"OFTENTIMES an uncertaintie hindered our going on so merrily, but by persevering the Difficultie was mastered, and the new Triumph gave stronger Heart unto us."­ RALEIGH.

"If everything which we cannot comprehend is to be called an impossibility, how many are daily presented to our eyes! and in condemning as false that which we consider to be impossible, may we not be depreciating a giant's effort to give an importance to our own weakness ? "­MONTAIGNE.

"They who aim vigorously at perfection will come nearer to it than those whose laziness or despondency makes them give up its pursuit from the feeling of its being unattainable."­CHESTERFIELD.

AS has been already stated, the steamengine is a machine which is especially designed to transform energy, originally dormant or potential, into active and usefully available kinetic energy.

When, millions of years ago, in that early period which the geologists call the carboniferous, the kinetic energy of the sun's rays, and of the glowing interior of the earth, was expended in the decomposition of the vast volumes of carbonic acid with which air was then charged, and in the production of a lifesustaining atmosphere and of the immense forests which then covered the earth with their almost inconceivably luxuriant vegetation, there was stored up for the benefit of the human race, then untreated, an inconceivably great treasure of potential energy, which we are now just beginning to utilize. This potential energy becomes kinetic and available wherever and whenever the powerful chemical affinity of oxygen for carbon is permitted to come into play; and the fossil fuel stored in our coal-beds or the wood of existing forests is, by the familiar process of combustion, permitted to return to the state of combination with oxygen in which it existed in the earliest geological periods.

The philosophy of the steamengine, therefore, traces the changes which occur from this first step, by which, in the furnace of the steamboiler, this potential energy which exists in the tendency of carbon and oxygen to combine to form carbonic acid is taken advantage of, and the utilizable kinetic energy of heat is produced in equivalent amount, to the final application of resulting mechanical energy to machinery of transmission, through which it is usefully applied to the elevation of water, to the driving of mills and machinery of all kinds, or to the hauling of " lightning " trains on our railways, or to the propulsion of the Great Eastern.

The kinetic heatenergy developed in the furnace of the steamboiler is partly transmitted through the metallic walls which inclose the steam and water within the boiler, there to evaporate water, and to assume that form of energy which exists in steam confined under pressure, and is partly carried away into the atmosphere in the discharged gaseous products of combustion, serving, however, a useful purpose, en route, by producing the draught needed to keep up combustion.

The steam, with its store of heatenergy, passes through tortuous pipes and passages to the steamcylinder of the engine, losing more or less heat on the way, and there expands, driving the piston before it, and losing heat by the transformation of that form of energy while doing mechanical work of equivalent amount. But this steamcylinder is made of metal, a material which is one of the best conductors of heat, and therefore one of the very worst possible substances with which to inclose anything as subtile and difficult of control as the heat pervading a condensible vapor like steam. The process of internal condensation and reevaporation, which is the great enemy of economical working, thus has full play, and is only partly checked by the heat from the steamjacket, which, penetrating the cylinder, assists by keeping up the temperature of the internal surface and checking the first step, condensation, which is an essential preliminary to the final waste by reevaporation.

The piston, too, is of metal, and affords a most excellent way of exit for the heat escaping to the exhaust side.

Finally, all unutilized heat rejected from the steamcylinder is carried away from the machine, either by the water of condensation, or, in the noncondensing engine, by the atmosphere into which it is discharged.

Having traced the method of operation of the steam-engine, it is easy to discover what principles are comprehended in its philosophy, to learn what are known facts bearing upon its operation, and to determine what are the directions in which improvement must take place, what are the limits beyond which improvement cannot possibly be carried, and, in some directions, to determine what is the proper course to pursue in effecting improvements. The general direction of change in the past, as well as at present, is easily seen, and it may usually be assumed that these will be no immediate change of direction in a course which has long been preserved, and which is well defined. We may, therefore, form an idea of the probable direction in which to look for improvement in the near future.

Reviewing the operations which go on in this machine during the process of transformation of energy which has been outlined, and studying it more in detail, we may deduce the principles which govern its design and construction, guide us in its management, and determine its efficiency.

In the furnace of the boiler, the quantity of heat developed in available form is proportional to the amount of fuel burned. It is available in proportion to the temperature attained by the products of combustion; were this temperature no higher than that of the boiler, the heat would all pass off unutilized. But the temperature produced by a given quantity of heat, measured in heatunits,
is greater as the volume of gas heated is less. It follows that, at this point, therefore, the fuel should be perfectly consumed with the least possible airsupply, and the least possible abstraction of heat before combustion is complete. High temperature of furnace, also, favors complete combustion. We hence conclude that, in the steamboiler furnace, fuel should be burned completely in a chamber having nonconducting walls, and with the smallest airsupply compatible with thorough combustion; and, further, that the air should be free from moisture, that greatest of all absorbents of heat, and that the products of combustion should be removed from the furnace before beginning to drain their heat into the boiler. A firebrick furnace, a large combustionchamber with thorough intermixture of gases within it, good fuel, and a restricted and carefullydistributed supply of air, seem to be the conditions which meet these requisites best.

The heat generated by combustion traverses the walls which separate the gases of the furnace from the steam and water confined within the boiler, and is then taken up by those fluids, raising their temperature from that of the entering " feedwater " to that due the steampressure, and expandling the liquid into steam occupying a greatlyincreased volume, thus doing a certain amount of work, besides increasing temperature. The extent to which heat may thus be usefully withdrawn from the furnacegases depends upon the conductivity of the metallic wall, the rate at which the water will take heat from the metal, and the difference of temperature on the two sides of the metal. Extended "heatingsurface,'' therefore, a metal of high conducting power, and a maximum difference of temperature on the two sides of the separating wall of metal, are the essential conditions of economy here. The heatingsurface is sometimes made of so great an area that the temperature of the escaping gases is too low to give good chimney draught, and a "mechanical draught " is resorted to, revolving "fanblowers " being ordinarily used for its production. It is most economical to adopt this method. The steamboiler is generallyconstructed of iron­sometimes, but rarely, of castiron, although "steel," where not hard enough to harden or temper, is better in consequence of its greater strength and homogelleollsness of structure, and its better conductivity. The maximum conductivity of flow of heat for any given material is secured by so designing the boiler as to secure rapid, steady, and complete circulation of the water within it. The maximum rapidity of transfer throughout the whole area of heatingsurface is secured, usually, by taking the feedwater into the boiler as nearly as possible at the point where the gases are discharged into the chimneyflue, withdrawing the steam nearer the point of maximum temperature of flues, and securing opposite directions of flow for the gases on the one side and the water on the other. Losses of heat from the boiler,by conduction and radiation to surrounding bodies, are checked as far as possible by nonconducting coverings.

