Significant Scots
John Elder - Engineer and Ship Builder

MEMOIR OF JOHN ELDER

The earliest accounts of the family of Elder show it to have flourished in the county of Kinross, in the east of Scotland, during the sixteenth and seventeenth centuries. The leading "branch of the family seems to have been that which possessed the estate of Arlarie, near the town of Milnathort. There are on record the names of two John Elders of Balbughtie, cadets of the family of Arlarie, one of whom lived in the sixteenth, and the other in the seventeenth century, both forefathers of the subject of this Memoir.

The line of his direct ancestors for nearly two centuries affords a remarkable example of a fact which is more common than is usually supposed— the hereditary transmission of skill and talent; for they all practised that art from which (as Fairbairn tells us) mechanical engineering has sprung — that of the millwright — and were all remarkable for ability and success.

The first of those regarding whom we possess definite information was Alexander Elder, wright at Craigo, about two miles west of Milnathort, born towards the end of the seventeenth century. He married Marion Ireland. His son David Elder, born in 1724, was a wright at Little Seggie, in the same neighbourhood; and from note-books of his, which are still preserved, he appears to have been a man of talent and information, and to have possessed considerable knowledge of mathematical and mechanical science. He was cut off in 1756, at the early age of thirty-two, leaving, by his wife Ellen Henderson, three sons and three daughters. His eldest son, Alexander, born at Little Seggie about 1748, carried on at that place, and afterwards at Milnathort, the same business. He married Elizabeth Morrison, by whom he had two sons and three daughters, and died at Milnathort in December 1823.

In his eldest son, David Elder, the talent of the race, handed down through so many generations, began to achieve public distinction. He was born at Little Seggie on the 7th of January 1784, and died at Glasgow on the 31st of January 1866, at the close of a vigorous old age, being then in his eighty-second year. A memoir of his life by Mr James E. Napier, in which full justice is done to his remarkable character and abilities, was published in the ' Transactions of the Institution of Engineers in Scotland' for 1865-66; and therefore it is sufficient now to recapitulate the leading events of his career only. He learned the practical part of his trade as an apprentice to his father, and its scientific principles by the private study of mathematical books during intervals of leisure.

In 1808 he succeeded to his father's business at Little Seggie, which he quitted a few years afterwards for Paisley; and in 1817 he removed to Glasgow, thus obtaining a wide field for the exercise of his knowledge and skill as a millwright and mechanical engineer. In 1821 he became manager of the works of Mr Eobert Napier, which office he continued to hold until induced by advancing age to retire.

In 1812 he married Grace, daughter of Mr John Gilroy; and the subject of the present Memoir was their third son.

John Elder was born at Glasgow on the 8th of March 1824. His elementary education was obtained in the High School of Glasgow. It does not appear that he applied himself to the study of the ancient classics; hut the result of his training in English scholarship became manifest in after-life; for in writing and speaking on those practical and scientific subjects which he understood so well, he showed himself master of a clear, concise, and energetic style of expression.

In arithmetic and mathematics, he was a pupil of Dr Connell, one of the most able and successful teachers of the time; and here he at once gave proofs of extraordinary talent and application, carrying off the principal prizes of the class.

In every branch of drawing—an art intimately connected with mechanical science—he was a most successful student.

The studies before mentioned constituted the principal part of his early school education. A constitution naturally delicate prevented him from deriving the full benefit of his attendance at the High School of Glasgow, and from pursuing his studies to any considerable extent at a university. A short attendance at the class of civil engineering in Glasgow College was all the university education he received. He was fortunate in being educated under the eye of his father, whose extensive information and high capacity were devoted to the training of his son, and under whose judicious advice he prosecuted his private studies with that ardour which was so marked a characteristic of his later years. The scientific knowledge of which he gave proof in after-life was not only varied and extensive, hut was complete and exact, and free from the defects in thoroughness and accuracy which often beset self-taught scholars.

To those who knew him well, and enjoyed the advantage of personal communication with him, it was manifest that his eminence was due not so much to teaching by others as to the fact that John Elder was that rare character—a man of genius; and therefore in a great measure independent of that external control and guidance which are necessary for the training of ordinary students. In other words, his mind was gifted with the faculty of subjecting itself to that systematic labour and discipline which has to be enforced in ordinary cases by academic authority, and with that strong and clear vision which gives the learner the power of finding his way through the mazes of science without a guide.

