- 
 
* 
 
 LIBRARY 
 
 'V 
 
 UN 
 
 IVERSITY OF CALIFORNIA 
 
 Received 
 Accessions No.^Z'fX'P Shelf No 
 
STEAM USING; 
 
 OR 
 
 Steam Engine Practice, 
 
 CHAS. A. SMITH, 0. E., 
 
 Professor of Civil and Mechanical Engineering at Washington University, St. 
 
 Louis, Mo.; Member of the American Society of Civil Engineers, the 
 
 Engineers' Club of St. Louis, and Associate Member of the 
 
 American Association of Railway Master Mechanics. 
 
 CHICAGO: 
 THE AMERICAN ENGINEER, 182-184 DEARBORN STREET. 
 
 1885. 
 
Entered according to act of Congress, in the year 1884, by * 
 
 PROPRIETORS OP THE AMERICAN ENGINEER, 
 in the Office of the Librarian of Congress, at Washington, D. C. 
 ^ 
 
 PRESS OK 
 
 JOHN W. WESTON, 
 CHICAGO, ILL. 
 
PUBLISHERS' PREFACE. 
 
 It is believed that some knowledge of the circumstances attending the 
 publication of this work, 'STEAM USING, "as well as its companion volume, 
 "STEAM MAKING," will be of interest to the reader. 
 
 The lamented author, Prof. Chas. A. Smith, had arranged with The 
 American Engineer for the publication of the two works. While the first, 
 "STEAM MAKING, " was going through the columns of the Engineer, Pro- 
 lessor Smith died, early in 1884, leaving also to the care of the Engineer 
 the recently completed manuscript of "STEAM USING." 
 
 To all who are familiar with the circumstances under which the books 
 were written the author suffering from a mortal illness and struggling 
 against death to thus round out his life work, only giving up to die on 
 their completion will appreciate and value the more highly the broad 
 and active experience thus crystallized. 
 
 To Mr. John W. Weston, so long connected with this journal, and 
 personally familiar with the>author and his writings, has been delegated 
 the pleasant duty of conducting these works through the various stages 
 of bookmaking, with the result now presented. The task has not been 
 without its difficulties, the most serious, perhaps, being the loss of the 
 invaluable assistance of their author in the work of revision of matter 
 and proof. 
 
 As the author died without leaving a preface to his second volume, we 
 desire to acknowledge for him the aid which he received from various en- 
 gine builders and professional men throughout the country, which the 
 following pages will disclose 
 
 It has been the aim, as far as possible, to preserve the exact style of 
 the author, and it is believed that the facts and features presented in both 
 books, the heirlooms of an admirable man, acknowledged to be profound 
 and exact in his particular lines of work, will be held to cover whatever 
 defects of minor importance may be encountered. 
 
 THE AMERICAN ENGINEER. 
 
 CHICAGO, MARCH 1, 1885. 
 
s& 
 
 SKETCH OF THE LIFE AND CHARACTER 
 OF THE AUTHOR. 
 
 Charles A. Smith was born in St. Louis, October 1, 1846. His parents 
 were both Massachusetts people who had been still further west. From 
 both father and mother he inherited the instincts of a sailor, and the 
 blood of several generations of ship-masters coursed through his veins. 
 Though he never became a sailor, he always showed a sailor's fondness for 
 "fixing things," for using his hands, for actual construction. 
 
 While he was still an infant, his mother died of cholera in St. Louis, 
 and he was placed in the care of his father's sister, in Newburyport, Mass. 
 This kind aunt was his mother, and her house was his home till he had a 
 home of his own. His mode of life was simple and plain, but young 
 Smith made warm friends and his boyhood was happy. 
 
 I first met him in 1860, when I became principal of the Boys' High 
 School, of Newburyport. He was then fourteen years old and a member 
 of the second class. He was a pleasant little fellow with a frank, earnest 
 look, and a forehead which suggested, brains. When the school gave 
 expression to its loyalty to the Union by the erection of a liberty pole and 
 publicly celebrated a flag-raising, young Smith was selected by his school- 
 mates to mount the platform and haul home the stars and stripes. 
 
 The school had a very good theodolite, and when we came to Loomis' 
 Surveying, a great enthusiasm for field work was developed, and young 
 Smith was never so happy as when on a surveying party. He took the 
 English course and graduated in 1862. The next spring he went into the 
 office of J. B. Henck, civil engineer, in Boston. At that time he probably 
 had no idea of going to an engineering school. In 1864 he was leveller on 
 the Boston, Hartford & Erie Railway. In 1865 he became chief assistant 
 in the City Engineer's office, Springfield, Mass. By this time he saw 
 clearly that an engineer requires a training far beyond a high school 
 education, and he resolved to enter the Massachusetts Institute of Tech- 
 nology, then first opened He had been reading ahead somewhat, with 
 occasional help from me, so that he entered what was organized as a 
 sophomore class. He lived again in Newburyport and went eighty miles 
 daily on his way to and from the Institute. President Rogers was his 
 teacher in physics, Professor Runkle in mathematics and applied mechan- 
 ics, and Professor Henck in civil engineering. 
 
 He graduated in the pioneer class in 1868. I never quite understood 
 how he managed to meet the cost of his course at the Institute. To be 
 sure he had carefully saved the earnings of three years, and he secured 
 
VI. STEAM USING; OK, STEAM ENGINE PRACTICE. 
 
 for his vacations most excellent employment under the celebrated hy. 
 draulic engineer, J. B. Francis, at Lowell, Mass. He there assisted in 
 determining the flow of water in pipes, over wiers, the efficiency of tur- 
 bines, etc. I left Massachusetts for St. Louis in 1865, so I did not follow 
 closely his career as a student. 
 
 After a year as engineer on the Union Pacific Railway in Utah, he 
 returned, on the completion of the road, to Boston and went into partner- 
 ship with Professor J. B. Henck, as civil engineers. While there associated 
 with Professor Henck, he took charge of a part of the Blue Ridge Rail- 
 way of North Carolina, as division engineer. 
 
 At that time, in 1870, the steady development of the Polytechnic 
 School of Washington University made it necessary to appoint an instruc- 
 tor of civil engineering. I took pleasure in recommending young Smith 
 for the position, and he was appointed. For the first year he made his 
 home in my family, and as a preparation for the work of the class room he 
 read with me Eankine's Civil Engineering entire. 
 
 After a brief experience as instructor, Mr. Smith was appointed pro- 
 fessor to the chair of civil and mechanical engineering, which was subse- 
 quently named in honor of William Palm. This chair Professor Smith 
 held till June, 1883, when compelled by his last illness to resign. 
 
 Though devoted at all times to the work of his professorship, Profes- 
 sor Smith found time to mingle in matters of practical engineering. For 
 five years he was consulting engineer of the Iron Mountain Railway, 
 among other things designing the DeSoto shops, and building a new pier 
 in the Black river. In a similar way he was associated with Messrs, 
 Shickle, Harrison & Co., designing the arched ribs of the roof over the 
 Chamber of Commerce, and the iron trestles of the Bessemer Iron Works. 
 Professor Smith was engaged as consulting engineer for the construction 
 of the water works of Hannibal, of St. Charles, in Missouri, and of Ames- 
 bury, Massachusetts. His last professional duties were in connection with 
 the last named. The pumping works at Richmond, Va., were designed by 
 him, his plans being entered in competition and receiving the first prize. 
 In 1879 he spent his summer vacation as resident engineer of the Balti- 
 more Bridge Company, building piers in the Mississippi river just below 
 Minneapolis. 
 
 Without attempting to give a full list of the professional enterprizes 
 of Professor Smith, I have said enough to show how tireless a worker he 
 was, and how closely he studied the practical details of engineering. But 
 it was in connection with the St. Louis Engineers' Club that his devo- 
 tion and enthusiasm were most fully shown. He was an active mem- 
 ber for twelve years, and the secretary for nine or ten years. The club 
 has not always been as flourishing as it is now. It has had its seasons of 
 depression when only the zeal and the courage of Secretary Smith seemed 
 to hold it together. Nothing but the direst necessity compelled him to 
 yield at last. 
 
 The fatal malady, which in the shape of a cancerous tumor, brought 
 his life to an untimely close on the 2nd of February, 1884, was born, as he 
 
LIFE AND CHARACTER OF THE AUTHOR. VII. 
 
 thought, of hard work, of exposure, and of physical neglect. He could 
 scarcely stop to eat or sleep; it was work first and comfort last. 
 
 Nothing in Professor Smith's life was more heroic than the way he 
 battled for two years against an impending fate. When too weak to stand 
 before his class, he taught reclining upon a lounge. One of his last 
 pupils speaks in a notice of his beloved professor of "the days of suffer- 
 ing spent in his study in the University, when we gathered round him 
 as he lay on the lounge, unable to stand, and listened to his exposition 
 of 'Economic Location,' taking as a basis the work of his friend, Arthur 
 Wellington." 
 
 In January, 1883, he was forced to give up his class work altogether, 
 and to keep his room. Still he was not idle. Lying on the bed, or reclin- 
 ing in an easy chair, he was hard at work upon his two books on "Steam 
 Making" and "Steam Using," which are just now being issued by the 
 American Engineer, in Chicago. The first was finished by the end of 
 1882, and arrangements were made for its publication, but the prospect 
 for the second book was gloomy enough. Nevertheless, he worked 
 at it with a terrible earnestness which no unfavorable symptom could 
 diminish. Nay, though clinging to the faintest glimmer of hope of 
 returning health, he toiled at his book with the resolute air of one who 
 was fully conscious that his days were numbered, and that the book must 
 speedily be finished. In spite of pain and the dark shadow of the inevit- 
 able, his mind seemed clear and his hand steady. In the spring of '83 
 he moved back to Newburyport, Mass., to be near his physician and his 
 family friends. There in a quaint old house, in a quiet neighborhood of 
 that quiet town, he finished his book, laying down his pen and the burden 
 of life at the same time. The readers of "Steam Using" may be glad to 
 know that the author's very life's blood went into that book; that it was 
 the last, the most perfect fruit of a very active and noble life. 
 
 Professor Smith is a good example of a poor boy who made his own 
 way; who fought his own battles; who earned and honored every position 
 he took. He was always a student. Some of you will remember with 
 what enthusiasm he studied quarternions and thermodynamics; with 
 what zeal and success he read all that he could get on graphical statics, 
 and how many important additions he suggested. The records of the 
 St. Louis Club probably will show that Professor Smith has presented 
 more papers than any other member, past or present. 
 
 As an engineer, Professor Smith was bold and trustworthy. His confi- 
 dence was based upon sound theory and careful practice. He was skillful 
 in preparing estimates and was always well informed both as regards the 
 latest improvements in engineering, and the best methods of working the 
 materials of construction. 
 
 These accomplishments added greatly to his value as an instructor of 
 young engineers. His students were brought veiy close to engineering 
 work. Though well read in theory, he loved to dwell on the details of 
 practice. He never lost an opportunity to learn a new process, or to study 
 a new machine. He used to tell how, while resident engineer on a road in 
 
VIII. STEAM USING; OK, STEAM ENGINE PRACTICE. 
 
 New England, lie tried his hand on the engine of the construction train 
 till he was able to "stoke" and to "drive." 
 
 Professor Smith left a wife and three children. During her husband's 
 long and discouraging sickness, Mrs. Smith was better than a faithful 
 nurse: she brought aid to his self-imposed labor, and hope and cheer to 
 his fainting spirit. So well did she understand the nature of his work 
 and his needs, and so helpful was the assistance she brought, that it is not 
 too much to say that without her positive cooperation and encouragement 
 the two books which he leaves behind would never have been finished. 
 
 I will not speak of personal losses. I prefer to feel that we all had 
 much to be thankful for in Professor Smith, and the nearest had the most. 
 Though dying in his thirty- eighth year, Professor Smith's memory may 
 well be preserved. The world is certainly the better for his having 
 lived in it. 
 
 C. M. WOODWAKD, 
 
 Dean Polytechnic School, 
 Washington University, St. Louis, Mo. 
 
 ST. Louis, December, 7, 1884. 
 
OOINTTIEiLTTS. 
 
 CHAPTER I. 
 
 PAGE. 
 ON THE NATURE OF HEAT AND THE PROPERTIES OF STEAM: 
 
 Heat ThermodynamicsRatios of Volume to Pressure : Regnault's Ratios 
 The Carnot Engine Making steam Measurement of Heat Expended 
 Table : The Properties of Saturated Steam Examples in Calculation 
 of Heat Expended, Etc. Table: Factors of Evaporation Its Use- 
 Table: Expansion and Density of Pure Water Entrained Water and 
 its Measurement... 114 
 
 CHAPTER II. 
 
 ON VALVE GEAR: 
 
 Action of the Valve Position of Valve with Regard to its Eccentric, Lap 
 and Lead Valve Diagrams and their Application Adjustable Eccen- 
 tricsThe Link and Problems Connected Therewith Gooch's Link- 
 MotionAllan's Link-Motion The Walschaert Link-Motion-Mar- 
 shall's Valve Gear Brown's Valve Gear Kirk's Valve Gear Joy's 
 Valve Gear Porter-Allen Link Motion Herr Kaiser's Gear The 
 Meyer Valve Cut-off and Problems Connected Therewith Ordinary 
 Slide Valves Piston Valves Poppet Valves Double Valve Trick's 
 Valve, Etc 15 56 
 
 CHAPTER III. 
 
 THE QUANTITY OF STEAM WHICH MIGHT BE, AND WHICH is USED: 
 
 Horse-Power, Indicated Horse- Power, Effective Horse-Power, Net Horse- 
 PowerExamplesCurve of Expansion and its Properties Data fur- 
 nished by Experiment Tables of Engine Trials Analyses of Results 
 of Practical Tests Internal Radiation Analysis of Experiments Show- 
 ing Effect of Internal Radiation, etc. Tables 57- 
 
X. CONTENTS. 
 
 CHAP TEE IV. 
 
 PAGE. 
 
 ON THE INDICATOR, THE INDICATOR DIAGRAM AND THE DIFFER- 
 ENT CLASSES OF ENGINES: 
 
 The Construction and Use of the Indicator Other Devices The Earlier 
 Forms of Engines Single Acting Engines: The Westinghouse, The 
 Brotherhood, and the Colt Disc Engine The Indicator Diagram and the 
 Effects of Internal Radiation, etc., from Actual Experiments Com- 
 pounding and Indicator Diagrams in Connection Therewith Clearance 
 Forms of Compound Engines Progress in Marine Engine Perform- 
 ancesConsumption of Fuel The Value of Jacketting and Compound- 
 ingPoppet Valve Eiver Engines and Diagrams High Service Pumping 
 Engine, St. Louis Water Works Indicator Diagram from Lawrence, 
 Mass., Pumping Engine Diagrams from Engines of Ocean Steamers 
 "Arizona" and "Aberdeen" Engines of Mississippi River Steamer "Mon- 
 tana" Engines of U. S. Lighthouse Steamer "Manzanita" Triple Ex- 
 pansion Engines of S. S. "Aberdeen" Compound Engines of S. S. 
 "Grecian" Three Cylinder Compound Engines of S. S. "Parisian" 
 Automatic Expansion and Governors The Wheelock Engine The 
 Porter- Allen Engine The Rider Automatic Expansion Gear Various 
 Forms of Expansion Slides and Valve Gears The Cummer Engine 
 Governor The Armington & Simms Engine Engine of the Steam 
 Yacht "Leila" The Locomotive The Buckeye Automatic Cut-off En- 
 gine The Reynolds' Corliss Engine The Lambertville, N. J., Auto- 
 matic Cut-off Engine The Porter-Allen Engine 87187 
 
 CHAPTEK V. 
 