The mechanical equivalent of the heat generated in the boiler is easily calculated when the conditions of working are known. A pound of pure carbon has been found to be capable of liberating by its perfect combustion, resulting in the formation of carbonic oxide, 14,500 British thermal units, equivalent to 14,500 X 772 = 11,194,000 footpounds of work, and, if burned in one hour, to 5.6 horsepower. In other words, with perfect utilization, but 0.177, or about onesixth, of a pound of carbon would be needed per hour for each horsepower of work done. But even good coal is not nearly all carbon, and has but about nine tenths this heatproducing power, and it is usually rated as fielding about 10,000,000 footpounds of work per pound. The evaporative power of pure carbon being rated at 15 pounds of water, that of good coal may be stated at 13.5.

In metric measures, one gramme of good coal should evaporate about 13.5 grammes of water from the boilingpoint, producing the equivalent of about 3,000,000 kilogrammetres of work from the 7,272 caulories of heat thus generated. A gramme of pure carbon generates in its combustion 8,080 calories of heat. Per hour and per horsepower, 0.08, or less than onetwelfth, of a kilogramme of carbon burned per hour evolves heatenergy equal to one horsepower.

Of the coal burned in a steamboiler, it rarely happens that more than threefourths is utilized in making steam; 7,500,000 footpounds (1,036,898 kilogrammetres) is, therefore, as much energy as is usually sent to the engine per pound of good coal burned in the steamboiler. The "efficiency " of a good steamboiler is therefore usually not far from 0.7.5 as a maximum. Rankine estimates this quantity for ordinary boilers of good design and with chimneydraught at E = (0.92)/(1.5) in which E is the ratio of weight of fuel burned per square foot of grate to the ratio of heating to grate surface; this is a formula of fairly close approximation for general practice.

The steam in the engine first drives the piston some distance before the induction or steam valve is closed, and it then expands, doing work, and condensing in proportion to work done as the expansion proceeds, until it is finally re leased by the opening of the exhaust or eduction valve. Saturated steam is modified in its action by a process which has already been described, condensing at the beginning and reevaporating at the end of the stroke, thus carrying into the condenser considerable quantities of heat which should have been utilized in the development of power. Whether this operation takes place in one cylinder or in several is only of importance in so far as it modifies the losses due to conduction and radiation of heat, to condensation and reevaporation of steam, and to the friction of the machine. It has already been seen how these losses are modified by the substitution of the compound for the single cylinder engine.

The laws of thermodynamics teach, as has been stated, that the proportion of the heatenergy contained in the steam or other working fluid which may be transformed into mechanical energy is a fraction.

Were the first expression strictly accurate, a hotair engine working from 413.6° Fahr. or 874.8° absolute temperature, down to 122° Fahr. or 583.2° absolute, should have an efficiency of 0.263, transforming that proportion of available heat into mechanical work. The engines of the steamer Ericsson closely approached this figure, and gave a horsepower for each 1.87 pound of coal burned per hour.

Steam expands in the steamcylinder quite differently under different circumstances. If no heat is either communicated to it or abstracted from it, however, it expands in an hyperbolic curve, losing its tension much more rapidly than when expanded without doing work, in consequence both of its change of volume and its condensation. The algebraic expression for this method of expansion is, according to Rankine, PV = C, a constant, or, according to other authorities, from PV = 35 = C to PV = 33 = C. The greater the value of the exponent of V, the greater the efficiency of the fluid between any two temperatures. The maximum value has been found to be given where the
steam is saturated, but perfectly dry, at the commencement of its expansion. The loss due to condensation on the cooled interior surface of the cylinder at the commencement of the stroke and the subsequent reevaporation as expansion progresses is least when the cylinder is kept hot by its steamjacket and when least time is given during the stroke for this transfer of heat between the metal and the vapor.

It may be said that, all things considered, therefore, losses of heat in the stcamcylinder are least when the steam enters dry, or moderately superheated, where the interior surfaces are kept hottest by the steamjacket or by the hotair jacket sometimes used, and where pistonspeed and velocity of rotation are highest.1 The best of compound engines, using steam of seventyfive pounds pressure and condensing, actually require about two pounds of coal per hour­20,000,000 footpounds of energy at the furnace­to develop a horsepower, i. e., about ten times the heat

1 In some cases, as in the Allen engine, the speed of piston has become very high, approaching 800 strokes.

equivalent of the mechanical work which they accomplish. Were the steam to expand like the permanent gases, they would have a theoretical efficiency of about onequarter; actually, the efficiency is nearly sixtenths. The steam- engine, therefore, utilizes about onesixth the heatenergy theoretically available with the best type of engine in general use. By far the greater part, nearly all, in fact, of the fivesixths wasted is rejected in the exhaust steam, and can only be saved by some such method as is hereafter to be suggested of retaining that heat and returning it to the boiler.

The mechanical power which has now been communicated to the mechanism of the engine by the transfer of the kinetic energy of the hot steam to the piston is finally usefully applied to whatever " mechanism of transmission" forms the connection with the machinery driven by the engine. In this transfer, there is some loss in the engine itself, by friction. This is an extremely variable amount, and it can be made very small by skillful design and good workmanship and management. It may be taken at onehalf pound per square inch of piston for good engines of 100 horsepower and upward, but is often several pounds in very small engines. It is least when the rubbing surfaces are of different materials, but both of smooth, hard, closegrained metal, well lubricated, and where advantage is taken of any arrangement of parts which permits the equilibration of pressure, as on the shaftbearings of double and triple engines. The friction of a steamengine of large size and good design is usually between five and seven per cent. of its total power. It increases rapidly as the size of engine decreases.