He acquired, as his father also had done, considerable knowledge and practical skill in music, especially that of the organ.

He served his apprenticeship of five years as an engineer in the works of Mr Robert Napier, under the direction of his father, working successively in the pattern-shop, moulding-shop, and drawing-office. He was then employed for about a year as a patternmaker in the works of Messrs Hick at Bolton-le-Moors, and afterwards as a draughtsman on the works of the Great Grimsby Docks.

Before 1849 he returned to the works of Mr Napier to take charge of the drawing-office, a most important appointment.

In the summer of 1852, the firm formerly called Randolph, Elliott, & Co. of Glasgow, well known and of high standing as millwrights, was joined by Mr Elder as a partner, and undertook the business of marine engineering, which they had never practised before, and which they were now enabled to undertake through possessing a partner with a thorough knowledge of the principles and practice of that branch of applied mechanics. The firm from that time was known by the designation of "Randolph, Elder, & Co." until 1868; and subsequent to the retirement of the other partners, its name became at first "John Elder," and then "John Elder & Co."

About 1860 the firm added shipbuilding to the other branches of its business.

The career of Mr Elder as a marine engineer and shipbuilder is so closely connected with the application of the compound expansive steam-engine to the propulsion of ships, that it now becomes necessary to introduce into this Memoir a brief explanation of the principles of that class of steam-engines, and a summary of their history from the time of their first invention.

In every machine a certain quantity of energy, or power of doing work, is expended, in order that a certain amount of work may be done. In every machine, and under all circumstances, the whole work done is equal to the energy expended; but only part of that work is useful, the remainder being useless, so that the energy expended in doing it is wasted. For example, in a pumping steam-engine the useful work consists in raising, in a given time, a certain quantity of water to a certain height: the useless, or wasteful work, is that done in overcoming friction. The proportion which the useful work done bears to the energy expended is called the efficiency. In an absolutely perfect machine, the efficiency would be represented by unity—but no such machine exists; and in every actual machine, the efficiency is expressed by a fraction which falls short of unity by an amount corresponding to the energy that is wasted.

In a steam-engine there are several successive causes of waste of energy. In the first place, the whole of the energy which the fuel is capable of producing by its combustion is not communicated to the water in the boiler, but only a certain fraction of that energy, ranging in ordinary cases from six-tenths to eight-tenths: this fraction is the efficiency of the boiler; and the amount by which it falls short of unity corresponds to the heat lost by imperfect combustion, by conduction and radiation, and by the high temperature at which the furnace-gas escapes through the chimney.

Secondly, the whole of the energy which in the form of heat is communicated to the water in the boiler, so as to raise its temperature and convert it into steam, is not obtained in the form of mechanical work done by the steam in driving the piston. In fact, it has for some years been known, through the progress of the science of thermodynamics, that the work done by the steam in driving the piston (often called the indicated work, because its amount can be registered by a self-acting instrument called the indicator) corresponds to a quantity of energy which has disappeared from the form of heat, "being the difference between the heat brought by the steam from the boiler, and the heat carried away by the same steam when it leaves the cylinder. That difference, in every case which can occur in practice, is but a small fraction of the whole heat brought by the steam from the boiler, such as a twentieth, or a tenth; and that fraction is the efficiency of the steam.

Thirdly, the whole of the energy exerted by the steam in driving the piston is not communicated to the machine which it is the purpose of the engine to drive; for a fraction of that energy, say from an eighth to a fourth, is wasted in overcoming the friction of the engine—the difference between that fraction and unity being the efficiency of the mechanism.

Fourthly, when the machine which it is the purpose of the engine to drive is an instrument for propelling a ship, a fraction of the energy is wasted in agitating the water in which the propeller works, the remainder only being usefully expended in overcoming the resistance of the vessel, and driving her ahead; and the ratio which this last remainder bears to the whole energy expended by the engine in driving the propeller is a fraction called the efficiency of the propeller.

The efficiency of the whole combination, made up of furnace, boiler, engine, and propeller, is found by multiplying together the four fractions already mentioned—viz., the efficiency of the boiler, the efficiency of the steam, the efficiency of the mechanism, and the efficiency of the propeller—and is of course a smaller fraction than any of the factors of which, it is the product.