 THE EXPERIMENTS OF HIRN AND HALLAUER: 
 
 Report on a Memoir Upon Steam Engines Experiments with a Steam En- 
 gineExperimental Study Comparing the Influence of Expansion in 
 Simple and Compound Engines 188285 
 
 CHAPTEK VI. 
 STEAM HEATING: 
 
 The Theory of Steam Heating Various systems in use in the United States 
 Description of Apparatus and Experiments Used by the Author- 
 Project for Heating a Cotton Mill. . . . . 286298 
 
STEA.M 
 
 OR 
 
 STEAM ENGINE PRACTICE. 
 
 CHAPTER I. 
 
 ON THE NATURE OF HEAT AND THE PROPERTIES OF STEAM. 
 
 By the term heat we understand that property of bodies by which they 
 grow hot, and give the sensation with which we are all familiar. 
 
 Heat is produced in three ways: 
 
 By chemical action, A. 
 
 By mechanical action, B. 
 
 By electrical action, C. 
 
 A. When certain chemical elements or compounds are combined under 
 certain circumstances, the result is a union accompanied by an increase of 
 temperature and the development of heat; as for example, carbon or hydro- 
 gen combining with oxygen; sulphuric acid, or quick lime, with water. 
 
 B. By the mechanical work of friction or percussion: Examples of 
 this are continually before us. 
 
 C. By the passage of an electric current in a conductor, as in wires of 
 too great resistance; or the electric arc. 
 
 The pf6perty of heat is thought by some to consist of a kind of motion 
 or vibration of the molecules of which bodies are supposed to consist; for 
 solid and liquid bodies in vibration, and for gaseous bodies in the real mo- 
 tion of the molecules. With the arguments, pro. or con., concerning this 
 hypothesis we have little to do further than to state that, its truth appears 
 very probable, and in such event the production of heat by chemical com- 
 bination or the passage of an electric current is simply a kind of mechani- 
 cal action; in the one case, the vibration resulting from the shock of mole- 
 cules attracting each other; in the other, from the setting up of a wave 
 movement, or kind of wave, in the path of the electric disturbance, what- 
 ever that may be. 
 
 That heat was produced by mechanical means has been long known. 
 While the identity of heat and mechanical force was suspected by Count 
 Rumford nearly a hundred years ago, it was reserved for Joule to prove 
 (by long continued experiment), that the same quantity of work always 
 gave the same quantity of heat, and to Bankine and Clausius to show, 
 theoretically, that the same quantity of heat always gives the same amount 
 of work, which has since been proved beyond all doubt by experimental 
 investigations. 
 
 By the labors of the two great men, Rankine and Clausius, the 
 
STEAM USING; OX, STEAM ENGINE PRACTICE. 
 
 science of thermodynamics was created, the application of mathematics 
 it the laws of heat. Of this interesting and beautiful science we shall, 
 however, only state the two fundamental principles: 
 
 First Principle "Heat and mechanical energy are mutually convertible, 
 "and heat requires for its production and produces by its disappearance 
 "mechanical energy in the proportion of 772 foot-pounds for each British 
 "unit of heat." 
 
 The British unit of heat, just mentioned, is: "The quantity of heat 
 "which corresponds to an interval of one degree of Farenheit's scale in the 
 "temperature of one pound of pure liquid water at and near its temperature 
 "of greatest density (39.1F)." 
 
 The second principle, as given by Clausius, is as follows: 
 
 Second Principle. "Heat, of itself, never passes from a cold body to a 
 hotter one." 
 
 Kankine states the second principle in a way that has been severely 
 criticised by Maxwell, but which appears to mean that, a unit of heat in a 
 cold body can do as much work as in a hot body, with the implied reserva- 
 tion that there must be yet a colder body into which it may pass. 
 
 Heat is converted into mechanical work through the agency of some 
 body that is expanded by heat, such as air or water. The heat is transferred 
 into these mediums, usually enclosed within limits of changeable volume, 
 the expanding medium enlarging the volume against a resistance thereby 
 does mechanical work. 
 
 It has been taken for granted that the word temperature was under- 
 stood to have its ordinary meaning, and that neither the ordinary thermo- 
 metric scales of temperature, nor the ordinary instruments used for meas- 
 uring temperature required description; but when great accuracy was 
 required, the use of the air thermometer drew attention to a very con- 
 venient scale. Dry air and some of the other gases increase in volume or 
 pressure from the temperature of melting ice to that of boiling water under 
 the atmospheric pressure as follows: 
 
 From the volume or pressure 1 to: 
 
 Constant Volume. 
 
 Constant Pressure. 
 
 Air 
 
 
 L 3665 
 
 
 3670 
 
 Hydrogen. 
 
 
 I 3667 
 
 
 3661 
 
 Nitrogen 
 
 
 3668 
 
 
 
 Carbonic Acid 
 
 
 3688 
 
 
 3669 
 
 Carbonic Oxide . . 
 
 
 3667 
 
 
 3719 
 
 Nitrous Oxide . 
 
 
 3676 
 
 
 3719 
 
 Cyanogen 
 
 
 3829 
 
 
 3877 
 
 Sulphurous Acid 
 
 
 .3843 
 
 
 .3903 
 
 NOTE. The above ratios are from Reenault. 
 
 With the air thermometer the change in volume of a portion of dry air 
 was used to measure the change in temperature, and the natural result was 
 that the temperature at which the dry gas would have no volume, if the 
 law should hold so far, was taken as the zero or starting point of such a 
 scale. This zero is 461 F. or 273 C., and is called "absolute zero," and 
 
XA TURE OF HE A T AND PROPERTIES OF STEAM. 3 
 
 temperatures measured on this scale are called "absolute temperatures." 
 We shall give later another and better reason for this scale and its name, 
 for we know now that all the gases above given can be reduced to liquids 
 and solids and therefore are not perfect gases. 
 
 A perfect or "reversible" engine was devised by Sadi Carnot; and 
 although such an engine cannot be constructed, and if constructed, could 
 not be worked; still it is extremely useful in assisting our conceptions and 
 in giving us a limit beyond which we cannot hope to proceed with im- 
 provements. 
 
 The operation of the Carnot engine is as follows: From a hot body, at 
 temperature T lt a working body receives heat at the same temperature T lt 
 expanding and doing work from the heat in the hot body directly. After 
 a time the hot body is withdrawn, leaving the working body at the same 
 temperature T lt and it then expands by virtue of the heat which it contains 
 until its temperature has fallen to T 2 . In expanding, more work has been 
 effected, which, of course, goes to the credit of the engine as work done. 
 At the temperature T 2 , the working body is brought into contact with a 
 body called the "cold body," at the same temperature T 2 -, work is then done 
 on the working body from the outside in compressing it to such a point, 
 heat meanwhile passing from the working body to the cold body at the 
 same temperature. So that by continuing the process of compression after 
 the removal of the cold body, the working body will have just reached its 
 first state of volume, pressure and temperature; the work expended in 
 the two compression processes is, of course, to the debit of the engine, but 
 there is on the whole a balance of work done by the engine. 
 
 It can be shown in this case, whatever be the working substance 
 used: First. That this engine utilizes more heat than can be utilized by 
 any other kind of engine working between the same temperatures 2\ and 
 T 2 . Second. That the work done, or heat utilized, is to the heat expended 
 from the hot body, as the difference between the temperatures between 
 which the engine works, T, !F 2 is to the absolute temperature of the hot 
 body jP,. Hence the fraction 
 
 2*1 
 
 where T is an absolute temperature, is known as the efficiency of the 
 engine, and is the maximum efficiency which can be reached by theory. 
 
 The proof of the above statements is given in any work on thermody- 
 namics, so that we shall not enter upon it here, believing it out of place 
 in a work of a practical character. 
 
 From the properties of the Carnot engine, a scale of temperature, 
 based upon the work done by a body when T x T 2 = 1, is established; 
 and it has been shown that the scale thus established coincides in origin 
 and amount with that of the perfect gas thermometer, which places it upon 
 a more substantial basis. 
 
 When heat is put into any body it may either increase the agitation of 
 its molecules, thereby heating it or raising its temperature; or it may ex- 
 pand it against an external resistance doing external work; or it may 1 
 
4 STEAM USING; OB, STEAM ENGINE PRACTICE. 
 
 change its condition, overcoming molecular attractions, doing what is 
 called internal work; or it may do two or three of these three things at the 
 same time. 
 
 When a fire is lighted under a boiler containing cold water, the heat 
 generated by the chemical action of combustion passes from the fire and 
 the gaseous products of combustion to the iron of the boiler, through the 
 iron of the boiler to the surface in contact with the water and thence into 
 the water. The volume of the water slightly increases with the tem- 
 perature, raising the level partly by its own increase in volume and partly 
 by the increase in volume of the air contained in the water. The heat 
 increases the molecular agitation of the water, till, usually at the tempera- 
 ture of 212 F., the boiler begins to make steam. If, as in many of the 
 boiler trials, the man-head or safety valve is open; or, as in a common tea 
 kettle, there is no other pressure than that of the air upon the water, at 
 this temperature the water remains; and all the heat going into it is 
 expended in overcoming the molecular attraction of one atom of water 
 for another, and in forcing the molecules apart. In thus overcoming the 
 molecular attraction it is doing internal work, and at the same time in lift- 
 ing the atmosphere by the steam formed, it is performing external work. 
 
 When the quantities of heat which a pound of water requires to raise it 
 from the temperature of melting ice into steam at any given pressure are 
 measured, that which it takes to raise the temperature is not exactly the dif- 
 ference in the temperatures which would be required if the specific heatof 
 water were constant, but a unit of heat raises the temperature of a pound 
 of water a little less than one degree at the higher temperatures. When a 
 boiler is making steam at a given pressure other than that of the atmos- 
 phere, there is a temperature at which steam forms from the water and 
 above which the water cannot be raised. This is known as the tempera- 
 ture of evaporation for the pressure. It is to be noted that the pressure of 
 the atmosphere may be partly removed and low pressure steam formed at 
 less than atmospheric pressure. 
 
 The quantity of heat required to evaporate a unit of weight of water 
 at different pressures, and to raise the temperature up to that of evapora- 
 tion, was carefully determined by Eegnault in an extensive series of 
 experiments made at the expense of the French Government. The volume 
 of one pound weight of steam, and, of course, its reciprocal, the density or 
 weight of a cubic foot of steam, was determined by experiments made by 
 Fairbairn and Tate. 
 
 From the heat of evaporation, the volume of steam, the pressure under 
 which it was evaporated, and the volume of the water from which it was 
 formed are computed: 
 
 First. The external work in foot-pounds, or the product of the pres- 
 sure in pounds per square foot by the difference in cubic feet of the volume 
 of one pound of steam and one pound of water. 
 
 Second. The external work in heat units obtained by dividing the ex- 
 ternal work in foot-pounds by 772. 
 
.V. I y/ '/,'/; OF ///;. I T A XT) PROPERTIES OF STE. 1 M. 5 
 
 Third. The internal work of evaporation obtained by deducting from 
 the heat of evaporation the external work found above. 
 
 Fourth. The sum of the internal work of evaporation and the heat 
 expended in raising the temperature, sometimes called the total internal 
 heat. 
 
 Fifth. The sum of the heat expended in raising the temperature of the 
 water, and the heat of evaporation; or, the sum of the total internal heat 
 and the external work in heat units; or, the sum of the heat expended in 
 raising the temperature, in internal work of evaporation and in external 
 work, is called the total heat. These quantities may all bo stated in foot- 
 pounds, and some writers prefer to use them in this way. But, although the 
 measurement of mechanical work is usually made in foot-pounds, all meas- 
 urements of heat and steam which require measurements of temperature 
 are best made with a thermometer, and by heat units; we shall, therefore, 
 retain the heat units. There is also this advantage, that in computation 
 there will be smaller numbers and less figures involved. 
 
 The measurement of the heat expended in raising the temperature of 
 water, in the total internal heat and the total heat, are all based on a start- 
 ing point of one pound of water at the temperature of melting ice. As, how- 
 ever, such quantities are usually used by differences, many writers give 
 these data from F. Of course this does not require any real existence 
 to the imaginary pound of water, as water assumed in this way. It gives 
 a little less numerical work with feed water at low temperature, but is of 
 no help when the specific heat has varied so as to alter the heat expended 
 in raising the temperature of the water from the difference between the 
 temperature and 32. We adhere to the basis of melting ice. 
 
 Most of the theoretical writers use as a base for the tables the temper- 
 ature of evaporation, although others use the pressure, a much more 
 practical starting point for engineers. But these writers have not given 
 the internal and external heats, have used in some cases the F. start- 
 ing point referred to above, and have given extended decimals. In our own 
 table we have only given the nearest heat unit, and have given a table, not 
 for every pound of pressure, it is true, but one in which it is very easy to 
 interpolate the nearest unit. We believe this table to be convenient for 
 use and sufficiently extended and accurate. 
 
 The heat of evaporation is called latent heat of evaporation, but as the 
 term latent has now no meaning we shall not retain it. 
 
 As the Kegnault experiments on steam are always considered models 
 in every respect, and as being of unapproachable accuracy, we shall only 
 say that they were made in all circumstances and conditions in a thoroughly 
 practical way, and that the values reached have been computed from purely 
 theoretical grounds; so also with densities. The table is to be relied 
 upon, and we shall not explain the experiments or comment further upon 
 them, but will illustrate by a few examples the use of the table here given: 
 
STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 TABLE I. THE PROPERTIES OF SATURATED STEAM. 
 
 Pressure in Ibs. 
 per sq. inch 
 above the at- 
 mosphere. 
 
 Temperature 
 of steam in 
 degrees Fah- 
 renhejt. 
 
 Heat above 32 
 F. in water 
 at boiling 
 point. 
 
 External work 
 in heat units. 
 
 H 
 
 Internal work 
 of evaporat'n 
 in heat units. 
 
 Latent heat of 
 evaporat'n in 
 heat units. 
 
 1 ** * 
 .2 eo 
 
 .2 
 .g 5 
 
 Weight of 1 cu. 
 ft. of steam 
 in pounds. 
 
 Volume of 1 It), 
 in cubic feet. 
 
 14 
 
 90 
 
 
 
 1109 
 
 
 
 
 
 
 13 
 
 121 
 
 99 
 
 62 
 
 1118 
 
 967 
 
 1029 
 
 3 
 
 0.006 
 
 172.0 
 
 12 
 
 138 
 
 106 
 
 65 
 
 1124 
 
 943 
 
 1018 
 
 
 0.008 
 
 117.5 
 
 11 
 
 160 
 
 118 
 
 67 
 
 1127 
 
 942 
 
 1009 
 
 d 
 
 0.011 
 
 89.6 
 
 10 
 
 160 
 
 128 
 
 67 
 
 1130 
 
 935 
 
 1002 
 
 3 
 
 .014 
 
 72.6 
 
 9 
 
 168 
 
 136 
 
 67 
 
 1133 
 
 925 
 
 993 
 
 <u . 
 