Having now traced somewhat minutely the growth of the steamengine from the beginning of the Christian era to the present time, having rapidly outlined the equally gradual, though intermittent, growth of its philosophy, and having shown how the principles of science find application in the operation of this wonderful machine, we are now prepared to study the conditions which control the intelligent designer, and to endeavor to learn what are the lessons taught us by science and by experience in regard to the essential requisites of efficient working of steam and economy in the consumption of fuel. We may even venture to point out definitely the direction in which improvement is now progreasing as indicated by a study of these requisites, and may be able to perceive the natural limits to such progress, and possibly to conjecture what must be the character of that change of type which only can take the engineer beyond the limit set to his advance so long as he is confined to the construction of the present type of engine.

First, we must consider the question: What is the problem stated precisely and in its most general form that engineers have been here attempting to solve?

After stating the problem, we will examine the record with a view to determine what direction the path of improvement has taken hitherto, to learn what are the conditions of efficiency which should govern the construction of the modern steamengine, and, so far as we may judge the future by the past, by inference, to ascertain what appears to be the proper course for the present and for the immediate future. Still further, we will inquire, what are the conditions, physical and intellectual, which best aid our progress in perfecting the steamengine.

This most important problem may be stated in its most general, yet definite, form as follows:
To construct a machine which shall, in the most perfect manner possible, convert the kinetic energy of heat into mechanical power, the heat being derived from the combustion of fuel, and syeam being the receiver and the conveyor of that heat.

The problem, as we have already seen, embodies two distinct and equally important inquiries:
The first: What are the scientific principles involved in the problem as stated?
The second: How shall a machine be constructed that shall most efficiently embody, and accord with, not only those scientific principles, but also all of those principles of engineering practice that so vitally affect the economical value of every machine ?

The one question is addressed to the man of science, the other to the engineer. They can be satisfactorily answered, even so far as our knowledge at present permits, after studying with care the scientific principles involved in the theory of the steamengine under the best light that science can afford us, and by a careful study of the various steps of improvement that have taken place and of accompanying variations of structure, analyzing the effect of each change, and tracing the reasons for them.

The theory of the steamengine is too important and too extensive a subject to be satisfactorily treated here in even the most concise possible manner. I can only attempt a plain statement of the course which seems to be pointed out by science as the proper one to pursue in the endeavor to increase the economical efficieney of steamengines.

The teachings of science indicate that success is economically deriving mechanical power from the energy of heat-motion will, in all cases, be the greater as we work between more widely separated limits of temperature, and as we more perfectly provide against losses by dissipation of heat in directions in which it is unavailable for the production of power.

Scientific research, as we have seen, has proved that, in all known varieties of heatengine, a large loss of effect is unavoidable from the fact that we cannot, in the ordinary steamengine, reduce the lower limit of temperature, in working, below a point whieh is far above the absolute zero of temperature­far above that point at which bodies have no heatmotion. The point corresponding to the mean temperature of the surface of the earth is above the ordinary lower limit.

The higher the temperature of the steam when it enters the steam cylinder, and the lower that which it reaches before the exhaust occurs, the greater, science tells us, will be our success, provided we at the same time avoid waste of heat and power.

Now, looking back over the history of the steamengine, we may briefly note the prominent improvements and the most striking changes of form, and may thus endeavor to obtain some idea of the general direction in which we are to look for further advance.

Beginning with the machine of Porta, at which point we may first take up an unbroken thread, it will be remembered that we there found a single vessel performing the functions of all the parts of a modern pumpingengine; it was, at once, boiler, steamcylinder, and condenser, as well as both a lifting and a forcing pump.

The Marquis of Worcester divided the engine into two parts, using a separate boiler.
Savery duplicated that part of the engine of Worcester which performed the several parts of pump, steamcylinder, and condenser, and added the use of water to effect rapid condensation, perfecting, so far as it was ever perfected, the steamengine as a simple machine.

Newcomen and Calley next separated the pump from the steamengine proper, producing the modern steamengine­the engine as a train of mechanism; and in their engine, as in Savery's, we noticed the use of surface condensation first, and subsequently that of the jet thrown into the midst of the steam to be condensed.

Watt finally effected the crowning improvements, and completed the movement of "differentiation " by separating the condenser from the steam-cylinder. Here this process of change ceased, the several important operations of the steamengine now being conducted each in a separate vessel.
The boiler furnished the steam, the cylinder derived from it mechanical power, and it was finally condensed in a separate vessel, while the power which had been obtained from it in the steamcylinder was transmitted through still other parts, to the pumps, or wherever work was to be done.

Watt, also, took the initiative in another direction. He continually increased the efficiency of the machine by improving the proportions of its parts and the character of its workmanship, thus making it possible to render available many of those improvements in detail upon which effectiveness is so greatly dependent and which are only useful when made by a skillful workman.

Watt and his contemporaries also commenced that movement toward higher pressures of steam and greater expansion which has been the most striking feature noticed in the progress of steamengineering since his time. Newcomen used steam of barely more than atmospheric pressure and raised 105,000 pounds of water one foot high with a pound of coal consumed. Smeaton raised the pressure somewhat and increased the duty considerably. Watt started with a duty double that of Newcomen and raised it to 320,000 footpounds per pound of coal, with steam at 10 pounds pressure. Today, Cornish engines of the same general plan as those of Watt, but worked with 40 to 60 pounds of steam and expanding three or four times, do a duty probably averaging, with the better class of engines, 600,000 footpounds per pound of coal. The compound pumpingengine runs the figure up to about 1,000,000.