The object of improvements in the economy of the marine steam-engine is to increase as far as practicable, consistently with due regard to economy in first cost, each of the four factors of the efficiency.

Judgment, as well as skill, is specially required in applying to practice in marine steam-engineering improvements whose objects are to increase the mechanical efficiency of the furnace and boiler, of the steam in the cylinder, and of the mechanism; for those improvements for the most part tend more or less to increase the cost of construction; and thus there arises in each case the commercial question, "Whether the economy in working to be attained by means of a given increase of efficiency is sufficient to warrant the additional expenditure? In decidiug that question, regard must be had to many different^-circumstances—such as the length of the voyage, the intended speed, the price of fuel, and the nature of the traffic, ' For example, it would be a waste of

money and labour to make elaborately-designed boilers and engines of very high efficiency for vessels intended to run short trips between places where coal is cheap and abundant; while for ships designed to make long voyages, with few and distant coaling stations, and expensive fuel, every improvement that increases efficiency may be a profitable investment. It is not sufficient, then, for success in the business of marine engineering, that the engineer should possess knowledge of the mechanical principles of his art, and skill in their practical application—for these qualifications alone might lead him into needless expense in the production of a degree of mechanical efficiency not required by the circumstances of particular cases; he ought also to have a sound judgment regarding the commercial result of the adaptation of engines of a given kind to a given vessel, intended for a given trade.

Those different qualifications are so seldom found united in one man, that the tendency of popular opinion is to regard them as incompatible, and to look especially upon the knowledge, skill, and enterprise which lead an engineer to adopt new or unusual improvements in practice as being fraught with danger to his success in business; and so no doubt they are, unless regulated by commercial sagacity.

The success of Mr Elder and of his firm proved that his commercial sagacity was not inferior to his knowledge, skill, and enterprise, and that his was one of those rare minds in which was realised that uncommon combination of talent.

It now becomes necessary to point out more in detail the nature of the improvements which ]\Ir Elder, by himself, or with the co-operation of his firm, carried out in the practice of marine engineering ; and as the most important of these were connected with the second and third factors of efficiency already referred to—that of the steam in its action on the piston, and that of the mechanism— the circumstances on which that second factor depends will, in the first place, be explained.

The expenditure of energy in the form of heat required in order to produce a given weight of steam, when the water is supplied to the boiler at a given temperature, increases when the pressure and temperature of the steam increase, but at a comparatively slow rate. Eor example, the expenditure of heat required to produce a given weight of steam at the pressure of ten atmospheres (about 147 lb. on the square inch of absolute pressure, or 132.3 lb. on the square inch above the atmosphere), the feed-water being at the temperature of about 100° Fahrenheit, is greater than that required to produce an equal weight at the atmospheric pressure, in the proportion only of 1.04 to 1, or 26 to 25 nearly. Hence the problem of obtaining the greatest possible quantity of indicated work from a given expenditure of heat in producing steam, is so nearly identical with that of obtaining the greatest possible quantity of work from a given weight of steam, that in practice the difference between those two problems may be neglected.

All mechanical work is done by the exertion of a force through a space, and is calculated and expressed as a quantity by multiplying the mean amount of the force into the space through which it acts. In the case of steam, the space is the distance through which the piston is driven in a given time; the force is the excess of the forward pressure exerted by the steam behind the piston as it comes from the boiler, and afterwards expands, above the backward pressure exerted by the steam in front of the piston while it is being expelled from the cylinder into the condenser in condensing engines, or into the atmosphere in non-condensing engines. In a non-condensing engine the back-pressure is a little greater than that of the atmosphere, say from 15 lb. to 18 lb. on the square inch. In a condensing engine the back-pressure is lower than that of the atmosphere, to an extent depending on the efficiency with which the condenser acts (or on the goodness of the vacuum, as it is commonly called), and ranges in ordinary cases from 3 lb. to 5 lb. on the square inch.