 .016 
 
 61.2 
 
 8 
 
 175 
 
 143 
 
 68 
 
 1134 
 
 923 
 
 991 
 
 ,Qg 
 
 .019 
 
 52.9 
 
 7 
 
 181 
 
 150 
 
 68 
 
 1137 
 
 918 
 
 987 
 
 
 .021 
 
 46.7 
 
 6 
 
 187 
 
 156 
 
 69 
 
 1138 
 
 913 
 
 982 
 
 a 8 
 
 .024 
 
 41.8 
 
 
 
 
 
 
 
 
 ft 
 
 
 
 5 
 
 192 
 
 161 
 
 69 
 
 1140 
 
 909 
 
 979 
 
 
 .026 
 
 37.8 
 
 4 
 
 196 
 
 165 
 
 70 
 
 1141 
 
 906 
 
 976 
 
 V-rl 
 
 .029 
 
 34.6 
 
 3 
 
 201 
 
 170 
 
 70 
 
 1143 
 
 903 
 
 973 
 
 5 S 
 
 .031 
 
 31.8 
 
 2 
 
 205 174 
 
 71 
 
 1144 
 
 899 
 
 970 
 
 j _ i o 
 
 .034 
 
 29.5 
 
 1 
 
 209 
 
 178 
 
 71 
 
 1145 
 
 896 
 
 967 
 
 5 
 
 .036 
 
 27.6 
 
 
 
 212 
 
 181 
 
 72 
 
 1146 
 
 893 
 
 965 
 
 1074 
 
 .038 
 
 26.3 
 
 1 
 
 215 
 
 184 
 
 72 
 
 1147 
 
 890 
 
 962 
 
 1074 
 
 .041 
 
 24.3 
 
 2 
 
 219 
 
 188 
 
 72 
 
 1148 
 
 888 
 
 960 
 
 1076 
 
 .043 
 
 23.0 
 
 3 
 
 222 
 
 191 
 
 73 
 
 1149 
 
 887 
 
 958 
 
 1078 
 
 .046 
 
 21.8 
 
 4 
 
 225 
 
 194 
 
 73 
 
 1150 
 
 885 
 
 956 
 
 1079 
 
 .048 
 
 20.7 
 
 5 
 
 227 
 
 196 
 
 73 
 
 1151 
 
 882 
 
 953 
 
 1079 
 
 .050 
 
 19.7 
 
 6 
 
 230 
 
 199 
 
 74 
 
 1152 
 
 879 
 
 951 
 
 1079 
 
 .053 
 
 18.8 
 
 7 
 
 233 
 
 202 
 
 74 
 
 1152 
 
 877 
 
 950 
 
 1079 
 
 .055 
 
 1S.O 
 
 8 
 
 235 
 
 204 
 
 74 
 
 1153 
 
 876 
 
 948 
 
 1079 
 
 .058 
 
 17.2 
 
 9 
 
 237 
 
 206 
 
 74 
 
 1154 
 
 873 
 
 947 
 
 1080 
 
 .060 
 
 16.6 
 
 10 
 
 239 
 
 208 
 
 74 
 
 1154 
 
 872 
 
 945 
 
 1080 
 
 .062 
 
 16.0 
 
 11 
 
 242 
 
 211 
 
 75 
 
 1155 
 
 869 
 
 944 
 
 1080 
 
 .065 
 
 15.4 
 
 12 
 
 244 
 
 213 
 
 75 
 
 1156 
 
 867 
 
 942 
 
 1080 
 
 .067 
 
 14.9 
 
 13 
 
 246 
 
 215 
 
 75 
 
 1156 
 
 866 
 
 941 
 
 1081 
 
 .070 
 
 14.4 
 
 14 
 
 248 
 
 217 
 
 75 
 
 1157 
 
 864 
 
 939 
 
 1081 
 
 .072 
 
 13.9 
 
 15 
 
 250 
 
 220 
 
 75 
 
 1158 
 
 863 
 
 938 
 
 1083 
 
 .074 
 
 13.4 
 
 16 
 
 252 
 
 222 
 
 75 
 
 1158 
 
 862 
 
 937 
 
 1083 
 
 .076 
 
 13.0 
 
 17 
 
 254 
 
 224 
 
 76 
 
 1159 
 
 859 
 
 935 
 
 1084 
 
 .079 
 
 12.7 
 
 18 
 
 256 
 
 226 
 
 76 
 
 1159 
 
 858 
 
 934 
 
 1084 
 
 .081 
 
 12.3 
 
 19 
 
 257 
 
 227 
 
 76 
 
 1160 
 
 857 
 
 933 
 
 1084 
 
 .083 
 
 12.0 
 
 20 
 
 259 
 
 229 
 
 76 
 
 1160 
 
 856 
 
 932 
 
 1085 
 
 .086 
 
 11.6 
 
 22 
 
 262 
 
 232 
 
 76 
 
 1161 
 
 853 
 
 929 
 
 1085 
 
 .090 
 
 11.0 
 
 24 
 
 266 
 
 236 
 
 77 
 
 1162 
 
 850 
 
 927 
 
 1086 
 
 .095 
 
 10.6 
 
 26 
 
 269 
 
 239 
 
 77 
 
 1163 
 
 848 
 
 925 
 
 1087 
 
 .099 
 
 10.0 
 
 28 
 
 272 
 
 242 
 
 77 
 
 1164 
 
 846 
 
 923 
 
 1088 
 
 .104 
 
 9.6 
 
 30 
 
 274 
 
 244 
 
 77 
 
 1165 
 
 844 
 
 921 
 
 1088 
 
 .109 
 
 9.2 
 
 35 
 
 281 
 
 251 
 
 78 
 
 1167 
 
 838 
 
 916 
 
 1089 
 
 .120 
 
 8 3 
 
 40 
 
 287 
 
 257 
 
 78 
 
 1169 
 
 834 
 
 912 
 
 1091 
 
 .131 
 
 7.G 
 
 45 
 
 293 
 
 263 
 
 78 
 
 1171 
 
 830 
 
 908 
 
 1093 
 
 .142 
 
 7.0 
 
 60 
 
 298 
 
 268 
 
 79 
 
 1172 
 
 825 
 
 904 
 
 1093 
 
 .154 
 
 6.& 
 
 55 
 
 303 
 
 273 
 
 79 
 
 1174 
 
 822 
 
 901 
 
 1095 
 
 .165 
 
 6.1 
 
 60 
 
 307 
 
 278 
 
 79 
 
 1175 
 
 818 
 
 897 
 
 1096 
 
 .176 
 
 5.7 
 
 65 
 
 312 
 
 282 
 
 80 
 
 1176 
 
 814 
 
 894 
 
 1097 
 
 .187 
 
 5.3 
 
 70 
 
 316 
 
 287 
 
 80 
 
 1178 
 
 811 
 
 891 
 
 1098 
 
 .198 
 
 5.0 
 
 75 
 
 320 
 
 291 
 
 80 
 
 1179 
 
 808 
 
 888 
 
 1099 
 
 .209 
 
 4.8 
 
 80 
 
 324 
 
 294 
 
 80 
 
 1180 
 
 806 
 
 886 
 
 1100 
 
 .220 
 
 4.5 
 
 85 
 
 328 
 
 298 
 
 81 
 
 1181 
 
 802 
 
 883 
 
 1100 
 
 .231 
 
 4.3 
 
 90 
 
 331 
 
 301 
 
 81 
 
 1182 
 
 800 
 
 881 
 
 1101 
 
 .241 
 
 4.1 
 
 95 
 
 334 
 
 305 
 
 81 
 
 1183 
 
 798 
 
 878 
 
 1101 
 
 .252 
 
 4.0 
 
 100 
 
 338 
 
 308 
 
 81 
 
 1184 
 
 795 
 
 876 
 
 1102 
 
 .263 
 
 3.8 
 
.Y.I 77 /,'/; OF HEAT AND PROPERTIES OF STEAM. 
 
 TABLE I. THE PROPERTIES OF SATURATED STEAM. 
 
 Pressure in Ibs. 
 per sq. inch 
 above the at- 
 mosphere. 
 
 Temperature 
 of steam in 
 degrees Fah- 
 renheit. 
 
 Heat above 32 
 F. in water 
 at boiling 
 point. 
 
 External work 
 in heat units. 
 
 Ik 
 
 o> 
 
 E 88 
 
 Internal work 
 of evajjorafn 
 in heat units. 
 
 Latent heat of 
 evaporat'nin 
 heat units. 
 
 ||| 
 
 a 
 
 ^ITI 
 
 Weight of 1 cu. 
 ft. of steam 
 in pounds. 
 
 Volume of 1 It), 
 in cubic feet. 
 
 165- 
 
 341 
 
 311 
 
 82 
 
 1185 
 
 792 
 
 874 
 
 1103 
 
 .274 
 
 3.6 
 
 110 
 
 344 
 
 315 
 
 82 
 
 1186 
 
 789 
 
 871 
 
 1104 
 
 .284 
 
 3.5 
 
 115 
 
 34^ 
 
 318 
 
 82 
 
 1187 
 
 787 
 
 869 
 
 1105 
 
 .295 
 
 3 4 
 
 120 
 
 350 
 
 321 
 
 82 
 
 1188 
 
 785 
 
 867 
 
 1106 
 
 .306 
 
 3.3 
 
 125 
 
 353 
 
 324 
 
 82 
 
 1189 
 
 783 
 
 865 
 
 1107 
 
 .316 
 
 3.2 
 
 130 
 
 355 
 
 327 
 
 82 
 
 1190 
 
 781 
 
 863 
 
 1108 
 
 .327 
 
 3.1 
 
 135 
 
 358 
 
 329 
 
 82 
 
 1191 
 
 779 
 
 861 
 
 1108 
 
 .338 
 
 3.0 
 
 140 
 
 361 
 
 331 
 
 83 
 
 1191 
 
 777 
 
 860 
 
 1109 
 
 .348 
 
 2.9 
 
 145 
 
 363 
 
 334 
 
 83 
 
 1192 
 
 775 
 
 858 
 
 1109 
 
 .359 
 
 2.8 
 
 150 
 
 366 
 
 337 
 
 83 
 
 1193 
 
 773 
 
 856 
 
 1110 
 
 .369 
 
 2.7 
 
 155 
 
 368 
 
 340 
 
 83 
 
 1194 
 
 771 
 
 854 
 
 1111 
 
 .380 
 
 2.6 
 
 160 
 
 371 
 
 341 
 
 83 
 
 1194 
 
 770 
 
 853 
 
 1111 
 
 .390 
 
 2.6 
 
 165 
 
 373 
 
 344 
 
 83 
 
 1195 
 
 768 
 
 851 
 
 1112 
 
 .400 
 
 2.5 
 
 170 
 
 375 
 
 347 
 
 84 
 
 1196 
 
 765 
 
 849 
 
 1112 
 
 .412 
 
 2.4 
 
 175 
 
 377 
 
 348 
 
 84 
 
 1196 
 
 764 
 
 848 
 
 1113 
 
 .422 
 
 2.4 
 
 180 
 
 380 
 
 351 
 
 84 
 
 1197 
 
 762 
 
 846 
 
 1113 
 
 .433 
 
 2.3 
 
 185 
 
 382 
 
 353 
 
 84 
 
 1198 
 
 761 
 
 845 
 
 1114 
 
 .443 
 
 2.3 
 
 195 
 
 386 
 
 357 
 
 84 
 
 1199 
 
 758 
 
 842 
 
 1115 
 
 .463 
 
 2.2 
 
 205 
 
 390 
 
 361 
 
 85 
 
 1200 
 
 754 
 
 839 
 
 1115 
 
 .484 
 
 2.1 
 
 215 
 
 394 
 
 365 
 
 85 
 
 1201 
 
 751 
 
 836 
 
 1116 
 
 .505 
 
 2.0 
 
 225 
 
 397 
 
 368 
 
 85 
 
 1202 
 
 749 
 
 834 
 
 1117 
 
 .525 
 
 1.9 
 
 235 
 
 401 
 
 373 
 
 85 
 
 1204 
 
 746 
 
 831 
 
 1119 
 
 .546 
 
 1.8 
 
 245 
 
 404 
 
 376 
 
 85 
 
 1205 
 
 744 
 
 829 
 
 1120 
 
 .567 
 
 1.8 
 
 255 
 
 408 
 
 380 
 
 85 
 
 1206 
 
 741 
 
 826 
 
 1121 
 
 .587 
 
 1.7 
 
 265 
 
 411 
 
 383 
 
 85 
 
 1207 
 
 739 
 
 824 
 
 1122 
 
 .608 
 
 1.6 
 
 275 
 
 414 
 
 386 
 
 85 
 
 1208 
 
 737 
 
 822 
 
 1123 
 
 .627 
 
 1.6 
 
 285 
 
 417 
 
 389 
 
 86 
 
 1209 
 
 734 
 
 820 
 
 1123 
 
 .649 
 
 1.5 
 
 335 
 
 430 
 
 392 
 
 86 
 
 1213 
 
 725 
 
 811 
 
 1127 
 
 .750 
 
 1.3 
 
 385 
 
 445 
 
 417 
 
 86 
 
 1217 
 
 714 
 
 800 
 
 1131 
 
 .850 
 
 1.2 
 
 435 
 
 457 
 
 428 
 
 87 
 
 1220 
 
 705 
 
 792 
 
 1133 
 
 .950 
 
 1.05 
 
 485 
 
 467 
 
 440 
 
 87 
 
 1224 
 
 697 
 
 784 
 
 1137 
 
 1.049 
 
 0.95 
 
 585 
 
 487 
 
 460 
 
 87 
 
 1230 
 
 683 
 
 770 
 
 1143 
 
 1.245 
 
 0.80 
 
 685 
 
 504 
 
 477 
 
 88 
 
 1235 
 
 670 
 
 758 
 
 1147 
 
 1.439 
 
 0.69 
 
 785 
 
 519 
 
 493 
 
 88 
 
 1240 
 
 659 
 
 747 
 
 1152 
 
 1.632 
 
 0.61 
 
 885 
 
 534 
 
 507 
 
 88 
 
 1244 
 
 649 
 
 737 
 
 1156 
 
 1.823 
 
 0.55 
 
 985 
 
 516 
 
 520 
 
 88 
 
 1248 
 
 640 
 
 728 
 
 1160 
 
 2.014 
 
 0.50 
 
 Values below * * * are computed and not experimental. 
 
 NOTE. For all values of Total Internal work below the atmosphere 1070 heat units 
 may be taken. All decimal parts of heat units have been neglected and the last one 
 may therefore be in error. 
 
 Example /.How much more heat is needed to boil a pound of water 
 at 200 pounds per square inch boiler pressure than at five pounds per 
 square inch, the feed being at 60 F. in either case. 
 
 AT FIVE POUNDS. 
 
 Units. 
 
 Heat required to raise 1 pound water from 32 to boiling at 5 pounds pressure 196 
 
 Deduct heat to raise from 32 to 60 not used .' 28 
 
 ' 
 
 Heat to raise from 60 to boiling 168 
 
 Internal work of evaporation.* 882 
 
 External work of evaporation 73 
 
 Heat required to boil from feed at 60 at 5 pounds 1,123 
 
8 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 AT TWO HUNDRED POUNDS. 
 
 Units. 
 
 Heat required to raise 1 pound water from 32 to boiling at 200 pounds per sq. inch. . 359 
 Deduct heat to raise from 32 to 60 not used 28 
 
 331 
 
 Internal work 756 
 
 External work . 84 
 
 1,171 
 
 Heat required to boil 1 pound of water from feed at 60 at 200 pounds: 
 1,171 1,123 = 48 units. 
 
 j-|L = 4 per cent., nearly. 
 