The increase in steampressure and in expansion since Watt's time has been accompanied by a very great improvement in workmanship­a consequence, very largely, of the rapid increase in perfection, and in the wide range

of adaptation of machinetools­by higher skill and intelligence in designing engines and boilers, by increased pistonspeed, greater care in obtaining dry steam, and in keeping it dry until thrown out of the cylinder, either by steam-jacketing or by superheating, or both combined; it has further been accompanied by a greater attention to the important matter of providing carefully against losses by radiation and conduction of heat. We use, finally, the compound or doublecylinder engine for the purpose of saving some of the heat usually lost in internal condensation and reevaporation, and precipitation of condensed vapor from great expansion.

It is evident that, although there is a limit, tolerably well defined, in the scale of temperature, below which we cannot expect to pass, a degree gained in approaching this lower limit is more remunerative than a degree gained in the range of temperature available by increasing temperatures.(l)

Hence, the attempt made by the French inventor, Dr. Trembly, twentyfive years ago, and by other inventors since, to utilize a larger proportion of heat by approaching more closely the lower limit, was in accordance with known scientific principles.

We may summarize the result of our examination of the growth of the steamengine thus:
First. The process of improvement has been one, primarily, of "differentiation; "2 the number of parts has been continually increased; while the work of each part has been simplified, a separate organ being appropriated to each process in the cycle of operations.
Secondly. A kind of secondary process of differentia

1 The fact here referred to is easily seen if it is supposed that an engine is supplied with steam at a temperature of 400° above absolute zero and works it, without waste, down to a temperature of 200°. Suppose one inventor to adapt the engine to the use of steam of a range from 500° down to 200°, while another works his engine, with equally effective provision against losses, between the limits of 400° and 100°, an equal range vith a lower mean. The first ease gives an efficiency of onehalf, the second threefifths, and the third threefourths, the last giving the highest effect.
2 This term, though perhaps not familiar to engineers, expresses the idea perfectly.

tion has, to some extent, followed the completion of the primary one, in which secondary process one operation is conducted partly in one and partly in another portion of the machine. This is illustrated by the two cylinders of the compound engine and by the duplication noticed in the binary engine.

Thirdly. The direction of improvement has been marked by a continual increase of steampressure, greater expansion, provision for obtaining dry steam, high pistonspeed, careful protection against loss of heat by conduction or radiation, and, in marine engines, by surface condensation.

The direction which improvement seems now to be taking, and the proper direction, as indicated by an examination of the principles of science, as well as by our review of the steps already taken, would seem to be: working between the widest attainable limits of temperature.

Steam must enter the machine at the highest possible temperature, must be protected from waste, and must retain, at the moment before exhaust, the least possible amount of heat. He whose inventive genius, or mechanical skill, contributes to effect either the use of higher steam with safety and without waste, or the reduction of the temperature of discharge, confers a boon upon mankind.

In detail: In the engine, the tendency is, and may probably be expected to continue, in the near future at least, toward higher steampressure, greater expansion in more than one cylinder, steamjacketing, superheating, a careful use of nonconducting protectors against waste, and the adoption of still higher pistonspeeds.

In the boiler: more complete combustion without excess of air passing through the furnace, and more thorough absorption of heat from the furnacegases. The latter will probably be ultimately effected by the use of a mechanically produced draugllt, in place of the far more wasteful method of obtaining it by the expenditure of heat in the chimney.

In construction we may anticipate the use of better materials, and more careful workmanship, especially in the boiler, and much improvement in forms and proportions of details.
In management, there is a wide field for improvement, which improvement we may feel assured will rapidly take by place, as it has now become well understood that great care, skill, and intelligence are important essentials to the economical management of the steamengine, and that they repay, liberally, all of the expense in time and money that is requisite to secure them.

In attempting improvements in the directions indicated, it would be the height of folly to assume that we have reached a limit in any one of them, or even that we have approached a limit. If further progress seems checked by inadequate returns for efforts made, in any case, to advance beyond present practice, it becomes the duty of the engineer to detect the cause of such hinderance, and, having found it, to remove it.

A few years ago, the movement toward the expansive working of high steam was checked by experiments seeming to prove positive disadvantage to follow advance beyond a certain point. A careful revision of results, however, showed that this was true only with engines built, as was then common, in utter disregard of all the principles involved in such a use of steam, and of the precautions necessary to be taken to insure the gain which science taught us should follow. The hinderances are mechanical, and it is for the engineer to remove them.

The last remark is especially applicable to the work of the engineer who is attempting to advance in the direction in which, as already intimated, an unmistakable revolution is now progressing, the modification of the modern steam-engine to adapt it safely and successfully to run at the high pistonspeed, and great velocity of rotation which have been already attained and which must undoubtedly be greatly exceeded in the future. As there is no known and definite limit to the economical increase of speed, and as the limit set by practical conditions is continually being set farther back as the builder acquires greater skill and attains greater accuracy of workmanship and the power to insure greater rigidity of parts and durability of wearing surfaces, we must anticipate a continued and indefinite progress in this direction­a progress which must evidently be of advantage, whatever may be the direction that other changes­not even excepting the change of type to be presently referred to­may take.

It is evident that this adaptation of the steamengine to great speed of piston is the work now to be done by the engineer. The requisites to success are obvious, and may be concisely stated as follows:

1. Extreme accuracy in proportions.

2. Perfect accuracy in fitting parts to each other.

3. Absolute symmetry of journals.

4. Ample area and maximum durability of rubbing surfaces.

5. Perfect certainty of an ample and continuous lubrication.

6. A nicely calculated and adjusted balance of reciprocating parts.

7. Security against injury by shock, whether due to the presence of water in the cylinder or to looseness of running parts.

8. A " positivemotion " cutoff gear.

9. A powerful but sensitive and accuratelyworking governor determining the degree of expansion.1

l The author is not absolutely confident on the latter point. It may be "found more economical and satisfactory, ultimately, to determine the point of cutoff by an automatic apparatus adjusting the expansiongear by reference to the steampressure, regulating the speed by attaching the
governor elsewhere. The author has devised several forms of apparatus of the kind referred to.