In an expansive steam-engine, the forward pressure exerted by the quantity of steam that is admitted behind the piston at each stroke has two stages in its action—the admission and the expansion. During the admission the steam is coming from the boiler into the cylinder, and it exerts a pressure less than that in the boiler only by the amount required to overcome the friction of pipes, passages, and valve-ports : say about a twelfth of the absolute pressure in the boiler in ordinary cases. The admission is terminated by the cut-off—that is, by the closing of the valve which admits the steam into the cylinder. Then follows the expansion of the steam which is confined in the cylinder; and during this stage of its action, it goes on occupying a continually increasing space as it drives the piston before it, and exerting a continually diminishing pressure. The exact law according to which the pressure diminishes while the steam expands is complicated, and is different under different circumstances as to heat. For ordinary practical calculations, however, it is sufficiently accurate to assume the simple approximate" law that the pressure varies inversely as the volume, falling to one-half of its original intensity when the volume is doubled, to one-third when the volume is trebled; and so on.

It is obvious that work continues to be done by the steam in driving the piston so long as the pressure behind the piston, or forward pressure, continues to be greater than the pressure in front, or back-pressure, exerted by the steam which has already done its work, and which" the piston is expelling from the cylinder; and hence it follows, that in order to realise the greatest quantity of work which the steam is capable of performing, the expansion ought to be carried on until the forward pressure of the steam behind the piston has fallen so low as to be just sufficient to overcome the back-pressure, and that to end the expansive working of the steam at an earlier period of the stroke is to throw away part of the power of the steam.

This statement must, however, be taken with the qualification that when the excess of the forward pressure above the back - pressure falls below the pressure which is just sufficient to overcome the friction," the work done is no longer partly useful and partly wasteful, but is wholly wasteful; whence it follows that, although in order to obtain the greatest indicated work from a given weight of steam the expansion should be continued until the forward pressure becomes just equal to the back-pressure, the greatest useful work is obtained by making the expansion cease when the forward pressure is just equal to the back-pressure added to a pressure equivalent to the friction of the engine.

Another obvious principle is, that both the indicated and the useful work obtained from a given weight of steam must be the greater the greater the proportion in which the forward pressure exceeds the back-pressure. To take an extreme case: If the mean forward pressure be simply equal to the backpressure, no indicated work whatsoever is obtained from the steam; and if the mean forward pressure is simply equal to the back-pressure added to the friction, no useful work is obtained. Hence the higher the forward pressure, and the lower the back-pressure, the greater is the efficiency of the steam in an engine; and as the pressure increases and diminishes with the temperature, the same principle may be otherwise expressed by saying that the temperature of the steam on its admission ought to be as high as possible, and that in a condensing engine the temperature in the condenser, on which the back-pressure depends, ought to be as low as possible. In a non-condensing engine, the back-pressure, as formerly-stated, is a little above that of the atmosphere.

The foregoing principle, as applied to the temperature in a condensing engine, was first distinctly stated by James Watt; and he invented the separate condenser as a means of carrying it into effect.

The pressure at which the steam is admitted is limited only by the strength and safety of the boiler. In Watt's time, he, in common with most other engineers, was very cautious in the use of high pressures ; and he therefore relied more on a low backpressure than on a high forward pressure for the efficiency of his engines. Improvements in the construction of boilers, and experience of their safety under high pressures when properly designed and managed, have caused subsequent engineers to become gradually bolder in the use of such pressures.

In order to realise the greatest theoretical efficiency in the expansive working of steam, the expansion ought to take place in a non-conducting cylinder, with a non-conducting piston. This condition cannot be absolutely realised in practice; but means may be taken to diminish the loss of efficiency arising from the conducting power of the cylinder and piston until it becomes unimportant.

If that loss arose solely from the waste of heat by its passage through the metal of the cylinder to the air outside, it would be sufficient for its practical prevention to clothe the cylinder with bad conductors, such as wood and felt. But by far the greater part of that loss arises in a different and more complex way, which was not thoroughly understood until about 1849 or 1850, when the consequences of the disappearance of heat in performing mechanical work were demonstrated. Until that time it was erroneously believed, from reasoning based on the hypothesis of caloric, that a given weight of steam, after performing work by expansion, contained exactly as much heat as before, and was therefore superheated; because the quantity of heat sufficient to keep it in the vaporous state at the higher pressure was more than sufficient to produce the same effect at the lower pressure; and that statement was so confidently believed that it was distinctly laid down as a fundamental principle in all, or almost all, writings on the theory of the steam-engine.