 The same result could be reached more directly. 
 
 Units. 
 
 Total heat from 32 at 200 pounds 1,199 
 
 Total heat from 32 at 5 pounds 1,161 
 
 Difference 48 
 
 Deducting from the 1,151 the 28 units not used, from 32 to 60, the feed 
 being at 60, we have 1,123 for the divisor to reduce to per cent, as before. 
 
 We advise the reader to use the former method, by preference, in his 
 computations, as serving to keep in full view the different uses and the va- 
 rious amounts of heat required for them; although there is, of course, more 
 numerical work required to do so. 
 
 The reason so much more difficulty is experienced in maintaining 
 high pressure than low pressure steam is to be found, not in the boiling of 
 equal weights of water, but in the fact that the high pressure steam 
 leaves the boiler more easily. If, for example, it be employed in an 
 engine, the engine can be made to do more work thereby. If, in running 
 a boat, the boat going faster the engine uses more steam; if employed 
 in heating a building, the radiators act more energetically with the higher 
 pressure, transmit more heat, condense more steam, and the skillful 
 attendant suits his fire to the work. 
 
 Example //.How much saving of fuel can be made by raising the 
 temperature of the feed- water from 100 F. to 200 F., the boiler pressure 
 
 being 120 pounds per square inch. 
 
 Units. 
 
 Total heat for 120 pounds 1,188 
 
 Deduct in the one case the units not used in raising the water from 32 F. to 100 F . . 68 
 
 Required from 100 F. to boil at 120 pounds 1,120 
 
 In the other case deduct for not using from 32 to 200 169 
 
 Required to boil at 120 pounds from water at 200 F 1,019 
 
 Difference between 1,120 and 1,019 is 101 units, or about 9 per cent. 
 In order to compare the performance of different boilers working with 
 different pressures and fed with water at different temperatures, it is ne- 
 cessary to assume a standard pressure, temperature of evaporation, and 
 temperature of feed-water. Various temperatures of feed-water have been 
 used, F., 32 F., 100 F., the latter about the usual temperature of feed- 
 water for condensing engines, and 212 F., used more generally than any 
 of the others as a standard; while for the pressure and temperature of 
 evaporation the atmospheric pressure and 212 F. are usually taken. 
 
NATURE OF HEAT AND PROPERTIES OF STEAM. 9 
 
 Example III. By experiment with a boiler at 160 pounds per square 
 inch it was found that, one pound of coal evaporated 7.91 pounds of water. 
 The temperature of the feed- water was noted at 120 F.: required the 
 equivalent evaporation from and at 212 F. 
 
 Units. 
 
 Total heat of evaporation from 32 F. at 160 pounds 1,194 
 
 Deduct from 32 to 120, units not used 88 
 
 Heat to evaporate from 30 at 160 pounds 1,106 
 
 Internal heat of evaporation at 212 893 
 
 External work of evaporation at 212 72 
 
 Sum or heat of evaporation at 212 965 
 
 7.91 x 1>106 = 9.06 as the evaporation required. 
 965 
 
 In order to facilitate this computation the following table of factors of 
 evaporation is given: 
 
10 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 -l 00 CO h- C t- <N t- 8*1 t- H O ^ O ^ IO O Ift O IO OS * CJ H< OS 00 W ' 
 
 * -; <N e*i TH rH o o ci en oo oo i- c- ee o m >n -* eo co o <N rH rn 001 
 
 I t- IN t- <M t- 
 
 SS8 g 
 
 s 1-1 vc o w o o o "io'ci ^ cr, * e> co oo co x co t- <N t- ^ t- rH o rH o rH in o m o >n o 10 
 
 ! CO <M O4 -H r-i o O Ci OC OOt-t-OtC W in * -* CO CO <M C^ rH rH OCCiCiOO 00 t- t- 5 <O O 
 IC^OIC^C'I C^C^C^i tfH rH r- I rH I-H r- I i I rH rH rH r- ( rH rH rH rH i I rH rH O O O OOOOO O 
 
 rH ?Ci IOi-Hin OiCOknCs -*O-^OCO OOCO' 
 
 <> inin^-*co COCNCN^HO ocsooooo t-t-. 
 
 * rH rH i I rM rH i t i (i (r-ir-Hr-t rHOOOO OO 
 
 "* "i-i"o~o~ft o inomo^n 
 
 CO COC)<Mi-l^H OOClOiCO 
 
 I i i i-n rH rHr-trHrHi-H rH O O C> O O O < 
 
 ccc-<Nt- cMt-(Mt-i-( o^HOr-im 
 
 ^C^r-lilHO OCiC5X t- t- <0 10 10-*^CO<N <^<-<i-ta>0 05C5QCOOt- t- O O IO 1O ^< 
 C^C^CNCMC^ C* T j TH i I rH rHi li-HrHrM rHrHi IrHrH rH i ( rH rH i* OOOOO OOOOO O 
 
 e 
 
 -^Oicoooco oocooocot* c^t^ot H coi^Or-iin oinoino^ ^Cyi-^Oico oocooccooo co 
 
 SNr-Hr^OO OSC^OOOOC- t- JO 5O JO VO * ^J CO CO CJ 2^rtSS OOOOO OOOOO O 
 
 ""~ or-#c>~*& oo" co " """ "~ 
 
 CCCCCO 
 C- 
 
 OOd l^C'lt-C^SO r-( 
 
 OOO 
 
 O5 c I-H m o in o~ m oino -*O5-*Oico oocooocot- (Mt-<Mt-i-i i-iOrHi 
 
 ?H rHT-HOOOS O 5 **^^ SSSrnS J32SSIH JnSSoO OOOo! 
 
 t--i o^-ioi-iin omowos -*O5-*cico oocoooi 
 
 -*-* COCO<N<MrH l-nOOCSOO OOt^t~OO 1O U5 >* 
 
 0000 t^O SS54I? C?S?5222 OOOiSoO SSt^SS Si 
 
 cr -~ cr -^ X ^ "O OCC^ t C^ t O3 tO rnOrHCOO inO 
 
 c^cnoo oot-t~oo inmHt-^tco coc^<MrHrH OO 
 
 IrHrHrH rHrHrHrHrH rHrHrHrHrH rHrHrHrHrH rH^H 
 
 COOOCOOOCO OO 
 
 in-*-*coco c* 
 
 OOOOO O 
 
 rH oooiCROo oot^t^oo inln-*5)co COC^CJTHO ocsoj 
 
 t- 1- o o-*-*c 
 
 OOOOO OOO 
 
 oin 
 
 t-o 
 
 oocooocot c^t cMtrn o r-i c.^ r- in oinoi 
 ^^rHrnS S 05050000 t ~ t ~ oy5 " 5 "^g^!; 
 
 i on (-<l t (M t-'NtC^O THOrHCOO lOOiOO-^ Oi^*iCi^OO COOOCOOOC^ t- CN t OJ t 
 
 )O5C50ooo t-t-oom in-*-^co( <M<MrHrno ascftaoooc- t-o3inm -*-*coco<;<) 
 
 8 H 
 
 c- <N o rH o rn o o in o o o m 
 
 SSS ooooo ooooo 
 
 SS 
 
 CCC- C^t 
 
 
 i iilll lllii I 
 
 S ISSSS ISSS3 Slip Slsis liiil iilll ISIsI S 
 
 lliiiisssaa sssas sins 
 
 S Hill SrlSSrH SSrHrnS SSlll Illll 1111 11S8 II 
 
 
3,'ATUJiE OF SEAT AND PROPERTIES OF STEAM. 
 
 11 
 
 The use of this Table of Factors of Evaporation is readily seen by tak- 
 ing the last example. The boiler evaporating 7.91 pounds at 160 pounds 
 per square inch from feed at 120 F., th3 evaporation factor from Table 
 II. for 120 and 160 pounds is 1.146. 7.91 x 1.146 = 9.06, as before, for the 
 equivalent evaporation from and at 212 F. 
 
 We introduce one other table here the weight of 1 cubic foot of water 
 at different temperatures. Very often in the trials of a boiler or engine 
 the most convenient unit of measurement of water is the cubic foot. This 
 will be the case when a weir measurement is made or when the water is 
 measured by a water meter. The use of a water meter involves many pre- 
 cautions, the most important being the following: The meter should work 
 under moderate head of supply and small head of delivery; it should be 
 set in such a manner that it can be tested in place under the exact condi- 
 tions of use; if a positive meter, it should be especially constructed to 
 work freely, if it is to be used in warm water. This table is also used for 
 estimating the weight of water in boilers, and for correcting boiler trials 
 for differences of water level. 
 
 TABLE HI. EXPANSION AND DENSITY OF PURE WATER. 
 FROM D. K. CLARK AND BY RANKINE. APPROXIMATE FORMULA. 
 
 Temperature in degrees 
 
 COMPA 
 
 BATTVE. 
 
 Density of Weight 
 
 Fahrenheit. 
 
 Volume. 
 
 Density. 
 
 per Cubic Foot. 
 
 32 
 
 1.00000 
 
 1.00000 
 
 62.418 
 
 35 
 
 1.99993 
 
 .00007 
 
 62.422 
 
 39.1 
 
 1.99989 
 
 .00011 
 
 62.425 
 
 40 
 
 1.99983 
 
 .00011 
 
 62.425 
 
 45 
 
 1.99993 
 
 .00007 
 
 62.422 
 
 46 
 
 .00000 
 
 1.00000 
 
 62.418 
 
 50 
 
 .00015 
 
 .99985 
 
 62.409 
 
 52.3 
 
 .00029 
 
 .99971 
 
 62.400 
 
 55 
 
 .00038 
 
 .99961 
 
 62.394 
 
 60 
 
 .00074 
 
 .99926 
 
 62.372 
 
 62 
 
 .00101 
 
 .99899 
 
 62.355 
 
 65 
 
 .00119 
 
 .99881 
 
 62.344 
 
 70 
 
 .00160 
 
 .99832 
 
 62.313 
 
 75 
 
 .00239 
 
 .99771 
 
 62.275 
 
 80 
 
 .00299 
 
 .99702 
 
 62.232 
 
 85 
 
 .00379 
 
 .99622 
 
 62.182 
 
 90 
 
 .00459 
 
 .99543 
 
 62.133 
 
 95 
 
 .00554 
 
 .99449 
 
 62.074 
 
 100 
 
 .00639 
 
 .99365 
 
 62.022 
 
 ioa 
 
 .00739 
 
 .99260 
 
 61.960 
 
 :uo 
 
 .00889 
 
 .99199 
 
 61.868 
 
 115 
 
 .00989 
 
 .99021 
 
 61.807 
 
 120 
 
 .01139 
 
 .98874 
 
 61.715 
 
 125 
 
 .01239 
 
 .98808 
 
 61.654 
 
 130 
 
 .01390 
 
 .98630 
 
 61.563 
 
 135 
 
 .01539 
 
 .98484 
 
 61.472 
 
 140 
 
 .01690 
 
 .98339 
 
 61.381 
 
 145 
 
 .01839 
 
 .98194 
 
 61.291 
 
 150 
 
 .01989 
 
 .98050 
 
 61.201 
 
 155 
 
 .02164 
 
 .97802 
 
 61.0% 
 
 160 
 
 .02340 
 
 .97714 
 
 60.991 
 
 165 
 
 .02589 
 
 .97477 
 
 6H.843 
 
 170 
 
 .02690 
 
 .97380 
 
 60.783 
 
12 
 
 STEAM USING; Oil, STEAM ENGINE PRACTICE. 
 
 TABLE III. CONTINUED. 
 
 Temperature in degrees 
 Fahrenheit. 
 
 COMPARATIVE. 
 
 Density of Weight 
 per Cubic Foot. 
 
 Volume. 
 
 Density. 
 
 175 
 180 
 
 1.02906 
 1.03100 
 
 .97193 
 .97006 
 
 60.655 
 60 .548 
 
 185 
 190 
 195 
 200 
 205 
 
 1.03300 
 1.03500 
 1.03700 
 1.03889 
 1.0414 
 
 .96828 
 .96632 
 .96440 
 .96256 
 9602 
 
 60.430 
 60.314 
 60.198 
 60.081 
 59.93 
 
 210 
 212 by formula. 
 212 by measurement. 
 
 1.0434 
 1.0444 
 1.0466 
 
 .9584 
 .9575 
 .9555 
 
 59.82 
 59.76 
 59.64 
 
 230 
 250 
 270 
 290 
 
 1.0529 
 1.0628 
 1.0727 
 1.0838 
 
 .9499 
 .9411 
 .9323- 
 .9227 
 
 59.36 
 58.78 
 58.15 
 57.59 
 
 298 
 338 
 366 
 390 
 
 1.0899 
 1.1118 
 1 . 1301 
 1 .1444 
 
 .9175 
 .8994 
 .8850 
 .8738 
 
 57.27 
 56.14 
 55.29 
 54.54 
 
 The use of the table of the properties of steam is more frequent in the 
 study of engine performance and indicator diagrams than of boiler per- 
 formance, but there is an important point in determining the evaporation 
 of a boiler in which it becomes of use. 
 
 As bubbles of steam formed on the hot iron of a boiler rise through 
 the water to the surface, breaking and scattering spray, a portion of water 
 thus thrown up into the steam room is carried along with the steam, and 
 unless more heat be supplied to evaporate this water it increases the volume 
 caused by the steam condensed in the pipes in the upper portion of the 
 boiler. This water carried with the steam is said to be "entrained" with it 
 and is called "priming" by many writers. When the proportion of water 
 becomes so large as to be evident in the action of the engine or the exhaust, 
 it is usually called by engineers "foaming." The amount of such water is 
 increased if the water is dirty and covered with scum, or if grease and alkali 
 combine to form a soap. The amount of water which can be carried by 
 steam in suspension is very great, but depends somewhat upon the velocity 
 of the current of steam; if the passages are large, and the flow of steam of 
 moderate velocity the water has time to drop out of the steam by the 
 action of gravity. ( Iu some cases the amount of water carried in weight 
 has been known to be three times that of the steam carrying it, although 
 usually it does not exceed 10 to 15 per cent. 
 
 The higher the pressure of steam the greater its density and the 
 quieter, other things being equal, is the process of ebullition and the smaller 
 the quantity of entrained water. The amount of water thrown up in spray 
 is largely dependent on the circulation, being much diminished by im- 
 provements in that direction. The area of surface water in contact with the 
 steam seems to be an important matter according to some authorities, but 
 
NATURE OF HEAT AND PROPERTIES OF STEAM. 13 
 
 as this varies very greatly without any apparent effect, we are not inclined 
 to attribute much importance to it. A violent rush of steam close to the 
 top of a body of water is to be avoided, as even a current of air would 
 throw spray in such a case. 
 
 The accurate determination of the water entrained with steam is a 
 matter of great difficulty and at the same time of great importance in the 
 determination of the performance of boilers and engines. 
 
 Four methods have been devised to measure the amount of water en- 
 trained, and two of them have been used in practice. 
 