10. Wellbalanced valves and an easyworking valvegear

11. Small volume of " deadspace," " clearance," and probably considerable "compression."

It would seem sufficiently evident that the engine with detachable (" drop") cutoff valvegear must, sooner or later, become an obsolete type, although the substitution of springs or of steampressure for gravity in the closing of the detached valve may defer greatly this apparently inevitable change. The "engine of the future " will not probably be a " drop cutoff engine."

As regards the construction of the engine as a piece of mechanism, the principles and practice of good engineering are precisely the same, whether applied in the designing of the compound or of the ordinary type of steamengine. The proportioning of the two machines to each other in such manner as to form an effective whole, by procuring approximately equal amounts of work from both, is the only essential peculiarity of compoundengine design which calls for especial care, and the method of securing success in practice may be stated to be, for both forms of engines,
as follows:

1. A good design, by which is meant­
a. Correct proportions, both in general dimensions and in arrangement of parts, and proper forms and sizes of details to withstand safely the forces which may be expected to come upon them.

b. A general plan which embodies the recognized practice of good engineering.

c. Adaptation to the specific work which it is intended to perform, in size and in efficiency. It sometimes happens that good practice dictates the use of a comparatively un-economical design.

2. Good construction, by which is meant­

a. The use of good material.

b. Accurate workmanship.

c. Skillful fitting and a proper " assemblage" of parts.

3. Proper connection with its work, that it may do its work under the conditions assumed in its design.

4. Skillful management by those in whose hands it is placed.
In general it may be stated that, to secure maximum economical efflciency, steam should be worked at as high a pressure as possible, and the expansion should be fixed as nearly as possible at the point of maximum economy for that pressure.(l) It is even more disadvantageous to cut off too short than to "'follow ' too far." 'with Considerable expansion, steam jacketing and moderate superheating should be adopted, to prevent excessive losses by internal condensation and reevaporation; and expansion should take place in double cylinders, to avoid excessive weight of parts, irregularity of motion, and great loss by friction.

External surfaces should be carefully covered by non-conductors and nonradiators, to prevent losses by conduction and radiation of heat. It is especially necessary to reduce backpressure and to obtain the most perfect vacuum possible without overloading the airpump, if it is desired to obtain the maximum efficiency by expansion, and it thenbecomes also very necessary to reduee losses by "deadspaces " and by badlyadjusted valves.The pistonspeed should be as great as can be sustained with safety.

The expansionvalve gear should be simple. The point of cutoff is perhaps best determined by the governor. The valve should close rapidly, but without shock, and should be balanced, or some other device should be adopted to mak it easy to move and free from liability to cutting or rapid wear.

1 In general, the number of times which the volume of steam may be expanded in the standard singlecylinder, highpressure engine with maximum economy, is not far from .5/P, where P is the pressure in pounds per square inch; it rarely execeds 0.76 P. This may be exceeded in doublecylinder engines.

The governor should act promptly and powerfully, and should be free from liability to oscillate, and to thus introduce irregularities which are sometimes not less serious than those which the instrument is intended to prevent.

Friction should be reduced as much as possible, and careful provision should be made to economize lubricants as well as fuel.

The Principles of SteamBoiler Construction are exceedingly simple; and although attempts are almost daily made to obtain improved results by varying the design and arrangement of heatingsurface, the best boilers of nearly all makers of acknowledged standing are practically equal in merit, although of very diverse forms.In making boilers, the effort of the engineer should evidently be:

1. To secure complete combustion of the fuel without permitting dilution of the products of combustion by excess of air.

2. To secure as high temperature of furnace as possible.

3. To so arrange heatingsurfaces that, without checking draught, the available heat shall be most completely taken up and utilized.

4. To make the form of boiler such that it shall be constructed without mechanical difficulty or excessive expense.

5. To give it such form that it shall be durable, under the action of the hot gases and of the corroding elements of the atmosphere.

6. To make every part accessible for cleaning and repairs.

7. To make every part as nearly as possible uniform in strength, and in liability to loss of strength by wear and tear, so that the boiler when old shall not be rendered useless by local defects.

8. To adopt a reasonably high "factor of safety" in proportioning parts.

9. To provide efficient safetyvalves, steamgauges, and other appurtenances.

10. To secure intelligent and very careful management.

In securing complete combustion, the first of these desiderata, an ample supply of air and its thorough intermixture with the combustible elements of the fuel are essential; for the second­high temperature of furnace­it is necessary that the airsupply shall not be in excess of that absolutely needed to give complete combustion. The efficiency of a furnace in making heat available is measured E = (T-T')/(t-t) in which E represents the ratio of heat utilized to the whole calorific value of the fuel, T is the furnacetemperature, T' the temperature of the chimney, and t that of the external air. The higher the furnacetemperature and the lower that of the chimney, the greater the proportion of heat available. It is further evident that, however perfect the combustion, no heat can be utilized if either the temperature of the chimney approximates to that of the furnace, or if the temperature of the furnace is reduced by dilution approximately to that of the chimney. Concentration of heat in the furnace is secured, in some cases, by special expedients, as by heating the entering air, or as in the Siemens gasfurnace, heating both the combustible gases and the supporter of combustion. Detached firebrick furnaces have an advantage over the "fireboxes " of steamboilers in their higher telnperature; surrounding the fire with nonconducting and highly heated surfaces is an effective method ofsecuring high furnacetemperature.

In arranging heatingsurface, the effort should be to impede the draught as little as possible, and so to place them that the circulation of water within the boiler should be free and rapid at every part reached by the hot gases. The directions of circulation of water on the one side and of gas on the other side the sheet should, whenever possible, be opposite. The cold water should enter where the cooled gases leave, and the steam should be taken off farthest from that point. The temperature of chimneygases has thus been reduced in practice to less than 300° Fahr., and an efficiency equal to 0.75 to 0.80 the theoretical has been attained.