One of the earliest consequences deduced from the principles of thermodynamics was, that when steam performs work by expansion, a quantity of heat disappears sufficient not only to lower the temperature of the steam to that corresponding to its lowered pressure, but to cause a certain portion of the steam to pass into the liquid state. The steam thus spontaneously liquified collects in the form of water in the cylinder; and if the cylinder and piston were made of a non-conducting material, that water would simply be discharged from time to time into the condenser, without causing any waste of heat. But the cylinder and piston, being made of a conducting material, give out heat to the liquid water which adheres to them, so as to re-evaporate it when the communication with the condenser is opened; and that heat is carried off to the condenser with the exhaust-steam, leaving the piston and the inside of the cylinder at a low temperature, even though the outside of the cylinder should be clothed with an absolute non-conductor. When steam from the boiler is admitted at the beginning of the next stroke, part of it is immediately liquified through the expenditure of its heat in raising the piston and the inside of the cylinder again to a high temperature, the result being that at the end of the second stroke the quantity of liquid water which is re-evaporated, and carries off heat to the condenser, is greater than it was at the end of the first stroke. At each successive stroke that quantity augments until it reaches a fixed amount, depending mainly on the difference of the temperatures of the steam at the beginning and end of the expansion; and the effect is the same as if a certain quantity of steam at each stroke passed directly from the boiler to the condenser without performing work. In some experiments lately made, the quantity of steam which thus ran to waste was found to be greater than that which performed work; so that the expenditure of steam was more than doubled.

The remedy for this cause of loss is to prevent that spontaneous liquifaction of the steam during its expansive working, in which the process just described originates; and that is done either by enclosing the cylinder in a jacket or casing supplied with hot steam from the boiler, or by superheating the steam before its admission into the cylinder; or by both those means combined. The steam is thus kept in a nearly dry state, so as to be a bad conductor of heat; and the moisture which it contains, though sufficient to lubricate the piston, is not allowed to increase to such an extent as to carry away any appreciable quantity of heat from the metal of the cylinder and piston to the condensers.

The steam-jacket outside the cylinder was invented and used by "Watt. "Whether he fully understood the nature of its action can never be known; for he did not publish any reason for using it except that of keeping the steam as hot as possible. Its real action was certainly not understood by Watt's immediate successors, nor indeed by any one, until the principles of thermodynamics were applied to the question about twenty years ago; and many engineers, reasoning correctly from the erroneous hypothesis of caloric, concluded that the steam-jacket was unnecessary, and abandoned its use. The fact of liquid water collecting in the cylinder was known, but was ascribed to "priming," or the carrying of spray from the boiler. The use of the steam-jacket was retained in a few special kinds of engines, such as the Cornish pumping-engines; and in them the economy properly due to high rates of expansion of the steam was realised; but in almost all other engines, and certainly in marine engines, the jacket was abandoned, with this result—that little or no practical advantage was found to result from expansive working when the steam was expanded to more than about double, or two and a half times its original volume; and this became a received maxim amongst engineers, and especially amongst marine engineers, for its truth in the case of unjacketed cylinders was established by practical experience, as well as by experiments made for the purpose of testing it.

The jacketing of the piston, by filling its internal hollow with hot steam from the boiler, was invented by M. Normand of Havre, and introduced into Britain by Mr Davison at a comparatively recent date, after the action of the steam-jacket had been explained according to the principles of thermodynamics, and its use revived in practice.

So far as the theoretical action of the steam on the piston is concerned, it is immaterial whether the expansion takes place in one cylinder, or in two or more successive cylinders. The advantage of employing the compound engine is connected with those causes which make the actual indicated work of steam fall short of its theoretical amount, and also with the strength of the engine and its framing, the steadiness of its action, and the friction of its mechanism.

