or the full value of its application; to Papin the piston for receiving and transmitting the force of air or steam within a cylinder; to Newcomen and Cawley the cylinder and piston independent of the boiler, as also the working beam, and the plan of internal condensation; to Beighton the successful introduction of automatic apparatus for the valves; and to Watt the separate condenser, saving the cooling of the cylinder and consequent waste of steam in reheating it, and the exclusion of air from the eylinder, with introduction of steam above the piston, changes which, with those that followed and grew out of them, rendered the engine at length practical, economical, and complete. This was still a "single-acting" engine; the steam pressure during the pushing down of the piston being that alone which took effect on the mechanism to be driven, and the only object of the subsequent admission of steam below being to return the piston to the top of the cylinder. This engine was also chiefly used for pumping and draining. It seems to have occurred to Watt, as early as 1769, that additional economy would be secured, especially, as he thought, in working light loads, by closing the steam pipe before the piston had descended the full length of stroke, thus saving the filling of the cylinder completely with steam of the initial density, and allowing the stroke to be completed by expansion of that already admitted, aided of course by momentum of the beam and piston. This principle he first applied in 1776 in an engine erected at Soho, but he published no account of it until on patenting this and certain other improvements in 1782; the variety he named the expansive engine. The first public announcement of benefit from expansive working of steam, however, was by Jonathan Hornblower, who in 1781 employed two cylinders, one larger than the other; the steam, of the boiler pressure, having first driven the smaller piston, was immediately transferred, and allow ed to act during its expansion (of course with diminishing pressure) upon the increased area of the larger piston, the two cylinders being thus approximatively equal in power. VOL. XV.-4

Watt's general patent, however, was judged to exclude this invention, and it did not at that time come into use. In the single-acting engine, one half the motion was still unaccompanied by useful effect; and the application of the force was ill adapted to impart any other than a simple reciprocating movement. Since the time of Savery, it had been an object of importance with inventors to convert this movement into one of revolution, as adapted to machinery; and Hulls, Fitzgerald, Stewart, and others had contrived various means of effecting this change. Watt early conceived of the use of the common winch or crank for this purpose; it was patented, however, by Wasbrough, and then by Steed, in 1781. As Watt was at this time engaged in his invention by which the engine was to be made "double-acting"-the steam being admitted to press alternately, and in turn condensed, both above and below the pistonthus fitting it to impart revolution to a shaft, to wheel work, &c., he was obliged to resort to other methods to secure this part of his purpose, among them to the "sun and planet" wheel. The adaptation of the engine to the production of a rotary motion prepared the way for its employment as the prime mover of every kind of machinery. The specification of the double-acting steam engine, with several kindred improvements, was enrolled July 4, 1782; and in 1784 patents were secured for the "par

allel motion," the throttle-valve, the governor, the indicator, &c.; as also for a form of locomotive engine, which however proved impracticable. Among the earliest of the double-acting or rotative engines produced for sale was the Albion mill engine, fig. 1; an inspection

of which will serve not merely to show the earlier form of construction, but also to illustrate many points hereafter to be referred to in respect to ordinary forms of engine. In 1787, or according to some authorities even before 1785, Oliver Evans of Philadelphia constructed the first complete and practically working non-condensing or high pressure engine, thus paving the way by economy of space and comparative simplicity of the mechanism required for the locomotive engine and the adaptation of steam to purposes of land carriage. Of the immense number of inventions since that date, intended to improve the form, the working, or the economy of the steam engine, or to adapt it to specific purposes, it would be impracticable here to detail the whole; some of the more important will be named in connection with the description of parts and forms of engine, or with the explanation of its working. II. Essential Parts of Ordinary Forms of Engine. Being a contrivance for applying motive power to machinery, pumps, carriages, the shafts of paddle or screw steamers, &c., the steam engine proper always terminates with that working piece which directly transmits the power; as, in the commonest forms, with the connecting rod from the beam to the crank. The fly-wheel, governor, valve motion, or other parts that receive their motion from the main shaft (crank shaft), and that are useful only in relation to the engine, must be considered in connection with it; through these parts a portion of the propulsive effect is thrown back, for certain needful purposes, upon the engine itself. The non-condensing engine is essentially of two main parts: the boiler, or generator of steam with its appendages; and the cylinder and piston, with their various adjuncts. To these in the condensing engine is added a third, the condenser, with its appurtenances. It will conduce to clearness to consider first in order the second of these three, as just named, the condensing apparatus being quite simple; the boiler and furnace will be reserved for a subsequent place. The cylinder is usually of cast iron, the toughest to be had; and in order to secure the requisite stiffness to preserve its form, its thickness is much greater than is required for strength. The desiderata in respect to the piston are, that it shall be steam-tight, or admit of no leakage; and yet that it shall move through the cylinder with the minimum of friction compatible with this condition. The cylinder should be truly bored; but it is never perfectly true, even with the best workmanship; hence the periphery of the piston must be so contrived that it shall be capable of adapting itself to the inequalities the surface presents. In common pistons the packing is of unspun hemp or soft rope, wound about the periphery, and between a plate and flange, and saturated with grease. Metallic pistons of several varieties the first being patented by Cartwright in 1797-are chiefly used in the better class of engines. These are usually of brass or cast