The extent of heatingsurface simply, in all of the best forms of boiler, determines the efficiency, and in them the disposition of that surface seldom affects it to any great extent. The area of heatingsurface may also be varied within very wide limits without very greatly modifying efficiency. At ratio of 25 to 1 in flue and 30 to 1 in tubular boilers represents the relative area of heating and grate surfaces as chosen in the practice of the bestknown builders.

The material of the boiler should be tough and ductile iron, or, better, a soft steel containing only sufficient carbon to insure melting in the crucible or on the hearth of the meltingfurnace, and so little that no danger may exist of hardening and cracking under the action of sudden and great changes of temperature.

Where iron is used, it is necessary to select a somewhat hard, but homogeneous and tough, quality for the firebox sheets or any part exposed to flames.

The factor of safety is invariably too low in this country, and is never too high in Europe. Foreign builders are more careful in this matter than our makers in the United States. The boiler should be built strong enough to bear a pressure at least six times the proposed workingpressure; as the boiler grows weak with age, it should be occasionally tested to a pressure far above the workingpressure, which latter should be reduced gradually to keep within the bounds of safety. In the United States, the factor of safety is seldom more than four in the new boilers, frequently much less, and even this is reduced practically to one and a third by the operation of our inspectionlaws.

The principles just enunciated are those generally, perhaps universally, accepted principles which are stated in all textbooks of science and of steamengineering, and are accepted by both engineers and men of science.

These principles are correct, and the deductions which have been here formulated are rigidly exact, as applied to all types of heatengine in use; and they lead us to the determination, in all cases, of the " modulus " of efficiency of the engine, i. e., to the calculation of the ratio of its actual efficiency to that efficiency which it would have, were it absolutely free from loss of heat by conduction or radiation, or other method of loss of heat or waste of power, by friction of parts or by shock.

The best modern marine compound engines sometimes, as we have seen, consume as little as two pounds of coal per horsepower and per hour; but this is but about onetenth the power derivable from the fuel, were all its heat thoroughly utilized. This loss may be divided thus: 70 percent. rejected in exhausted steam; 20 per cent. lost by conduction and radiation and by faults of mechanism and design; and only the 10 per cent. remaining is utilized. Thirty percent. of the heat generated in the furnace is usually lost in the chimney, and of the remainder, which enters the engine, 20 percent at most is all which we can hope to save any portion of by improvements effected in our best existing type of steamengine. It has already been shown how the engineer can best proceed in attempting this economy.

As it is easy thus to show that the existing common type of steamengine, even if perfect as a piece of mechanism, necessarily wastes a very large proportion of the heatenergy which is supplied to it, it follows that no possible improvement, short of a complete change of type, can greatly increase the efficiency of the best modern engine.

Now the question arises: Can we not eject the saving of some part of that immense proportio n of heat which we have been taught to believe to be invariably and unavoidably wasted ?
Let us see what is the one essential feature, of all heatengines in use, which defeats all attempts at such economy:

Heatengines may be divided, for present purposes, into three principal classes, according to their disposition of rejected heat:
I. Those which restore all heat rejected from the working cylinder to the reservoir from which it was derived.
II. Those which restore a part of the unutilized heat of the working fluid, discharging the remainder from the system and allowing it to be wasted.
III. Those which waste all heat rejected from the working cylinder.
No existing type of steamengine belongs to the first of the classes specified. Some forms of air and gas engines are theoretically assignable to that class, as, by means of some form of "regenerators" they store up rejected heat and restore it to the succeeding charge of working fluid as it enters the cylinder. Actually, however, these engines cannot perform this part of their task thoroughly, and they are thus really to be catalogued in Class II. Nearly all heatengines, including the steamengine, are most correctly assignable to Class III. In the steamengine, the rejected heat, on leaving the steamcylinder, is thrown into the condenser or discharged into the atmosphere, and, in either case, is wasted. A small portion is usually saved by supplying the boiler from the hotwell or from heaters in which it has acquired some increase of temperature by transfer from exhausted steam. In so far as this takes place, they fall into Class II. They will here be considered as belonging to Class III.

All actual steamengines may therefore be considered as belonging to one primary class. They all take steam from a source having a high temperature, degrade the heat thus obtained to a lower temperature, doing work and consuming a definite amount of heat in the production of a definite
amount of mechanical energy, and finally discharge unutilized heat into the atmosphere or into a large volume of condensing water, by which it is carried out of the system and thrown away.

The steamengine differs, aS has already been seen, radically from air and gas engines of its own class in one respecnt; and this difference, although passed over as apparently unimportant by accepted authorities on the theory of the steam-engine, has constituted a serious and hitherto unsurmounted difficulty in the process of calculation of the exact efficiency of the steamengine. As has been seen, the steamengine should be more economical than the air or gas engine working between the same limits of temperature.

That it is not, can only be due to the cause of loss already adverted to as peculiar to the steamengine; or, more correctly, to heatengines, in which the working fluid changes its physical state within the working limits of temperature and pressure. In consequence of the facts stated, it is seen that the usually accepted method of estimating economy, as in gasengines, is not correct as applied to the steamengine. Let us estimate the true theoretical economy of a steam-engine:

Taking the feedwater into the boiler at a temperature of 110° Fahr., which is a very usual temperature of feed, the total heat of steam supplied to the engine at a pressure of 90 pounds per square inch above a vacuum will be 1,211.5 ­110 = 1,101.5. The thermal contents of each pound of steam rejected from the cylinder at 2 pounds pressure, the temperature being 126.27, will be, if reckoned from the original temperature of feed, 1,152.5­110 = 1,042.5. The heat wasted with each pound of water so rejected will be 126.27­110 = 16.27 thermal units.