 The first method, that of M. G. A. Him, is the most used. It depends 
 upon the amount of heat given out by a known weight of a mixture of 
 steam and water and is best performed as follows: 
 
 A barrel is set on a platform scale and a known weight of water run 
 into it. It is convenient to put in 298 Ibs. of water. Steam is taken from 
 the top of the steam pipe by a rubber hose terminated by an iron pipe 
 capped on the lower end and perforated with holes drilled obliquely to 
 the radii, but in the plane thereof. . This pipe is placed in the barrel of 
 water and steam turned on; the scale is loaded 2 Ibs. more, and as the 
 steam comes into the water the fluid increases in weight, and when the 
 beam tips there is 300 Ibs. of water. The temperature of the water is then 
 carefully noted. The disposition of the jets keeps the water stirred up 
 thoroughly, and the flow of steam into the water being horizontal only, the 
 water remains steady. The weight is then increased 10 Ibs., and when the 
 the scale tips at 310 pounds the temperature is noted. 
 
 The number of heat units given to the water, in the barrel, by the 
 steam and water from the boiler, is found by multiplying the 300 Ibs. of 
 water by the rise in its temperature. 
 
 The portion which was dry steam gives up its internal heat of evapor- 
 ation in condensing, and the external work done by the air upon the fluid 
 in compressing it from steam to water, together make the latent heat of 
 evaporation; and the whole fluid then falls in temperature from that due 
 to the pressure in the boiler to the final temperature of the barrel. 
 
 Deducting from the heat gained by the water in the barrel, ten times 
 the difference between the boiler temperature and the final temperature 
 in the barrel, and dividing the remainder by ten times the latent heat at the 
 boiler pressure, the quotient will be the fraction of the whole which is dry 
 steam. 
 
 It is easily seen that with any other weight the process would be the 
 same; but in place of the ten we should use the number of pounds run in 
 between the noting of the temperatures. 
 
 The preliminary 2 Ibs. is to provide for any water which might have 
 collected in the hose or connections while standing, and to render the op- 
 eration uniform. 
 
 Sometimes a coil of pipe as a surface condenser is used, and the steam 
 which is condensed therein is kept separate from the condensing water; 
 but great care has to be used to get all the water condensed out of the coil. 
 The accuracy of this method is dependent upon the delicacy of weighing 
 
14 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 and the reading of the thermometer; in unskillful hands the results are 
 sometimes astonishing. 
 
 The second method is to put into the feed water a quantity of sulphate 
 of soda, and to draw from the boiler, at intervals, from the lower gauge cock 
 a small amount of water, keeping this water by itself; also to draw from the 
 steam, condensing either by a coil of pipe in water, or a small pipe in air, 
 taking care to draw only water without steam, at the same intervals, keep- 
 ing the one separate from the other. A chemical analysis defines the pro- 
 portion of sulphate of soda in each portion, and a division of the proportion 
 of sulphate of soda in the portion from the steam by the proportion in that 
 from the water gives the proportion of water entrained, the basis of the 
 method being the fact that steam does not carry the sulphate of soda, 
 this being only carried by the hot water entrained. This method was used 
 by Professor Stahlschmidt at the DusseldOrf Exhibition Boiler Trials. 
 
 A third method has been suggested: To enclose a portion of steam in a 
 vessel placed inside the steam pipe, then closing it and removing it from 
 the steam pipe, obtain the weight of the enclosed fluid, which, being 
 in a known volume, the proportion of water can be found from the volume 
 and density at the known pressure. There appear to be many practical 
 difficulties in this method, and we are not aware that it has been used to 
 any extent. 
 
 A fourth method is to have a small cylinder with piston enclosed in the 
 steam, and to put a known volume of the cylinder in connection with the 
 steam; then closing the communication, pull out the piston (which, 
 of course, passes through proper stuffing boxes into the air) until the 
 pressure in the cylinder begins to lower, the water contained evapor- 
 ating at the pressure, until, after it has been evaporated the pressure begins 
 to fall with increase of volume. The increase of volume at constant pres- 
 sure divided by the final volume is the proportion of water carried. This 
 method promises well, but we have no knowledge of its use. 
 
 Steam formed in the presence of water is always saturated, that is, it is 
 at the same temperature as the water, and cannot be raised above that tem- 
 perature until the water is all evaporated; but after this has been done, or 
 if the steam be heated in a separate vessel, the temperature rises nearly 
 2 F. for each unit of heat added to a pound in weight, while the steam 
 increases in volume at first not very closely, but afterwards very nearly as 
 a perfect gas, or by ^- s part of itself for each degree F. The amount of 
 heat required to raise 1 ft. weight of dry steam 1 F., is stated as 0.47 of a 
 unit, and 0.5 by different authorities, the first including Kankine, and the 
 second Hirn. Steam thus raised in temperature is said to be superheated, 
 but our knowledge of this condition is still very limited and confined to 
 the results of a few experiments. 
 
CHAPTER II. 
 
 ON VALVE GEAR. 
 
 As steam can, only do useful work by changing the volume of the 
 vessel enclosing it, we find that to employ it in moving a piston in a cyl- 
 inder is the most practical method of using it. There is, however, a cer- 
 tain class of pumping engines, in which steam in direct contact with a 
 fluid, without intervening mechanism, is used for pumping water. 
 
 The motion of the piston to and fro in the cylinder is communicated 
 to the outside of the cylinder by a rod passing through a steam-tight 
 opening, where it is then, in some way, connected to the resistance which 
 is to be overcome, either directly as to a pump rod, or indirectly by a 
 connecting rod and crank to a rotating shaft, which is to be revolved 
 against the resistance. 
 
 The steam has to be let in and out of one or both ends of the closed 
 cylinder, and for this purpose there are used one, two, three, or four 
 valves, which open and close two or four passage ways, or ports. 
 
 These valves are moved by mechanism of more or less complex 
 form by the engine itself, and are either slides, lifting valves, or rotative 
 valves. 
 
 The use of the slide valve and eccentric requires little illustration 
 until we examine them more closely. 
 
 The piston of the engine is connected to the piston rod, which passes 
 out of the cylinder, and is fastened to the cross- head, and this can only 
 move in a straight line. The connecting rod is attached, at one end, to 
 the cross-head by a pin and suitable box, and at the other end to the 
 crank pin; thus, one end of the connecting rod can only move in a 
 straight line while the other end moves in a circle. 
 
 The action of the slide valve by the eccentric is in a similar manner 
 one end of the eccentric rod moving in, or nearly in, a straight line, while 
 the other end moves in a circle; for, it must not be forgotten that the 
 eccentric may be, and sometimes is, defined as a crank with a small crank 
 arm and large crank pin. 
 
 It is well known that the motion of the two ends of the connecting 
 rod is not regular; that is to say, that for equal movements of the piston 
 equal ones are not moved over by the crank, and further, that for equal 
 distances moved by the piston from the end of the stroke, the angles 
 moved over by the crank from the dead points are not alike; and, also, 
 for equal arcs moved by the crank, the cross-head aijd piston move 
 unequal distances. This is shown in Fig. 2, where the arcs 0-1 is equal to 
 6-7, 1-2 equal to 5-6, and 2-3 equal to 4-5, but the distances on the straight 
 line between points with the same numbers are evidently not equal. 
 
16 
 
 STEAM rSING; OK, STEAM ENGINE PRACTICE. 
 
fL\ VALVE fiEAli. 
 
 17 
 
 An examination of this irregularity shows us that it increases as the 
 lengths of the connecting rod and crank are more nearly equal, and 
 decreases the more unequal they are. Now for any engine, the lengths of 
 the connecting rod and crank are more nearly equal than the eccentric 
 
 o i 
 
 FIG. 2. 
 
 rod and eccentric radius for the same engine, and if the latter are 50 
 inches and 2 inches long, respectively, the greatest error introduced by 
 them is less than 0.005 inch. We should be justified in neglecting this 
 entirely, but we must remember that we can not do so for the irregularity 
 due the connecting rod and crank; and it becomes convenient for us to 
 refer the position of the slide to that of the crank arm instead of referring 
 it to the piston, as is more usually done. While it is more natural to think 
 of the position of the valve with the piston at half stroke, for example, 
 yet when we reflect that this piston at half stroke gives two positions for 
 the valve instead of one, it becomes evident that to define the position of 
 the valve as that due a certain crank angle, 90 in the case taken, or crank 
 position given by drawing or otherwise, is more exact. For we so easily 
 find, by drawing, the place of the piston or cross-head when the crank 
 position is given, or the reverse, that no difficulty will be met with in our 
 study if we confine ourselves to the connection between crank and valve 
 position, in place of the more usual statement, the connection between 
 piston and valve positions. The methods used are, therefore, those of 
 drawing, and little computation is needed. 
 
 The position of the valve with regard to its eccentric is readily found 
 as follows: In Fig. 3, A B is the throw of the eccentric, F C the crank, 
 and C D the eccentric radius; D being the centre of the eccentric while C 
 
 FIG. 3. 
 
18 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 is the centre of the shaft. To find the position of the valve, draw D E 
 from E perpendicular to A B. Then will C E be the distance the valve 
 has moved from its middle position, and E B and E A the distances of the 
 valve from the end of its travel in either direction. By combining this 
 with C F, the position of the crank arm, we see that we have the angle 
 between the, crank and eccentric. A convenient method of examining the 
 whole motion at once, is found to be by laying out on C D or C F the dis- 
 tance CE=CCr=CH, and moving the point G or H as the shaft is 
 turned. 
 
 We will take a position of Fig. 3 on a larger scale in Fig. 4, and make 
 the construction for different positions by laying upon the eccentric arm 
 the amount the valve has moved from the centre for that position of the 
 eccentric. In doing this we see at once that if there be drawn from the 
 point A, or end of the travel, a perpendicular upon the eccentric arm, A F, 
 the distance C E C F, which is the required distance from the middle 
 position. Make C F f on the crank arm = C .Fon'the eccentric arm. By 
 
 FIG. 4. 
 
 drawing these for different positions it will be found that the point F 
 always falls upon one of two circles whose centres are at G and H, one 
 half of C A and C B from C. This may be verified by trial, or by consid- 
 ering that A F C is always a right angle and therefore inscribed in a semi- 
 circle, since C A is common to all. 
 
 The point F f will always lie on one of two circles having the same 
 
ON VALVE GEAR. 
 
 19 
 
 FIG. 5. 
 
 radii as before, but with centres I and J, which lie upon a line I J making 
 an angle with A B at the point C, where it intersects it, said angle I C A = 
 D C I", the angle between crank and eccentric; for when the crank is at 
 1 C, the eccentric is at C A, and the valve is therefore farthest from the 
 middle. We shall call the circles just found the distance circles, for by 
 drawing any position C F' of the crank, we find that the valve is distant 
 C f from its middle position. 
 
 We are now in a condition to state and solve our first problem: 
 Given, the angle between the crank and eccentric arms and the travel of 
 the valve; to find the position of the valve for any position of the crank. 
 
 In Fig. 5 lay off A B, the full 
 travel of the valve, and bisect A B in 
 C. Lay off D C A, the angle between 
 the crank and eccentric arms, and 
 make D C = i A C, also produce D C 
 and make C E = D C. Draw two 
 circles from D and E as centres with 
 radii = C D These are the distance 
 circles. For any position of the 
 crank, as C F or C F' , or C F", the 
 amount the valve has moved from its 
 , middle position is 
 given by the dis- 
 tance from C to G, 
 G', or G". If the 
 crank arm cut the 
 circle D the valve 
 is to the right of 
 the centre, if it 
 cut the circle E, 
 the valve is to the 
 left of the centre. 
 If a rock shaft 
 is used between 
 
 the eccentric rod and valve rod, as usual with American locomotives, the 
 same diagram can be used, but the crank arm must always be con- 
 sidered as if it were on the other side of the shaft from that which it 
 actually is. 
 
 Now this construction is not any better than the one given before in 
 Fig. 3, as far as finding any one relation, or the position of valve from one 
 given crank position; but it is more comprehensive and we can easily fol- 
 low the backward and forward movement of the valve as we have the 
 rotating motion of the crank, the valve always moving to and fro on its 
 seat, while its distance from the centre of its movement is seen always to 
 be the amount cut off on the crank arm by the distance circle. Figure 6 
 shows a longitudinal section of the slide valve as usually constructed. 
 When placed in its middle position it completely covers the cylinder ports 
 
20 
 
 STEAM USING; Oil, STEAM ENGINE PRACTICE 
 
 and projects beyond them on both sides. The name given to the exter- 
 nal projection is lap, or steam lap, and to the internal projection, exhaust 
 
 lap. These names are 
 used to denote the linear 
 amount of these projec- 
 tions as well as the pro- 
 jections themselves. 
 
 It is evident from the 
 FJG. fi. diagram that the cylin- 
 
 der port at one end can 
 
 not be opened to the steam until the valve shall have moved away from that 
 end a distance equal to the steam lap; nor can the same port be opened 
 to the exhaust until the valve is moved from the centre towards the port 
 a distance at least equal to the exhaust lap; and if we lay one of these laps 
 on the crank arm it will describe with them two circles upon the centre 
 C. Introducing this element into our diagram we are ready to'answer our 
 next problem: 
 
 During what portion of the revolution is the cylinder open to the steam 
 and to the exhaust, the angle between the crank and eccentric arms, the 
 travel, and the laps being given? 
 
 Lay off in Fig. 7 the full travel A B bisecting in C, and draw the line 
 D C E through C, making the angle D C A equal to the angle between 
 
 FIG. 7. 
 
 the crank and eccentric arms. On D and E, as centres, draw the two dis- 
 tance circles with D C and E C as radii. With the given steam and ex- 
 haust laps as radii, draw circles, or portions of circles, on C as a centre, 
 
VALVE 
 
 21 
 
 cutting the distance circles in eight points. Through these eight points 
 draw radii from C. When the valve distance is greater than the lap circle, 
 one side is open to the steam; and when the valve distance is greater than 
 the exhaust lap, one side is open to the exhaust. When the valve distance 
 is less than the steam lap, the steam is closed at one end and the exhaust 
 at the other end is open, or closed, as the valve distance is greater, or 
 less, than the exhaust lap. 
 
 From F to G the steam is open at the left end, and from F' to (?' at the 
 right end. From H to I the left end exhaust is opened, and from H' to /' 
 the right end. For an engine without a rock shaft the cylinder is sup- 
 posed to be at the right of the diagram, and for an engine with a rock shaft 
 the cylinder is supposed to be at the left. The rotation is from D to A 
 through G. 
 
 When the valve has moved to the right the right exhaust and left 
 steam ports are open. When the valve moves to the left the left end opens 
 to the exhaust and the right end opens to the steam. The valve moves to 
 the right, if there be no rock shaft, when the crank cuts the circle D, and 
 to the left when it cuts the circle E. With a rock shaft we may consider 
 the crank arm to be moved 180 on the shaft without other change in the 
 diagram. In this problem there are various points involved. First, the 
 angle between the crank and eccentric arms; second, the travel of the 
 valve; third, the position of the crank when the steam opens; fourth, the 
 steam lap; fifth, the position of the crank when the exhaust opens; sixth, 
 the exhaust lap; and we can, of course, with all these data, solve other 
 important questions, one or two of which we shall introduce and con- 
 struct. 
 
 Given, the travel and position of crank for opening and closing to the 
 steam; to find the angle between the crank and eccentric arms: 
 
 In Fig. 8 set off the 
 trave A B bisecting at C, 
 and through C draw the 
 <; two positions of the crank 
 
 when the ports are to 
 open and close to the 
 steam, C H and C F re- 
 spectively. Bisect the 
 angle H C F by the line 
 C D. Then will D C A 
 be the required angle 
 between the crank and 
 eccentric arms. 
 