iron, and composed of segments pressed outward by springs radiating from a centre. Steam packing is now much in use, the periphery of the piston being in segments, with cavities within the body of the piston, into which small orifices admit steam directly from the cylinder to secure the requisite pressure. The stuffing box is a packing about the piston rod, where it passes through the cylinder cover. The pipe conveying steam from the boiler to the cylinder is the "steam pipe" S, fig. 4; in it is the "throttle-valve," controlled by the governor G, fig. 1; sometimes a stop-valve, to be shut when the engine is not in use; and a "cut-off" or expansion valve may also be placed in it. The passages leading from the steam pipe into the cylinder are the steam passages, steam ways, or "nozzles," A B, fig. 2, and are opened and closed by the "induction valves;" if the openings to the exhaust pipe (that for escape of the steam after use from the cylinder), E, figs. 2 and 3, are controlled by separate valves, these are the "eduction valves." Very commonly, as will appear, the entrance and exit of steam for both ends of the cylinder are managed by two valves only, or by a single one. The orifices for steam from the passages directly into the cylinder are the "ports;" but this name is quite as commonly given to the opposite ends, oo, fig. 2, of the steam passages, directly under the valves. An escape valve, held by a spring, may be placed at each end of the cylinder, or "blow-through," or cylinder cocks, for discharge of water due to condensation or to "priming," i. e., the carrying over of liquid water from the boiler. The jacket and its use have been referred to; the covering of felt and wood often put upon the cylinder for a like purpose, or over the outer cylinder, is the "clothing," or "cleading." If the cylinder is bare, it is usual to keep it well polished to diminish the cooling that results from radiation. The valves, and sometimes portions of the steam passages, are included within a chest alongside of the cylinder, and of less size, and called the "valve chest," c, fig. 4. For controlling the entrance and exit of steam at the ports, many different sorts of valves are in use. The two classes most commonly employed are the "slide valve," certain varieties of which are called "D valves," and the spindle, "T," or "poppet valve," in certain forms known as the conical or button valve. Flat or double conical valves were the first used by Watt in his earlier engines. The slide valve was applied by Murray in 1799, being adapted from Lavoisier's slider for the air pump with two barrels. It is generally used in European countries, and in this country in locomotive and frequently in marine engines. It acts by sliding upon that inner surface of the valve chest, made accurately plane, which faces the cylinder, and is named the seat, or valve seat. The original or short D valve, still much in use, is shown in fig. 2, its form a partial D, with flat extremities for covering