The force exerted by the steam on the piston of an engine is transmitted by the piston-rod to the moving pieces of the machinery which it drives—such as the connecting-rod, crank, and crank-shaft; and by the bearings of the moving pieces it is transmitted to the framework. It produces straining actions on all those pieces, moving and fixed; and each of them must be made strong enough to bear safely the straining action produced, not by the mean or average force exerted by the steam, but by the greatest force. The mean force which the steam has to exert on the piston depends on the power required to do the work of the engine, and on the mean speed of the piston; and the greater the rate of expansion, the greater is the inequality between the greatest force and the mean force, and the stronger must the engine be made. For example, when the steam is expanded to twice its original volume, its pressure during its admission is about once and a fifth its mean pressure; when to five times, its pressure during admission is about double of its mean pressure ; and when to ten times, its pressure during admission is about three times its mean pressure; so that in this last example, if the engine is single cylindered, all parts of the mechanism and framing that are strained by the force of the steam must be made three times as strong as they would require to be in an engine of the same power working without expansion. That additional strength involves not only additional cost of construction, but additional friction, because of the greater size of the bearings ; and thus the economy of power due to expansion is partly neutralised.

It was to obviate this disadvantage in the use of high rates of expansion that the earliest form of compound steam-engine was contrived by Horn-blower in 1781. That engine was single-acting, and adapted to the pumping of mines ; it had two cylinders, standing side by side, and having their pistons hung from the same end of the walking-beam; the larger cylinder was of the dimensions suited for a single-cylinder engine of the same power and speed; but instead of admitting the steam at its comparatively high initial pressure to act upon the large area of the piston of that cylinder, and thus to exert a great straining force, it was admitted in the first place into the smaller cylinder, so as to exert a straining force equal to the initial pressure multiplied by the area of the smaller piston only; and after having done part of its work by expansion in the smaller cylinder, it was transferred to the larger cylinder in a state of increased volume and diminished pressure to complete its action there. The cylinders were called the high-pressure and low-pressure cylinders respectively, and the same terms are still used in describing compound engines.

The same principle of action was applied by "Woolf to engines with "Watt's separate condenser, and to double - acting steam - engines ; and consequently compound engines came to be very generally known as " "Woolf's engines."

In "Woolf's form of the compound engine, as well as in Hornblower's, the two piston-rods are hung from the same end of a walking-beam, so that the forces exerted through them act in the same direction at the same time; and the straining actions produced on the framing and mechanism are those due to the sum of those forces. The same is the case in those forms of direct-acting compound engines for marine purposes in which the high and low pressure piston-rods are hung from one cross-head. Hence, although the straining actions of the two rods are, in a well-designed engine of the construction just mentioned, less than in a single-cylindered engine of equal power, they are not so small as they may be made to become by causing the straining actions due to the two forces to oppose each other. This improvement, so far as the straining actions on a walking-beam and its bearings are concerned, was introduced by M'Naught, who hung the two piston-rods from the opposite arms of the walking-beam, so as to make the difference, instead of the sum of their straining forces, act on the main centre. The sum, however, of those forces still acts on the bearings of the shaft in M'Naught's engine, in the direct-acting engines already referred to, and in the forms of compound engine described in Mr Craddock's treatise on that subject. That book was published in 1847, and contains the descriptions and drawings of compound engines adapted to marine, locomotive, and other purposes, as patented by liim at different times from 1840 to 1846.

Craddock's compound engine, as described by him in the treatise just mentioned, is direct-acting. The high and low pressure cylinders, placed side by side, are not exactly parallel to each other, but make a small angle in order to enable the engine to " pass the centre." The two piston-rods are connected with one crank; upon which, therefore, and upon the shaft and its bearings, they exert a straining action due to the resultant of their forces, which, though not quite, is very nearly equal to their sum.

Craddock's compound engine, as described in his treatise, is further defective through the absence of steam-jackets, which are now known to be essential to the realising of the economy properly due to high rates of expansion; and unless that economy be fully realised, the additional cost and complexity of a compound engine are thrown away.

It is true that in some steamers fitted with Craddock's engines, or engines resembling them, at a later date (viz., in 1858 and subsequently) the straining actions of the pistons were opposed to each other, and the cylinders were jacketed; but this was long after the time at which the proper principles of the construction of compound marine engines had been brought into practical use by Messrs Randolph, Elder, & Co.

In 1850 a peculiar form of compound steam-engine called the " continuous expansion engine" was patented by Mr Nicholson. The pistons of the high and low pressure cylinders drive two cranks at right angles to each other; and the straining action is the resultant of those due to the forces acting through the two rods. This form has considerable advantages in certain cases ; but it was not brought into practical use till about six or seven years later.