The actual weights of steam and of the water rejected, per pound of steam supplied to the engine, may be readily calculated. The volume of one pound of steam, at the temperature and pressure at which it enters our steamcylinder, is 4.72 cubic feet. The value of P V § =C At = 90 X 144 .4.72t= 102.611. The volume of steam exhausted from the cylinder­expansion being taken to have been carried so far that the pressure at the opening of the exhaust valve is 2 pounds per square inch­must be V = (Gil= (102.611-at 28144)1 = 76.53. But the volume of 1 pound of steam at this pressure is 172.4 cubic feet. The quantity of uncondensed steam rejected from the cylinder, per pound of steam supplied, is, therefore, 76.531 . 172.4 =0.444 pound. The weight of water rejected is (172.4­7.5) .172.4 = 05.56 pound. The steam carries with it 1,042.5 X 0.444 = 462.85 thermal units, and the water loses 16.27 x 0.556 = 9.05 units. The sum, 471.90 units, is the total amount of heat wasted by rejection from the engine. The heat which has been utilized is 1,101.5­471.90= 630. The efficiency, E, is (1,101.5­471.9) . 1,101..5 = 0.57, or nearly sixtenths.

As has been seen, had the working fiuid been a permanent gas, the effieiency would have been onefourth. The difference illustrates the superiority of the eondensible vapor as a medium for conversion of heat into work. Assuming perfect efficiency, aside from the defect just estimated, this engine should require but (1,980,000 . 1,101.5X772) ­. 0.57 = 4.09 pounds of steam, or about twofifthsof a pound of coal per hour and per horsepower. An actual engine, working steam in this manner, usually consumes at least five times this amount. The difference between the estimated theoretical and the actual efficiency has usually been supposed to be much less, and it has been stated by the writer, as well as by other engineers, that the range left for improvement, by modification of the structure of the common engine, is much less than it is here shown to be.

The direction in which further improvement must take place in the standard type of engine is plainly that which shall most efficiently check losses by internal condensation and reevaporation by the transfer of heat to and from the metal of the steamcylinder. The condensation of steam doing work is evidently not a disadvantage, but, on the contrary, a decided advantage.

To secure this vitally important economy, it is advisable to seek some practicable method of lining the cylinder with a nonconducting material.1 The loss will also be reduced by increasing the speed of rotation and velocity of piston. Where no effectual means can be found of preventing contact of the steam with a good absorbent and conductor of heat, it will be found best to sacrifice some of the efficiency due to the change of state of the vapor, by superheating it and sending it into the cylinder at a temperature considerably exceeding that of saturation. With low steam and slowly-moving pistons, it is better to pursue the latter course than to attempt to increase the efficiency of the engine by greater expansion.

Increasing steampressure and expanding to a greater extent wig give theoretically slightly increased economy. Repeating the calculation made above, using as constants the temperature, volumes, and pressures of steam entering the cylinder at 250 pounds, and expanding down to the same point, it will be found that the gain in efficiency is but a few per cent.

Now, could a change of type be secured by which all heat could be utilized, giving a theoretical efficiency of unity, we might expect an actual efficiency of at least one

1 This plan, as has been seen, was adopted by Smeaton, in constructing Newcomen engines a century ago. Smeaton used wood on his pistons, and Watt tried wood as a material for steamcylinder linings. That material is too perishable at temperatures now common, and no metal has yet been substituted, or even discovered, which answers the same purpose.

half, and thus obtain a horsepower by the expenditure of five pounds of steam, and the combustion of a halfpound of coal per flour. The best existing type of engine has been seen to demand four times this amount, and it is very common for engines to require ten times as much.

It now remains to be determined whether there is any way by which these losses of rejected heat can be avoided. There are two forms of engines of Class I., in which­were it possible to fully avail ourselves of them­all this waste of energy may be avoided:

Type A. The working fluid, if expanded from the temperature and pressure of the boiler or reservoir quite down to the absolute zero, would have all its heatenergy transformed into mechanical work, and there would be no waste. The efficiency would be perfect.

Type B. All heat rejected from the cylinder unutilized may be gathered up and restored to the boiler, there to serve as a basis upon which to pile a new stock of transformable heatenergy, instead of being, as now, rejected from the system entirely and lost. This done, there could be no loss, as all heat leaving the machine would be transmitted to exterior bodies as mechanical energy. Nothing being lost as heat, the efficiency of the engine would be unity and its economy a maximum.

Forms of steamengines may be conceived in which these methods (of saving heat now wasted) may be applied. Practically, however, it is evident, the first form of these two ideal engines can never be made successfully, since it would require to be made of such immense size that all the power derivable from it would be insufficient to move it.

Complete utilization of all heat is not practicable with the heatengine of Type A under any known, or even conceivable, conditions in actual construction.l Type E, in which all unutilized heat is restored to the reservoir, comprehends engines of two very different kinds. These are:

1. Heatengines in which the rejected heat is transferred from the mass of working substance discharged from the working cylinder at the end of the stroke of the piston, to a mass of metal or other heatabsorbillg material, and from the latter again transferred to another and a new charge of working fluid which is about entering the reservoir to take up a new stock of heatenergy.

2. Heatengines in which, as in the ideal engines already described, the rejected working fluid itself, with its contained unutilized heat, is all returned to the reservoir.

Engines of the first of the two kinds into which Type B is here divided, and in which the working fluid is usually air, have frequently been designed. The mechanism by which the heat is stored and restored by transfer has been :called the " Regenerator."

The regenerator is entirely inapplicable to engines in which steam and other saturated vapors are employed. The rejected heat, in these engines, is principally conveyed from the working cylinder in the form of latent heat of vaporization, and can only be removed from the exhausted charge

l The two kinds of fluid, working in engines of the Class A, are seen to give two characteristic products of the tevo known processes of abstraction of heat from fluids: 1. Forcibly compress the fluid, removing the heat of compression as rapidly as it is generated; the product is. after all heat has been removed, a mass of maximum density, probably solid in all cases, and at the absolute zero of temperature. 2. Abstract the heat contained originally in the mass, by causing its expansion against a lesistanec; the product is a diffused mass, without tension, of minimum possible density, and at the absolute zero of temperature.

by the condensation of the steam. But this condensation involves the degradation of the rejected heat to a temperature at which it is no longer transferable to the entering charge of the working substance. The use of a regenerator is therefore out of the question in engines in which vapors of liquids similar to saturated steam are employed, except when, as in Siemens's "Regenerative SteamEngine," the vapor is highly superheated.