steam acting on A B from without, A is made larger; acting on CD from between the disks, D is made larger; and under this slight difference of pressures, though each double valve is kept in its seat, it is lifted or depressed by its rod without strain or jar. Many other forms of valve are in use. In engines of which the cylinders oscillate upon trunnions projecting from their sides, now much in use, and known as oscillating engines; and also in rotatory or rotary engines, in which the steam drives round a revolving piston, the methods of supplying steam differ greatly from those above. In oscillating engines the steam and the exhaust passages traverse each a trunnion and support on either side; and they may be self-acting, the movement of the face of the trunnion over that of the support establishing and closing connection of each passage at the right point in the movement of the piston; or a steam chest and valves are employed, controlled by one or two eccentrics. This form of engine has many advantages; but almost insuperable defects in it are the heating and wear of the rubbing surfaces, which, with the spreading effect of the steam pressure, tend to produce and increase leakage of steam. The piston rod of a beam engine is attached by its cross head to one end of the working beam; and the rise and descent of this rod imparts to the ends of the beam a reciprocating movement, which, through the connecting rod from the further end, turns the crank. When the power acts by thrust and pull of an inflexible rod upon a crank, there occurs at the same two points in each revolution a momentary cessation of the rotative impulse. These points of no action are called the "dead points," and the rod and crank standing at either is technically said to be "on the centre." Obviously, the rotative effect upon the crank will, in each half revolution, increase from O at the dead point to a maximum when the rod stands at right angles to the crank, and then diminish in reaching the next from maximum to 0 again. The earliest, and in stationary engines the still almost universal method of continuing the movement steadily over the dead points, is by application of the fly-wheel upon the crank shaft. This is an iron wheel, of large diameter, and the very great weight of which is concentrated almost entirely in its massive rim. Brought to move at a given speed, it serves as a reservoir either of momentum or of inertia; hence tends to equalize the movement of the crank, as well as of the engine and machinery. The diameter of the circle described by the crank necessarily equals the length of stroke of the piston. Another method of equalizing the rotative pressure on the crank, is that of causing a pair of engines to act at once on the same shaft, their pistons impelling a pair of cranks placed at right angles to each other, so that when the action of one crank is 0 the other shall be at its point of greatest force. The effect of the fly-wheel upon the engine or machinery can

only be to prolong the transition from one speed to another; it cannot confine the working of the engine within the limits of variation that shall correspond to a desired average speed. The changes in the load or resistance worked against, as well as in the formation of steam in the boiler, are liable to be sudden and extreme. To confine the speed of the engine within certain limits is the office of the "governor," W, fig. 1. The number of forms now given to this is very great; but the principle of action in all is the same. The centrifugal tendency of heavy balls at the lower and free ends of two bars on an upright axis, is made to move a sliding collar and levers, and through these to control the throttle valve, which is fixed within the steam pipe like an ordinary pipe damper. On account of the friction of its axis and joints, the governor does not respond at once and completely to the speed impressed by the engine, and does not instantly compensate it. The later forms have usually in view to obviate this defect. Among them is the so-named centrifugal governor, in which two small balls swing on short bars from a single joint at the apex of the cone described in their revolution, and are caused to turn at a very high velocity; this arrangement not only works with less friction, but also by the more rapid revolution develops more speedily a centrifugal action competent to overcome that which remains. Evidently, the movement of the piston and its rod through the stroke should be rectilinear, or true in respect to a line which may be regarded as the axis of the cylinder, as otherwise the piston, stuffing box, and rod must be rapidly worn or wrenched out of form. In the engines called vertical, parallelism of movement is usually secured by the ascent and descent of the cross head of the piston rod within guiding grooves, in oscillating engines, the mere rocking of the cylinder continually adjusts it to the line of action on the crank; in other forms, generally, the piston rod is jointed, and parallelism of the rod below the joint, secured either by Watt's parallel motion or by guides. The former is much employed in English beam engines; it consists of an arrangement of levers and linkwork, p m, fig. 1, attaching to fixed points and to the piston-rod joint in such way that, while the end of the beam necessarily moves through the arc of a circle, this joint and the rod below it shall move in almost an exact straight line. The real movement is, with most of these combinations, still in a double curve resembling a much flattened 8. It is because the long strokes and cranks generally preferred in American engines would exaggerate the departure from a right line, that in this country the use of guides to secure parallelism is almost universal. The "condenser" is a steam-tight and air-tight vessel or chamber, usually near the cylinder, into which, through the exhaust pipe, the spent steam from the cylinder is discharged during each movement of the piston, and in which it is condensed, either by circulation of