 Given, the travel of 
 the valve, angle between 
 crank and eccentric and 
 point of closing to steam,; 
 to find the lap. 
 In Fig. 8 set off A B equal to the full travel, bisecting in C; layoff A CD 
 
 l> 
 
 FIG. 8. 
 
22 STEAM USIXH; VK, STEAM ENGINE PliA < 'TI< 'K. 
 
 equal to the angle between the crank and eccentric arms; draw the distance 
 circle from Eas a centre through C with radius J^equal to A B, and draw 
 C F, the required position of crank at point of cut off. C G is the steam 
 lap, for where C F cuts the distance circle the valve is closing the steam 
 port. Drawing the arc G I from C as centre, with C G radius, we find the 
 steam opens at C I a little before the crank gets to the dead point. The 
 distance J Kis the amount the valve is open at the end of the stroke, and 
 the opening, and amount of opening, are known as steam lead. If this 
 opening be thought too great the eccentric must be moved on the shaft 
 and the lap found again as before. As D C always bisects H C F, this pre- 
 sents no difficulty. And it is to be remembered that D J is always at 
 right angles to A B. 
 
 Given, the travel, lap and lead; to find the cut off: 
 
 Set off in Fig. 8 the full travel A B bisecting at C. From C lay off 
 C K equal to the lap, and make K J equal to the lead. Draw J D at right 
 angles to A B, and make C D equal to A C, equal to C B, by taking C as a 
 centre and striking an arc with A C as radius cutting the perpendicular 
 J Din D. Bisect C D in E and draw the distance circle through D Jaud 
 C, using E C as radius. Draw also the arc I K G with the lap C K as 
 radius, and draw C F through G. C F is the position of the crank at 
 cutting off steam. 
 
 A little practice with this method, first upon actual valves, and then 
 by combining the foregoing problems, introducing also the exhaust, will 
 soon give a feeling of confidence not easily obtained with the usual 
 methods. 
 
 There is yet one case with a common slide valve which it is desirable 
 to examine. In the foregoing examples we have supposed the eccentric 
 rod to be nearly parallel to the line joining the centre of the shaft and the 
 crosshead, either by placing the steam chest on the side of the cylinder, or, 
 if it be on the top of the cylinder, by the use of rock. shafts. Now the 
 latter are always ugly, and, although much used, the former arrangement 
 is to be preferred when possible. There are, however, cases in which it is 
 convenient to lead the eccentric rod in an angle to the line from the shaft 
 centre to the crosshead, the steam chest being on top of the cylinder and 
 the valve rod jointed and guided; or else the steam chest is at one side of 
 the cylinder, but is above or below the centre. 
 
 The action of an oblique connecting rod is seldom explained, or more 
 rarely fully understood, and we shall therefore devote some attention 
 to it. Looking at the general case in Fig. 9, we see that the further 
 end of the connecting rod can come no nearer the shaft than the differ- 
 ence in length between the rod and crank, and can go no further from the 
 centre of the shaft than the sum of the lengths of the rod and crank, and 
 that these are absolutely the only limits imposed by the crank. Thus, in 
 Fig 9, by taking C B as the length of the connecting rod, and C A as the 
 length of the crank, by drawing from C as a centre two arcs with radii 
 equal to A C + C B and C B - A C, the only limit to the stroke is that its 
 ends shall lie on these lines and that it should take place between them. 
 
ox VALVE (;I-:MI. 
 
 23 
 
 The path of the outer end may be straight or curved. For slide valves 
 worked in this manner it is usually straight, the end moving in guides; or 
 nearly straight with attachment by a comparatively long link, so that the arc 
 it moves in is flat and close to a straight line. Suppose it is straight, then it 
 can be seen that the length of the stroke produced by a given crank may 
 be varied considerably, as, for example, at D E and F G, the latter being 
 plainly the greater. Another feature is also introduced: that is, that for a 
 uniform revolution of the crank, the times of forward and backward 
 strokes, which are the same for D E, are not equal for F G, because the 
 
 FIG. 9. 
 
 dead points, which always occur when the crank and rod are in the same 
 line with each other, when the motion changes, are, for the stroke F G, at 
 the points J and K. These are not on the same diameter, and it will take 
 longer to pass from J to K and F to G than K to J and G to F, the revolu- 
 tion being right-handed. This is taken advantage of as a "quick return 
 motion" in some slotting machines. If, now, the middle of the stroke F G 
 at H be found, and the straight line H L A r C be drawn, the dead points K 
 and J will not lie on this line but near it, and the longer the rod is com- 
 pared with the crank, and the smaller the angle E C H, the closer will be 
 the agreement; and when the crank is on this line the other end of the 
 connecting rod will be close to the points F or G, as the case may be. 
 If the motion were studied on the line A E only, being a parallel to 
 F G passing through C, it would take place as if moved by a crank arm 
 C A instead of C L, which is at the angle ACL from the other, and, in 
 fact, we may call C A an equivalent crank for C L, for it will cause the 
 stroke D E to be made at the same time as C L moves F G, the rod coming 
 to D in one case when the other comes to F, and to E when the other 
 arrives at G. 
 
24 
 
 STEAM T'SrXG; Oh', STEAM ENGINE P 7? AC 'TIC 'E. 
 
 In applying the valve diagram to an engine of this kind, the only 
 change we have to make is, that instead of using the actual angle between 
 the crank and eccentric arms, we must use in its place the angle between 
 the crank and the equivalent eccentric arm; that is, it must be changed by 
 the angle between C A and C L, or A C L, in other words, by the angle at 
 the centre of the crank shaft between the lines there from one to the 
 centre of the travel, and the one to the place where we have assumed our 
 investigation to be made. These lines may also be said to pass from the 
 centre of the shaft through the average dead points and the other through 
 the equivalent dead point. This change will be an increase or decrease 
 in the angle between the crank and eccentric arms, or rather, the travel 
 line and the line joining the centres of the distance circles, according as 
 the rotation is right or left-handed, there being no rock shaft used. 
 
 FIG. 9A. 
 
 Such engines are not very common, but the only objection lies in the 
 necessity of guiding the end of the valve rod. 
 
 In the preceding paragraphs we have discussed the case of an eccen- 
 tric fixed on the shaft, and moving a single slide valve, in such a manner 
 that no change is made in the relation or movement of parts while the 
 engine is in motion, all adjustments being made during rest. This 
 also may be taken to apply to the old-fashioned "D" slide, to "B" valve, 
 and to piston valve engines, and these types are in most general use. 
 The advantages of simplicity, durability and universal acquaintance are 
 
OJV* VALVE GEAR. 
 
 found to outweigh many points, even the waste of steam. The desire to 
 change the points of cut-off, either with the engine stopped or in motion, 
 and the necessity of reversing the direction of rotation in many engines, 
 have led to the use of two eccentrics working one or two slide valves, and 
 in some cases to three eccentrics working two slides. The necessity of 
 reversing was the cause of the adoption of two eccentrics before any change 
 in the expansion was considered. The first reversing gear used with the 
 slide valve was probably some form of hook attachment to a rock shaft 
 carrying pins on both sides of the centre. If a valve be made without lap 
 or lead the steam follows the piston full stroke, and the position of an 
 eccentric is 90 in advance of the crank, to reverse the position of the 
 eccentric it must be changed 180. 
 
 FIG. 9B. 
 
 When, however, there is lap, it is evident that there can not be an equal 
 opening for forward and for back gear with eccentric arms 180 apart; but, 
 as we have already stated, the eccentric arm is more than 90 in advance of 
 the crank, and all that is required is to use the same angle between the crank 
 and eccentric arms for each motion. Figs. 9 A and 9 B show two arrange- 
 ments for accomplishing this with a single eccentric. The eccentric is not 
 keyed to the shaft, but is bolted to the face of a disc of smaller diameter, 
 rigidly fastened to the shaft. In one of the arrangements shown, the 
 eccentric can move freely about one of the bolts as a centre, while the 
 other projects through a curved slot. This second bolt can either be 
 tightened, holding the eccentric to the disc in rny position between the 
 
26 
 
 STEAM rsi.\(, : (>/,', STEAM ENGINE PRACTICE. 
 
 2 C8 
 +a O .W 
 
 (*!{ 
 
 sii 
 
 I a -2 i 
 
 . .2 
 
 rQ CD CD 
 
 fl .2 p 
 
 * S 
 
 -O-'S 
 
 <u o Q a> 
 
 1.2^ 
 
 o g a 
 .2 'C * B 
 
 11 
 
ON VAL VE GEAR. 27 
 
 two extreme ones, thus allowing a change in the eccentric arm when the 
 engine is still; or, the engine being moved till the eccentric comes to rest, 
 with the end of the slot against the bolt, of course the eccentric will 
 follow the motion of the shaft, and the crank and slot being properly 
 arranged, the valve will keep the engine running in the direction in which 
 it was started. If, on stopping, the engine be moved by hand so as to run 
 in the other direction, the shaft and disc will move until the eccentric 
 is caught up with, and the engine will drag it properly to continue the 
 motion. It is therefore a simple reversing gear for engines small enough 
 to be easily turned by hand. Sometimes a small fly-wheel has been 
 mounted with the eccentric and a quick closing valve placed on the steam 
 pipe. By closing the steam valve quickly the fly-wheel will carry the 
 eccentric round so that the engine would start in the other direction; while 
 011 the other hand, if the steam valve be closed slowly the eccentric will 
 remain in contact with the disc and the engine will start in the direction 
 it had been moving an arrangement hardly promising to remain long in 
 working order. In larger engines the same curved slot, or the straight 
 slot shown in the other figure, 9 B, is used, and the eccentric is traversed 
 by a sleeve and key; the latter is spiral and is so arranged that by moving 
 the sleeve along the shaft the eccentric is shifted across the shaft. The 
 movement of the sleeve is effected by a yoked lever. 
 
 The construction of the valve diagram for different positions of these 
 eccentrics, and the change in the cut-off produced thereby, is a very 
 desirable study for the student. Where two eccentrics are used for moving 
 a slide valve they are usually fastened to the shaft, and the rods are 
 attached to the eccentric by straps or yokes; the other ends of the rods 
 are connected by pins to a piece called the link. There are three 
 varieties of the motion which goes under the name of link motion, but 
 of late years one or both of the eccentrics have been omitted and the 
 link moved from some other portion of the moving mechanism. The 
 three kinds noted have been long in use, and are known as the shift- 
 ing link, being the invention of Howe, a foreman in Stephenson's 
 Locomotive Works, or the Stephenson Link, the Gooch, or Fixed Link, 
 produced about the same time in the locomotive shops of Mr. Daniel 
 Gooch. and the straight link of Alexander Allan, a combination of the 
 other two. 
 
 Fig. 10 shows clearly the arrangement for the shifting link in its most 
 simple form. The link joining the ends of the two eccentric rods is slot- 
 ted out and the valve rod is pinned to a rectangular block, sliding in the 
 slot, but which in this simple form can be clamped to the link. When 
 loose, the link can be moved by the handle at its end until it occupies any 
 given or desired position. It will be seen that if the link be placed with 
 one end of the slot against the slider, the valve will move by the eccentric 
 connected to that end of the link almost entirely, and we shall have to ex- 
 amine the motion when the slider is clamped at some intermediate point 
 between the ends. In order that the centre of the valve may remain 
 unchanged for all positions of the link, a curvature must be produced in 
 
28 
 
 STEAM 
 
 07?, STEAM ENGINE PRACTICE. 
 
 the slot; otherwise, as the link is moved, say from its end to the centre, 
 the valve is pulled over toward the shaft. The valve still has the same 
 motion as a whole, its travel is simply displaced, thereby producing 
 unequal distribution of the steam to each end of the cylinder. By curv- 
 
 FIG. 10. 
 
 ing the link to an arc with a radius equal to the distance from the centre 
 of the shaft to the centre of the slider when the valve is placed in its cen- 
 tral position on its seat, this action is removed. 
 
 In Figure 11, for the sake of clearness, consider the link as a straight 
 bar, the extremities of which are hung in such a manner that the ends L 
 and L' can only move in lines nearly parallel to the line A C B; this may 
 be the case exactly for either, or nearly, for one or both of the points 
 L L' '. Inasmuch as C L and C L' both make angles with C A, both ends 
 
 FIG. 11. 
 
 of the link may come under the case of oblique eccentric rods. But 
 this presents no great difficulty. If the link were moved till L came to 
 the slider M, the motion received by the valve would be that due to the 
 
OX VALVE GEAR. 
 
 29 
 
 eccentric E; but for the given position, the point L receives motion 
 parallel to A C as if it were driven by a virtual eccentric at F, with angle 
 between crank and virtual eccentric arm, A C F equal to L C E, Avhile the 
 point L' receives motion from E', as if moved by a virtual eccentric at E*, 
 with the angle A C ^ equal to the angle L,' C E J ' . In general 
 
 A CF+ A CF f = L CE+ L' C F/ . 
 
 E C F/-(A C F + A C F') = ECF/ (L C E + L' C E^); 
 or, EC F+ F/ CF- = A CL + A C L' = L C L' = the link angle. 
 Hence, also, the angle, F C F' , between the virtual eccentric centres is 
 constant, and it swings around its vertex C as the link moves. If the link 
 be divided in any proportion, the link angle should be divided in the same 
 proportion ; and an angle may be set proportionately to the number in 
 which L M divides L L' . L and L' then move exactly as if they were 
 on the line C A driven by the virtual eccentrics, F and F' ; for E comes 
 to its dead points when F comes to C A, and I?", when F comes to C A, 
 the dead points for E and F being on, or very near the lines C L and 
 C Li' respectively. 
 
 We have now to examine the motion of a point on a bar, when the 
 bar is moved at two points, as if connected with a crank or eccentric 
 at each of such points. We have established for the points of connection 
 L and L', the virtual eccentrics from which they receive motion, as far as 
 the line A C is concerned, and we come to the motion of the point M, as 
 follows: 
 
 The motion of M may be found by considering that, if L f were fixed 
 while L moves, the motion of M would be definite; and, also, if L were 
 fixed while L' moves, the motion of M would again be definite; while if 
 L and Lf both move, the point 3/ would have a motion equal to the 
 
 resultant of these two motions. 
 Now, enlarging a part of our Fig. 
 11, we see in Figs. 11 and 12, using 
 the virtual eccentrics F and F*, 
 as it has been shown we must do, 
 that if Lf be fixed, F/ Lf discon- 
 nected, and C K be made the same 
 A part of C F that L' M is of LL', 
 
 the point M will move as if driven 
 by an eccentric with centre at K 
 and a rod at M K; and also that if 
 L be fixed while 'E L is discon- 
 nected, and C K f be made the 
 same part of C F f that L M is of 
 L L', the motion of M will be as if 
 derived from a single eccentric, 
 centre at K', by a rod K' ^f. To combine these two motions at once, 
 draw K N equal and parallel to C K' ; then does the point A" revolve 
 about A" as K' does about (\ And also draw K' N equal and parallel to 
 C K\ then does the point N revolve about K f as K does about C. Either 
 
 FIG. 12. 
 