It results, then, from the history of marine steam engineering, that previous to the formation of the firm of Randolph, Elder, & Co., the compound steam-engine had not been successfully applied in Great Britain to the propulsion of vessels; that compound engines such as Craddock's had been proposed for that purpose, but had not been designed so as fully to realise the advantages of that form of engine; that the abandonment of the steam-jacket in the practice of almost all marine engineers had made it useless, if not wasteful, to employ those high rates of expansion to which the compound engine is suited; and that as this practical error originated in an erroneous theory of the mechanical action of heat, founded on the hypothesis of substantial caloric, then universally prevalent, it was not to be expected that it should be reformed except by an engineer who had studied and understood the principles of the then almost new science of thermodynamics.

Such an engineer was Mr Elder. He knew, in common with other practical men, the fact that when high rates of expansion were used with a view to economy of fuel, their economical action was defeated by the gathering in the cylinders of large quantities of liquid water, which evaporated when the exhaust-port opened, and carried away heat to the condenser; but he had learned also—what was known to very few practical men fifteen years ago—that the formation of that liquid water originated in the disappearance of heat during the performance of work by the expansion of the steam, and that the remedy was to supply the cylinder with additional heat to replace that which so disappears, by returning to the practice of Watt and the Cornish engineers, and resuming the use of the steam-jacket.

Mr Elder had also mastered a subject which, before his time, had been almost wholly neglected, and which even now does not always meet with the attention that it deserves, and that is, the diminution of the friction of the engine by causing the forces which drive the shaft round to balance and neutralise, as far as possible, each other's actions on the bearings where the friction takes place. As an elementary illustration of this subject, suppose that a shaft is made to rotate by means of a single force applied to a single crank-pin. The whole of that force will be transmitted to the bearings, and will there produce a pressure which will cause a certain amount of friction in addition to that produced by the weight of the shaft. But if we now divide the force required to drive the shaft into two equal forces of half the amount, and apply them in opposite directions to a pair of cranks exactly opposite to each other, those two driving forces will balance each other as regards pressure on the bearings, and the friction will be that due to the weight of the shaft alone. It is impossible in practice to realise this balance of driving forces with absolute precision, but an approach to it can always be made. One of the most important advantages of compound cylinder engines with opposite cranks is their enabling that balance of driving forces to be approximately realised ; and that advantage had been neglected or very imperfectly developed before Messrs Eandolph, Elder, & Co. constructed their marine engines, which in this respect were a great improvement upon all compound engines previously invented.

The careful attention which Mr Elder had bestowed on the friction of engines and the means of diminishing it, is fully shown in an unpublished lecture which he delivered before the United Service Institution in April 1866. He there takes a practical example of a marine engine, and shows by detailed calculation how from 10 to 15 per cent of the whole indicated power of an engine may be wasted in unnecessary friction through neglect of proper arrangements for the mutual balancing of the forces exerted on the shaft. In fact, he took a more correct view of the real advantages of the compound engine than had previously been done by any practical engineer; regarding it as a means, not so much of increasing the indicated power produced by a given expenditure of steam, as of diminishing that waste of power which causes the effective power to fall short of the indicated power.

In the lecture already referred to, Mr Elder points out under what circumstances it becomes advantageous to employ a compound engine rather than a single-cylinder engine—viz., when the rate of expansion exceeds four. He adds that, should rates of expansion greater than nine be used, it will become advisable to expand the steam in three successive cylinders instead of two.

Most of the improvements introduced by Messrs Randolph, Elder, & Co. in marine engineering were secured by a series of patents, of which the following is a summary—the patents being distinguished by letters ; and it is also shown which of those patents were taken in the names of both partners, and which in the name of one only.

A. Charles Randolph and John Elder—dated 24th January 1853. An arrangement of compound engines adapted to the driving of the screw-propeller. The engines are vertical, direct-acting, and geared. The pistons of the high and low pressure cylinders move in contrary directions, and drive diametrically opposite cranks, with a view to the diminution of strain and friction.

B. John Elder—dated 28th February 1854. For an improved arrangement of the parts of horizontal direct-acting condensing engines for screw-steamers.