cold water over the surfaces of the reservoir containing the steam, or by injection of a shower or spray of cold water, or often in both ways at the same time. In condensing land engines, the injection water is supplied from the "cold well" which surrounds or encloses the condenser, and which is kept supplied by the "cold water pump," worked in beam engines by a rod from the beam. In marine engines the water enters directly from the sea by a tube or sort of inverted funnel pierced at top with small orifices, and rendered properly strong. The "blow-through" valves communicate from the cylinder to the condenser, and from the latter a "snifting valve" opens out to the atmosphere; all these are usually shut, but they are opened for blowing through steam in order to expel the air from the cylinder and condenser before the engine is set to work. The vacuum gauge on the condenser shows how much the pressure within it falls below the atmospheric pressure. Residual steam, air, and water are extracted from the condenser by the "air pump," also worked from the beam where that is present, and are discharged into the "hot well," from which by a tube these products are carried to the boiler and supply it with water, thus economizing much of the sensible heat resulting from condensation of the steam. "Surface condensers," now coming much into use, usually consist of a collection of tubes, of larger or smaller size, through or into which the spent steam is caused to circulate, and within which it is condensed by cold water surrounding the tubes, either as a current, or in form of a dense spray over their exterior. Of these the best now in use are probably Pirsson's and Sewell's. The cylinder and passages of the condensing engine being closed against the atmosphere, if a perfect vacnum could be instantaneously produced in advance of each movement of the piston, the total effect of pressure of the entering steam upon the piston would in all cases be utilized. Practieally, this instantaneous and perfect vacuum is never attained; owing to time required for escape through the passages and imperfection of the condensation in the condenser itself, there is always a body of steam of low tension in advance of the piston, producing a residual pressure, and subtracting so much from the effective pressure of the entering steam on the other side of the piston. Still the resistance is less than that of the atmosphere; and a proportionate saving of steam power is realized in condensing engines. With steam, indeed, at one atmosphere, or not much above, the condenser is indispensable. As the steam is generated and applied at higher pressures, the gain by condensation becomes a smaller fraction of the total pressure applied, and the condenser is then, either for safety or economy, less serviceable. So, again, without condensation, the greater the pressure of the steam the less is the fraction or part lost by atmospheric resistance. And, simple as is the principle of the condenser,

its application along with the parts subservient to it imparts to the engine a good degree of its complexity, and greatly increases the weight of materials and the space required for it. Hence, to render the engine portable, the condenser and its appendages had to be discarded; the piston must then be driven both ways against the pressure of the atmosphere. Beside, the steam which had driven the piston must be expelled after each movement of it into the atmosphere; here, first, the heat contained by this steam is wasted, and secondly, since the orifice for escape cannot be made as large as the mouth of the cylinder, there is some delay and compression of the escaping steam, and an added resistance upon the piston due to this cause. In its working, then, the non-condensing engine is the less economical, and requires greater strength; but the saving in original cost, to say nothing of the need of portability, goes far to compensate for those defects; so that this kind of engine has come into general use, not only for locomotives, but very commonly also in the forms of stationary engine, and frequently for steam craft, especially those intended for river navigation. The pumps being dispensed with in the non-condensing engine, the beam may be so likewise; the piston rod being directed by a cross piece and guides, and a crank rod from this cross piece directly impelling a crank with or without a fly-wheel. III. Classification and Varieties of Engines. This branch of the subject has of necessity been in part anticipated in considering the essential parts of ordinary engines. Dividing steam engines with reference to the most essential distinction to be made between them, i. e., the physical principles upon which they are worked, we have the following two classes and principal varieties under each:

Double-acting engines.

66

1. Acting by pressure only.... 2. Acting by pressure and expansion. Secondly, classifying with reference to the kind of movement the engine is to impart, we have: A, non-rotative engines, applying force directly by alternating movements, as in pumping, direct-acting steam hammers, &c.; B, rotative, giving movement to a revolving shaft, now rather the exceptions than the rule. Thirdly, in reference to the general purpose of their use, engines are: A, stationary, for propelling machinery, &c.; B, portable, for removal from place to place, but stationary while in use, as for sawing, pile-driving, &c.; C, marine, for propelling vessels; D, locomotive, for propelling vehicles on land. Fourthly, with reference to the mode of applying the action of the piston rod of ordinary cylinder engines, we have: A, beam engines; as 1, ordinary beam engines, the beam above; 2, "side lever" en