As the regenerator system is inapplicable here, it is necessary to secure the return of that unutilized heat by restoring to the reservoir all exhausted gas or liquid, or both, without seeking to remove from them and to restore separately the heat with which they are charged. Fortunately, the conditions which make the use of the regenerator system impracticable are favorable to the adoption of the expedient which has been described as characterizing the operation of engines of Type B.

As has been seen, the expansion of steam doing work results in the condensation of a portion of the vapor. The weight of steam condensed bears a proportion to the weight of steam supplied, which is greater as the work done by expanding that steam becomes a greater proportion of the mechanical equivalent of the thermal contents of the steam when supplied to the working cylinder.

Suppose the expansion to be carried so far that onehalf of the vapor becomes liquid. Imagine the water of condensation separated from the uncondensed vapor, and the two masses returned to the boiler separately by compressingpumps. The return of the water would involve the expenditure of an amount of work measured by the product of the volume of that water into the difference of pressures in the boiler and in the receiver into which the exhaust has been discharged. The return of the uncondensed steam would involve not only the return of the charge of heat contained in that steam, but also the return of all the energy which the vapor had yielded during its expansion, since an equivalent amount of heat would be generated by its recompression to boilerpressure.

Under the conditions now assumed, it is evident that only that portion of the heat entering the engine which is surrendered by the condensation of steam doing work can be utilized. It is also evident that, in this form of engine, no heat can be lost; and, consequently, that the engine of Type B. which is operated as just indicated, will have yielded the exact equivalent of the net amount of heat expended upon it. All heat rejected from the working cylinder unutilized being returned to the boiler, there to form " a basis on which to pile up a new stock of utilizable energy," the engine is a "perfect engine" in a broader sense than that adopted by Carnot. It is further evident that perfect efficiency is given for all ranges of temperature, and that what working fluid shall be adopted, and what temperatures shall be chosen, will be determined simply by practical conditions to be ascertained by experiment.

The work done in restoring the water of condensation to the boiler being so small in amount that it may be neglected, the ratio of the total work done in restoring rejected heat to the work done by the steam on the piston will be the ratio of the weight of steam restored to the boiler of the engine to that supplied to the working cylinder at each stroke. Altering an engine of the common type, having an efficiency of 0.50, into an engine of Type B. the quantities of work done during expansion and during compression will have the ratio, nearly, of 1: (1­0.50) = 0.50.

The indicator diagram of our new engine will be similar to that shown in the accompanying figure, in which the ordinates of the curve measure the pressures, and the abscissa are proportional to the volumes of the expanding flnid. The steam expands from the volume g O to the volume f O, when doing work, and from the pressure O a or b g to the pressure c A, when expanding according to Boyle's law. Doing work, however, the gradual condensation of the fluid reduces the pressure during expansion, and the line becomesb k d instead of b c. When the attempt is made to restore the rejected fluid to the boiler, the compression would naturally cause the line d k b to be retraced; but having removed the water of condensation from the engine separately, it cannot add, by its reevaporation, to the tension of the steam under compression, and the latter must behave more nearly like a perfect gas, causing the line to take the direction to be. The net result is the production of mechanical

energy represented by the area e b k d I e, and the expenditure of an equivalent amount of heatenergy, plus the amount of energy required to return the water of condensation to the boiler. Had, in this case, the steam been cut off at an earlier point in the stroke, as at e the area e n m would have represented the quantity of heat transformed into work. It is evident that, the more the steam is expanded the greater will be the proportion of steam condensed, and the less the amount remaining to be compressed to boilerpressure, and the less the weight of steam required to be supplied to the engine per unit of work done. It follows that, to secure an engine combining high efficiency with small volume, it will be advisable to employ steam of high pressure, and to expand it as much as may be found practicable. The new type of engine can, if at all, probably only supersede the common form when engineers can employ steam of very high pressure, and adopt much greater range of expansion than is now usual. Great velocity of piston and high speed of rotation are also essential in the attempt to make this revolution in steam-engine construction a success.

When this possible " engine of the future " is likely to be introduced, if at all, can be scarcely even conjectured. It seems evident that its success is to be secured, if its introduction is ever attempted, by the adoption of high steam-pressures, of great pistonspeeds, by care and skill in design, by the use of exceptionally excellent materials of construction, by great perfection of workmanship, and by intelligence in its management. There seems no tangible obstacle to its introduction.(1)

Experiment and experience will probably lead gradually to the general and safe employment of much higher steam-pressures and very greatly increased pistonspeeds, and may ultimately reveal and remove all those difficulties which must invariably be expected to be met here, as in all other attempts to effect radical changes, however important they may be. We are continually learning that the use of steam as a medium for transformation of heat into mechanical energy is more advantageous than it had been supposed. The corrected estimates of the efficiency of the steamengince, already given, show it to be capable of far more perfect utilization than has been generally supposed possible. It is not impossible that even these estimates give a very inade

1 The idea first suggested itself to the mind of the writer in 1858'59, when at Brown University, and while designing a peculiar form of "drop cutoff engine," which was intended to work with exceptional economy. The plans then conceived have gradually assumed a different shape, but still embody the essential principles here outlined. Mr. C. E. Emery, in 1868, proposed a similar plan.

quate idea of the true efficiency of the perfect steamengine. We know almost nothing of the physical properties of steam which are mostly concerned in its utilization in heatengines at the high pressures which we know are attainable; and it is impossible to say that the modifications of specific heat and of pressures at such high temperatures may not be such that it may prove a vastly more efficient working substance for the heatengine, at such temperatures and pressures, than we are led to consider it from our knowledge of its properties at the now usual working pressures.(1)

l Vide a paper "On a New Type of Steam Engine," etc., by the author, Journal of the Franklin Institute, OctoberDecember, 1877, and in the "Transactions of the American Association for the Advancement of Science," 1877, in which this proposed type of ongoing is treated of at greater length, and in which this portion of these lectures was first published.

Previous Chapter

Title Page