3O 
 
 STEAM USING; OR. STEAM ENGINE PRACTICE. 
 
 way we look at it N revolves about K while K revolves about C; or N 
 revolves about K f while K' revolves auout O, and hence the point M 
 moves as if directly connected with the point N by a rod M N of fixed 
 length. The point ^Yrnay therefore be called the virtual eccentric centre, 
 and C N the virtual eccentric arm, for the point M. It is also seen that 
 the point Nis on the line F F f ', which it divides in the proportion that M 
 divides L L' ; for the triangles F K N, N K' F f and F C F' , are all similar, 
 and the line F F' may be drawn and the point N found at once by making 
 F f N the same part ofFF* that ML is of L I/. This is a more conven- 
 ient construction for the point N than the other, which was, however, only 
 intended for demonstration. It may be necessary again to caution the 
 reader that the components of motions are all understood to be parallel 
 to A O. 
 
 If the link was, as a whole, fixed, or was not to be changed, we could 
 find once for all the virtual eccentric for the end points L and L' , and 
 draw FF f at once. But, as in the shifting link motion, we use all portions 
 of the link which is moved about C A, we have F coinciding with E when 
 L coincides with M, and then N falls on E. As the link is moved the tri- 
 angle F C F f is swung about C, and the point N travels along F F' till 
 the other end, L' ', of the link is brought to ilf when the point N reaches 
 F' which coincides with E f . The line which includes all positions of the 
 point N is a kind of spiral, but is approximated by Kankine to a circle, 
 and by Zeuner, to whom the whole method is due, to a parabola. We 
 will content ourselves with drawing this curve as the arc of a circle, and 
 with finding a third point thereon by which it may be constructed. We 
 have already found F and F / on this curve coinciding vriih.' E and E' 
 respectively; then if, as is usually the case, C F - C E, the middle point 
 of the curve between F and F' is easily found. Set off in Fig. 13 the 
 angles E C F, E f C F' , each equal to one-half the link angle, L C L', then 
 the point found by the intersection of F F f with A C is, for this case, the 
 middle point of the arc E N E' desired, and a circle is passed through 
 
 FIG. 13. 
 
ON VALVE 
 
 31 
 
 these three points. If the link be drawn in mid-gear and we take the 
 intersection of the rods L E, L' E', combining this with our valve dia- 
 gram, we find a complete mastery over the link motion, and we will try to 
 solve some of the cases which are of frequent occurrence. 
 
 Given, the full travel, the laps and the lead in full gear and the link 
 angle; to flnd the mid gear travel and lead, and also the travel and lead with 
 
 points of admission, cut off, compression and 
 release for any given position of the link. 
 
 Set off, in Fig. 14, C L = the lap and 
 L P, equal the full gear lead; from C as a 
 centre swing with a radius = one-half the 
 full travel, the arc E F, and erect the per- 
 pendicular P E cutting this arc in E. Lay 
 off E C F one-half the link angle, or one- 
 half the angle between the end radii of the 
 link itself (which is not shown in the 
 figure), and drop F Q perpendicular to 
 C L. Then C Q is one-half the mid-gear 
 travel, and L Q is the mid-gear lead. By 
 producing E P to E' making P E' = P E 
 and drawing an arc through E Q E' , we 
 have the curve on which the single virtual 
 eccentrics are found. Suppose we wish to 
 find the steam and exhaust openings and 
 closings for that position of the link which 
 is three- eighths of the way from full for- 
 ward to full backward gear. Divide the arc 
 E E" by G, so that E G is three-eighths of 
 E E' measured on the arc. Join C G. This 
 is one -half the valve travel required; from 
 the middle point K of C G as a centre, draw 
 a circle with radius C K; and from C as a 
 centre, draw the lap circles or portions of 
 them; that is, the radius C L = the steam 
 lap, and C V= the exhaust lap; and where 
 these arcs cut the distance circle from A", 
 the points of intersection give, by drawing 
 lines from C, the positions of the crank 
 
 C S f or steam admission. C T for cut-off, C produced for release, and 
 V C produced for compression. 
 
 It sometimes happens that the forward eccentric is the lower one when 
 we make our examination, and in such a case we must lay on the other 
 
 side of C E and C E' to find the virtual eccentric centres Fand F f ; making 
 the necessary construction we have the middle point of the line F F^, or 
 N, as the single virtual eccentric which drives the block in mid-link, and 
 we may pass our curve through E, E' and this point as we did before; but 
 
32 
 
OX VALVE GEAR. 
 
 33 
 
 we find it curved in the other direction, and the mid-gear lead is less than 
 the full gear lead, while in the first case it was greater. 
 
 FIG. 15. 
 
 The fixed link, or Gooch's link-motion, is shown on page 32. A swing- 
 ing rod called radius rod is attached to the valve stem and carries the slider 
 at its free end. This rod is controlled from the foot board in the same 
 way as the link in Stephenson's motion, and the actual eccentric centres 
 
 are the virtual eccentrics. As 
 the eccentric rods do not change 
 their mean angle with the mo- 
 tion line the single virtual ec- 
 centric is found on the straight 
 line E&, Figs. 13 and 15, join- 
 ing the real eccentric centres, 
 and the lead does not change for 
 any position of slider. This gear 
 is a favorite in England, but is 
 rarely seen in the United States. 
 In Fig. 16 the lap and valve 
 circles are drawn for different 
 gears and will be readily un- 
 derstood. 
 
 The difficulty experienced 
 in fitting the curved links for 
 the shifting and fixed link-mo- 
 tions, and the fact that the cur- 
 vature in the link, required to 
 keep the valve at the same place 
 in mid- travel, was in different 
 directions, led Mr. Alexander 
 Allan, then employed at the Crewe shops, to design a link, shown on page 
 32, in which the radius rod was retained, although the link was also shifted; 
 
34 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 and by proper proportions of the rocker arms, attached to link, and of 
 radius rod, a straight link was secured. Of course these straight links 
 were easier to fit than the curved ones. They are in quite general use in 
 England and on the Continent, while in the United States they have been 
 rarely, if at all, used. 
 
 In studying this link-motion we call the angle from the centre of shaft 
 through which the link is shifted a, and set off one-half of it from the 
 actual eccentric centres to find the position of the virtual forward and 
 backward eccentric centres for mid-gear; and we then find the curve on 
 which all our single virtual eccentric centres lie as in the case of the 
 Stephenson link. The curve in this case is flatter, and the change from 
 full to mid- gear lead is less than for the Stephenson link. The mid-gear 
 lead may be greater or less than full gear lead according as the upper or 
 lower eccentric is used for forward gear. 
 
 No drawing is required. The change of cut-off for different positions 
 of the slider is shown by the distance and lap circles as already explained 
 for the Stephenson and the Gooch links. 
 
 The rock shaft arms have to be proportioned to the segments into 
 which the line from centre of shaft to centre about which radius rod 
 vibrates is divided by the vertical through the centre of rock shaft. 
 
 All three of the gears described as having two eccentrics are used for 
 reversing gears as well as for changing point of cut-off; and, in fact, were 
 used only for the former purpose years before their value for the latter was 
 appreciated. For many years these three were the only forms employed. 
 Of late years, however, a variety of gears for varying point of cut-off and 
 for reversing the motion of the engine have come into use, which we will 
 discuss in order of complexity rather than of age. 
 
 One of the most widely extended modifications of the link-motion with 
 two eccentrics is met with in the valve gear of Walschaert, or the Heusin- 
 ger von Waldegg link-motion, shown on page 35. 
 
 In the ordinary form of link-motion we have two eccentrics attached 
 to two points on the link, and we get 'more or less of the motion of one 
 eccentric by connecting the valve rod to different points upon the link. 
 Suppose, however, that while one eccentric attached to the combination 
 arm remains constant as to length of its arm and the angle it makes with 
 crank arm, we have the means of varying the length of the other eccentric 
 arm, which is attached to a second point on the combination arm as shown 
 in illustration. If, now, the valve stem be connected with a third point on 
 the combination arm the movement of the valve can be controlled by 
 giving more or less travel to the variable arm, and can be reversed by 
 reversing the motion of this arm. 
 
 In the Walschaert gear, the link, or combination arm, is moved by 
 an attachment rod from the cross- head, and its lowest point moves with 
 the piston or main crank. This combination arm is attached at the upper 
 end to the valve stem, and just below this attachment it is pivoted to a 
 centre. If this centre were fixed the valve would receive only the motion 
 due to a virtual eccentric in line with the crank, and with an arm which 
 
ON VALVE VEAK. 
 
 35 
 
 
 
 o: 
 LU 
 O 
 
 C/5 
 
 D 
 UJ 
 
 I 
 
 LJ 
 
 I 
 h 
 
 CC 
 
 O 
 
 cr 
 
 LU 
 < 
 
 1 
 
 LU 
 
 I 
 
STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 bears to the crank length the fraction that the short distance from valve 
 stem to pivot is of the length from pivot to pin at cross-head connection. 
 The other element of the motion is obtained by connecting the pivot of 
 this lever to a vibrating slotted link, pivotted at its centre, by means of a 
 rod attaching to a slider in the link. The vibration of this link is given 
 by a small eccentric set at right angles to the crank, or by a return crank 
 from the main crank pin. The effect of moving this slider in the vibra- 
 ting link is to change the travel of the pivot in magnitude, or to reverse 
 its motion; the pivot always reaches the end of the stroke and its centre 
 at the same part of the revolution: i. e., its period of vibration, or motion 
 with respect to the motion of crank remains unchanged, however much 
 the radius rod may be changed; but the amount of movement received by 
 the valve stem is greater than that received by the pivot in the ratio of 
 the whole length of the combination arm to the portion between the pivot 
 and cross-head connection pin. 
 
 To illustrate: Suppose the stroke of the piston is 24 inches, while the 
 combination arm is 26 inches, with the pivot 2 inches from the valve 
 
 stem connection. The 
 amount which the valve 
 moves due to the pis- 
 ton movement only, is 
 24(26-24) 
 
 ^ 
 
 nearly, not exactly, a 
 slight error being intro- 
 duced by the connecting 
 bar from the cross-head 
 to the link, which, in 
 changing its inclination, 
 reduces the travel near 
 the ends where inclined 
 to the stroke. 
 
 If the radius of the 
 eccentric or return crank 
 be 2 inches, and the ex- 
 treme length of the vi- 
 brating link be 6 inches, 
 3 above and 3 below the 
 centre, and the attach- 
 ment of the vibrating link 
 to eccentric is at 2 inches 
 from the centre of the vi- 
 brating link, we can easily 
 draw valve diagram. In Fig. 17, set off horizontally 1 inch, which equals lap 
 
 0/ 
 
 + lead, from the centre C on line A C, and then vertically a distance 
 the distance the slider is from the centre of the vibrating link = 2^ inches; 
 
 FIG. 
 
ON VALVE <;EAR. 
 
 37 
 
 joining this point with C we use this line as the diameter of the valve circle, 
 and by drawing the lap circles we have all the points as before for the 
 opening and closing of the steam ports. This gear has been used in Europe, 
 and was introduced into this country by the late William Mason, who 
 placed it on many of the light engines used for passenger travel at Coney 
 Island beach and about New York. 
 
 In designing the preceding forms of valve gear, many little points 
 arise with regard to equalizing quantity of steam used at each end of the 
 cylinder, and for so arranging the details that the wear shall be a mini- 
 
 FIG. 18. 
 
 mum. A detailed study must be made in each case, while for many practi- 
 cal points the works on "Link and Valve Motions" by Zeuner, and also by 
 Auchincloss, will be found of great service. 
 
 If in Fig. 18 we compel the end of the eccentric rod to move in the path 
 H K we know that the end of the rod will always reach the centre and ends 
 
 i-: . 
 
 FIG. 18 A. 
 
38 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 of the travel in defined time, while with regard to a motion at right angles 
 thereto in this case the end will have none. Now, if the path is from F to 
 G, it can be seen that the end of the rod has a very respectable component 
 vertically, but that it reaches the end and centre of the travel at the same 
 time as before, or nearly so, and the horizontal travel is H K nearly, and 
 the vertical is FH + KG. By changing the angle of the path we find the 
 magnitude of the path may be changed and even the motion reversed, as 
 F passes C K, and at the same time the vertical motion of the rod at the 
 end may be increased or diminished. We have then all the elements for a 
 link in the eccentric rod itself. One end moves up and down with the 
 motion of the eccentric, and the other end moves up and down with the 
 in and out motion and the eccentric rod becomes the link. 
 
 We can, of course, connect the valve stem to a slider, leaving the path 
 
 FIG. 19. FIG. 19 A. 
 
 MARSHALL'S GEAR FOR MARINE ENGINES. 
 
ON VALVE GEAR. 
 
 39 
 
 at the outer end fixed, or we can [connect the valve stem to any point on 
 the link, as F, and change the angle of the path at the end, thus causing 
 the travel to.\vary; the time, however, of the outer end up and down 
 motion being the same as that of the in and out motion. The valve stem, 
 of course, in this figure moves vertically. 
 
 This form has been adopted by Mr. F. Marshall for moving the valves 
 of vertical marine engines. The arrangement is shown in Figs. 19, 19 A, 
 and 19 B.j^In Fig. 19, the form of path adopted for the outer end is the 
 arc of a circle,^the end of the eccentric rod beiDg connected by a radius 
 
 FIG. 19 B. 
 MARSHALL'S VALVE MOTION ENLARGED. 
 
 rod swung from the point 0. The form may be either a slot or the arc of 
 a circle. Fig. ISA shows a straight slot in the face of a disc attached to 
 the end of shaft connected with a reverse lever. 
 
 In Fig. 19, C is centre of shaft; A, crank pin; A D, connecting rod; E, 
 centre of eccentric; E I, link; V. valve stem connection; O L, arm fixed to 
 axis of geared arc Z Z^ at L; W, handwheel which moves the geared arc 
 by means of worm w; O 7, radius rod which controls movement of end of 
 link, or eccentric rod. 
 
40 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 Mr. Charles Brown, of Winterthur, Switzerland, introduced many 
 varieties of this form, see Fig. 20. In most of them the eccentric was 
 dispensed with, and the link attached to the connecting rod in such a 
 manner that while the movement in the direction of the inclination 
 is reduced, as in the Walschaert gear, the other component, that due 
 the transverse vibration of the rod, is the one which is governed by 
 the path used. 
 
 Mr. Brown usually employed some form of parallel motion at the 
 outer end in place of the slotted guide. Most of his work was applied 
 to engines with rods quite long in comparison with the movements 
 taken from them, and both Marshall and Brown attached the valve 
 stem either between the eccentric and guide, or outside of it as de- 
 sired. 
 
 A* 
 
 FIG. 19C. 
 DIAGRAM FOB MARSHALL'S VALVE GEAK. 
 
 Mr. David Joy has introduced, quite extensively, a form of gear 
 similar to Brown's in some respects, but more carefully worked out in 
 others; Figs. 21 and 21 A. Mr. Joy attaches to a point near the centre 
 of the connecting rod a bar H I; Fig. 21 A. This bar is carried by a 
 hanger from J, a fixed point in the frame of the engine. Now, as the 
 
ON VALVE GEAR. 
 
 41 
 
 FIG. 20. 
 BROWN'S VALVE GEAR. 
 
 FIG. 21. 
 
 JOY'S VALVE GEAR ON A 
 MARINE ENGINE. 
 
 cross-head moves from D to E the point 
 / moves in the arc of a circle, but because 
 the length H I is short it gets an extra 
 pull in when the cross- head nears the 
 end of the stroke. Mr. Joy therefore 
 carries his link from K, a point on H I, 
 to the guides L, and attaches the valve 
 stem M N to a point in K L, in this case 
 produced, at M. The results have been 
 very good, and in most cases pin joints 
 are used for eccentric straps. In many 
 cases Mr. Joy replaces the guide L, 
 which he makes curved to radius de- 
 pending on N M, by a swinging link 
 having the same centre as the guide; 
 this centre is mounted on an arm of 
 a rock shaft, and the rock shaft cen- 
 tre coincides with that of the guide. 
 Two more pin-joints are required but 
 the wear on the guide is removed. 
 