C. Charles Randolph and John Elder—dated 15th March 1856. This describes an arrangement of compound engines which was applied with most successful results to a long series of steamers. There are two diametrically opposite cranks and four cylinders, making a pair of compound engines; the high and low pressure cylinder of each engine lie side by side in an inclined position, and their pistons move in contrary directions; and this arrangement not only promotes the balance of driving forces, but enables the steam to pass from the high pressure to the low pressure cylinder in the most direct manner possible, without having to traverse long crooked passages as it did in Hornblower's and Woolf's engines.

The directions in which the cylinders of the two engines lean are contrary—that is to say, for example, in a paddle-wheel steamer the forward engines incline backwards, and the after engines forward; and in a screw-steamer the starboard and port engines lean respectively to starboard and to port, so that their piston-rods make with each other an angle which, in different engines, ranges from 60° to 90°. The whole arrangement is one of the most simple and compact that is possible in a pair of compound engines, and it produces as near an approach to a balance of driving forces as is practicable when each engine has two cylinders only.

An ingenious contrivance for reversing the engine is described, consisting in an arrangement of epicyclic gearing, whereby a loose eccentric is made when required to overrun the shaft until it reaches the position for backward gear.

The specification fully states the importance of providing each cylinder with a steam-casing or jacket to prevent liquifaction: but this is not claimed; for it was not a new invention, but, as has been already explained, the revival of a practice which had fallen into neglect, though essential to the economical use of high rates of expansion.

D. John Elder—dated 29th January 1858. The specification of this patent describes an arrangement of cylinders in the compound engine by which a nearly perfect balance of driving forces is obtained, and not merely a good approximation to such balance, as in the arrangements previously described; and consequently it may be regarded as embodying the principles of the construction of steam-engines of which Mr Elder approved, in their most complete form, calculated to realise the greatest possible efficiency of the mechanism as well as of the steam. There are three cranks on the shaft—two pointing diametrically opposite to the third, which lies between them. Each engine has three cylinders, lying parallel to each other and side by side ; in the middle is the high-pressure cylinder, whose piston drives the middle crank; at its two sides are a pair of low-pressure cylinders, whose pistons move simultaneously in the contrary direction to that of the middle cylinder, and drive the other two cranks. Thus the resultant of the forces exerted through the two low-pressure piston-rods is not merely contrary in direction, but directly opposed to the force exerted through the high-pressure piston-rod; and if the rates of expansion in the high and low pressure cylinders are properly adjusted to their dimensions, there is an exact balance of the actions of those forces on the bearings.

When there is only one low-pressure cylinder, as in the engines described under patent G, the forces exerted through the two piston-rods may be equal and contrary, but they are not directly opposed, because they are exerted at different points in the shaft; and hence the balance of driving forces cannot be quite exact.

Two or more three-cylindered compound engines can be placed at suitable angles of inclination to each other, so as to drive one shaft, as in the arrangement of two-cylindered engines described in specification G.

As the three-cylindered compound engine is somewhat more expensive than a two-cylindered compound engine of the same power, it has been used only in certain cases where special economy of power was desired. Its success in practice will be described further on.

In specification D, as well as in specification C, the importance of the steam-jacket is mentioned; but, for the reason already stated, that part of the engine is not claimed.

E. John Elder—dated 7th June 1858. This patent is for a very simple but very important improvement —the making of paddle-floats of plates of iron or steel, bevelled to a sharp edge, instead of thick wooden planks. The broad edges of wooden paddle-floats oppose a resistance to the plunging them into and drawing them out of the water; and the inventor considered that the substitution for them of comparatively thin sharp-edged metal plates caused a gain of from 4 to 6 per cent in the speed of a given vessel with engines of a given power. This invention was perfectly successful in practice, and was applied to several steamers in the course of the year in which the patent was obtained, and it still continues to be put in practice by the firm with beneficial results.

F. Charles Randolph and John Elder—dated 28th April 1859. This patent is for a variety of improvements in engines and boilers, which it is unnecessary to describe in detail. Amongst other inventions, it describes the making of a boiler with two or more uptakes, in order to increase the surface for superheating the steam.

G. John Elder—dated 15th October 1859. This patent relates to details of mechanism for moving slide-valves.