 Mr. Kirk has patented a form of 
 valve gear. In a marine engine he 
 places a vibrating link on the air pump 
 side levers in such a a manner that the 
 centre of the link is moved thereby 
 from the piston rod. The link is caused 
 to vibrate by the transverse motion of 
 the connecting rod and a compensation 
 due the obliquity is introduced. In this 
 case the levers adopted are in the 
 
42 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 form of a Watt parallel mo- 
 tion. 
 
 The motion used for a 
 link in the Allen engine for 
 driving a slide valve, and in 
 the Porter- Allen for moving 
 the steam valves, is exceed- 
 ingly elegant and clearly set 
 forth in Fig. 22. The link is 
 part of the eccentric strap, 
 and the centre of the link, of 
 which only one-half is con- 
 structed for non-reversing 
 engines, is guided, in an ap- 
 proximately straight line, 
 by a swinging rod attached 
 to the frame; so that the 
 small versed sine of half 
 the arc of swing is bisected 
 by the centre line of engine 
 and the chord of the arc is 
 parallel to it. The eccen- 
 tric is forged on the shaft 
 and corresponds in Fig. 23 
 to the virtual arm A B. As 
 we proceed up the link we 
 find it also acting as a bell 
 crank, and we see that by 
 coming up the vertical B to 
 G and making B G propor- 
 tional to distance of slider 
 above centre line, we have 
 the virtual eccentric and 
 distance circle for any point 
 on the link. 
 
 To draw the valve dia- 
 gram for any point on the 
 link, lay off first the real 
 eccentric radius on the 
 motion line, and then from 
 that point lay off at right 
 angles to the motion line a 
 distance equal to the real 
 eccentric radius multiplied 
 by the ratio of the distance 
 of the slider above centre 
 line to the distance from 
 
 FIG. 21 A. 
 JOY'S VALVE GEAK ENLARGED. 
 
ON VALVE GEAR. 
 
 43 
 
 FIG. 22. 
 
 centre of eccentric to the point of attachment of rocker arm. The point 
 thus found is the virtual eccentric centre moving the valve. The link is 
 of course curved to the radius of length equal the radius rod. For a 
 reversing gear the link is continued beyond the centre or pin from which 
 it is hung. 
 
 Another form of gear was used in Germany by Herr Kaiser, of Berlin, 
 and is very simple. Two lugs project from the eccentric strap at right 
 angles. To one of them the valve stem is attached, and the other is 
 
 guided in a slot which can be placed 
 at different inclinations with line of 
 motion.^, When the slot is in the line 
 of the motion, the valve is moved as 
 if by a single eccentric found in the 
 diagram as follows: 
 
 Set off on the.line'of the motion 
 the real eccentric radius, and at right 
 angles to the motion, a distance 
 equal to the real eccentric radius 
 multiplied by the ratio of the dis- 
 tance from the centre of the eccentric 
 to the centre of pin in the valve stem, 
 to the distance from the centre of the 
 eccentric to centre of pin in slider. 
 
44 
 
 STEAM VSING; OR, STEAM ENGINE PRACTICE. 
 
 When the slide is inclined, the amount of vertical movement of the 
 slide block caused thereby must be added to the eccentric radius before 
 multiplying length of the lugs by the above ratio: as the sum of these 
 
 two movements is 
 used it is apparent 
 that this motion is 
 not well suited for 
 a reversing gear. 
 
 Instead of a 
 slide a swinging 
 link is often used, 
 and by moving the 
 point of suspen- 
 sion in an arc with 
 a centre at the 
 middle point of 
 the motion a very 
 good distribution 
 is obtained. 
 
 FIG. 24. 
 
 In all the foregoing motions, or valve gears, we have considered the 
 valve as the ordinary slide. By examining the valve diagrams already 
 given, we find that when an early cut-off is given by the use of lap the ec- 
 centric has to be set forward, and that either release occurs very early, or 
 if this be prevented by giving lap on the exhaust 
 side the exhaust closes early and cushion begins. 
 Now just where it is best to stop in either direc- 
 tion has not been decided, but the greater the 
 clearance and number of revolutions the earlier 
 the cut-off can be used with advantage. With 8 
 per cent, clearance and less than 100 revolutions 
 it is not desirable to cut off before half stroke, 
 but if the speed be increased to over 300 revolu- 
 tions a cut-off at one-fourth stroke may be em- 
 ployed. If the travel of the valves on a loco- 
 motive for full gear be 4| to 5 inches, for a lead 
 of i"e inch at full gear, and -, 5 6 inch at mid gear, 
 a steam lap of f inch and no exhaust lap will 
 
 secure excellent results; but if the engine was never to run at a speed 
 of over 20 miles an hour, an exhaust lap of | inch could be used to 
 advantage. Most builders of stationary engines give so much exhaust lap 
 that a considerable back pressure is caused, and the engines can not be 
 run at high speed, and for two reasons: 1st, that the steam does not get 
 out of the cylinder fast enough, and, 2nd, there is not enough cushion to 
 tase up the fling of the connections at high speed. An early release and 
 strong cushion are required for high speeds. 
 
 At moderate speed an early release and strong cushion deadens the 
 motion of the engine over the centres, and the use of two slide valves. 
 
 u 
 FIG. 25. 
 
O.V VALVE 
 
 45 
 
 one on top of the other, was suggested by Meyer. A false valve seat was 
 suggested by Eankine, with the object of obtaining a quicker cut-off, 
 the seat being moved by one eccentric while the valve was moved by 
 another. In this way the effect of an eccentric with greater throw was 
 obtained. 
 
 The first change consisted in making the steam chest in two cham- 
 bers. In the one next the cylinder the ordinary slide was employed while 
 the steam came in through openings from the other chamber, these open- 
 ings were covered by a simple slide moved by an eccentric. Thus the 
 inlet and exhaust were regulated by the ordinary slide, but the second one 
 cut off the supply of steam. As the principal objection to this was the 
 large clearance space left in the main steam chest and the consequent 
 waste of steam, the Meyer gear became the favorite. 
 
 isnffifcz^ 
 
 FIG. 26. THE MEYER YALVE. 
 
 The use of an expansion valve on the back of the main valve allows 
 the main valve to govern the admission, release, and cushion, but the cut- 
 off is effected by the expansion slide closing the steam ports of the main 
 valve. This combination enables the cut-off to take place more quickly 
 with the sum of the motions of the two valves. 
 
 In Fig. 26, A A are the steam ports and B the exhaust port in the cyl- 
 inder metal; C C are the steam ports and D the exhaust port in the main 
 slide. The main slide is moved by rod E, the plates G G, on the top of 
 the main slide, by rod F. Steam is usually admitted by the outer edges of 
 the cut-off plates. 
 
 The eccentric of the expansion valve is usually placed in line with the 
 crank, either on the same or opposite side of the shaft. This, however, is 
 
46 
 
 STEAM USING; OR, STEAM ENGINE^PRACTIGE. 
 
 not essential, for there are four ways in which the cut-off may be varied by 
 means of an expansion valve: 
 
 1. By changing the lap of the expansion valve, usually by means of a 
 right and left thread on the valve stem. 
 
 2. By changing the travel of the expansion valve, usually by means 
 of a radius rod joined to the valve stem, and a rocker link moved by the 
 eccentric. 
 
 3. By moving the eccentric round the shaft, of which, perhaps, the 
 "Buckeye Engine" is the best example. 
 
 4. By the use of a link motion for the expansion valve. 
 
 FIG. 27. 
 
 We shall examine hereafter these methods more fully; but to begin 
 with will take up the case in which the expansion valve is without lap, 
 and is moved by an eccentric placed opposite the crank, the main valve 
 being without lap or lead, or, "line and line," having, of course, its eccen- 
 tric set at right angles to the crank. The valve diagram may be drawn 
 from the known position of the eccentrics, each valve being represented 
 by its own distance circle. 
 
 In Fig. 27, let C B be the diameter of the distance circle for the main 
 valve and C A the diameter of the distance circle of the expansion valve. 
 
ON VALVE GEAR. 
 
 Then, it being remembered that with the piston at the right end of the 
 cylinder, the position of the crank arm is C A, the expansion valve is 
 farthest from its mid position and the main valve is at its mid position. 
 As the crank moves on towards the position C F, the main valve rapidly 
 opens the port while the expansion valve moves inward, at first slowly, but 
 with increasing speed. At C F it is evident that the main valve and ex- 
 pansion valve are at the same distance from mid position. If, therefore, 
 there be no lap on the expansion slide, it will at this point cover and 
 
 close the opening in the 
 main valve, and C F is 
 therefore the position of 
 the crank at cut-off. The 
 release of steam to the 
 exhaust, being under the 
 main valve, can in no way 
 be dependent upon the 
 \A* upper, or expansion 
 valve; but there is one 
 thing to be carefully 
 guarded against, which 
 with this arrangement 
 might happen: the main 
 valve moves on outward 
 and the expansion valve 
 inward, until at C B the 
 port in the main valve is 
 wide open, while the ex- 
 pansion valve is at mid 
 position. The continuation of the motion draws the main valve back, grad- 
 ually closing the port while the expansion valve is now moving beyond mid 
 position and would of itself cover the cylinder port if placed thereon. At 
 C D, for instance, the main valve has not yet returned to the centre by the 
 amount C E, -while the expansion valve is past the centre by the amount C 
 E? '. We see that in this case there is no risk of opening the steam before 
 the end of the stroke is reached, which was the danger to be shunned. Of 
 course, at the end of the stroke steam is admitted to the other end of the 
 cylinder by the main valve, its port at that end having been uncovered by 
 the expansion valve when the crank was in the position (7 D'. 
 
 If with the given eccentrics we should desire to have the cut-off take 
 place before the crank reach C F, we must add lap to the expansion valve 
 so that its edge shall meet and cover the port in the main valve before the 
 expansion valve becomes central thereto. On the other hand, if the cut- 
 off is to come after C F. a strip must be taken from the edge of the expan- 
 sion valve which must meet the port of the main valve after the slide has 
 become central thereto, or the lap must be negative. We will consider 
 this case more fully hereafter. With the simple figure we readily see the 
 effect of changing the angular position of the eccentrics on the shaft, 
 
 FIG. 28. 
 
48 
 
 STEAM USING; OK, STEAM ENGINE PRACTICE. 
 
 for the main valve is rarely set without lead cushion or a more prompt 
 release. 
 
 If the expansion eccentric be brought nearer the main eccentric, or 
 moved backwards on the shaft, we see in Fig. 28 that the intersection of 
 the two circles takes place later, and that it may range to the end of the 
 -stroke, or more strictly speaking, to the cut-off given by the main valve 
 due to its lap. 
 
 FIG. 29. 
 
 It may also be seen, that in the use of an expansion valve without lap 
 the effect of a change of travel, see Fig. 29, in the expansion valve, the 
 position of its eccentric remaining the same, will also vary the cut-off 
 through a considerable range, but in this case we must expect inconven- 
 ience to arise from the greater length of steam chest required to accom- 
 modate the increased travel. 
 
 The effect of changing the lap on the expansion valve is best examined 
 by combining the distance circles of the two valves as follows, see Fig. 30: 
 
 Take as before, C B and C A, the distance circles of the main valve and 
 expansion valve, respectively. Join A B and draw C G parallel to A B, 
 and B G parallel to A C. Upon C G as a* diameter draw a circle. Then, 
 for any position, as C D, of the crank, it may be easily shown that the dis- 
 tance of the centre of the main valve from the centre of the expansion 
 
ON VALVE GEAR. 
 
 49 
 
 FIG. 30. 
 
 valve, CD C E = E D, is equal to the chord, C L, of the arc of this last 
 circle intercepted by the crank arm. Therefore, the circle C G may be 
 regarded as a resultant distance circle giving the position of the expan- 
 sion valve upon the back of the main valve, without regard to the motion 
 of the latter. If there be no lap on the expansion valve, we find it central 
 with the main valve at C F, which is at right angles to A B; and we have 
 already seen that with positive lap the cut-off takes place before C F, while 
 with negative lap it takes place after CFha.s been passed. Thus we see 
 that by cutting from the edge of the expansion valve, or by increasing the 
 negative lap, we can delay the cut-off till C G is reached, at which point 
 the expansion valve just closes the port in main valve for an instant only, 
 and the steam continues to pass into cylinder until main valve closes. In 
 order that the expansion valve may not open again before the main valve 
 closes, it must close the port in the main valve at C B, or at half stroke in 
 the case before us; that is, the distance of the edge of the expansion valve 
 from the far edge of the port in main valve, when both are in mid position, 
 must not exceed C B = C A'. This distance evidently depends for its 
 value upon the lap of the expansion valve. 
 
 If the cut-off is variable, its maximum limit should not be beyond the 
 point of cut-off of the main valve; if it is coincident with that of the 
 main valve, it is evident that the diameter, C G, coincides with the posi- 
 tion of crank arm when the main valve closes, and equals the distance of 
 
50 
 
 AM USING; Oil, STEAM ENGINE PRACTICE. 
 
 the edge of expansion valve from far edge of port as before stated. We 
 have thus a limit which we did not meet with in our former case. The 
 remedy is to increase the throw of the expansion eccentric, thereby ren- 
 dering the angle C A B more acute, or else, of course, to move the expan- 
 sion eccentric nearer the main eccentric. 
 
 By proper use of the distance circles and resultant circle all problems 
 on the Meyer valve gear may be readily solved. 
 
 NOTE. Figures 31, 32 and 33 work from the left instead of right, as heretofore. 
 
 We will illustrate by a few examples: 
 
 Given, the lap and lead of the main valve, and the travel of both valves; 
 to find lap of the expansion valve for a given cut-off. 
 
 First, in Fig. 31, from C set off horizontally C D the lap and D E the 
 lead of the main valve, and with C F the half travel and E F a vertical line, 
 construct the right angle triangle C E F. On C F as a diameter draw the 
 main valve distance circle, and the lap arc D K f defines the crank position 
 C K when the main valve closes. Set off C G the half travel of the expan- 
 sion valve with the eccentric opposite the crank, and join G F. On C H, 
 equal and parallel to G F, draw the resultant distance circle; and where it 
 meets C I, the position given for the crank at cut-off, gives CD, the 
 negative lap of the expansion valve, which we carry round to J in order 
 to see that by the time the expansion valve again uncovers the port of the 
 main valve at C J, the main valve has already closed the cylinder port 
 at C K. 
 
ON VALVE GEAR. 
 
 51 
 
 Given, the lap, lead and travel of the main valve, and the travel of an 
 expansion valve having no lap; to find position of expansion eccentric to 
 produce a given cut-off. 
 
 Find in Fig. 32 the distance circle and closure of the main valve, as 
 before, and let C I be the crank for given cut-off. With I, where this inter- 
 sects the distance circle, as a centre, and with C as the centre, swing radii 
 C A and I A, each equal to the travel of the expansion valve, denning by 
 
 FIG. 32. 
 
 their intersection at A, the centre of the distance circle for the expansion 
 valve passing through I and C. The position of the expansion eccentric 
 is 180 angle E CA ahead of the crank, or distant from the main eccentric 
 by the angle F C A. If the point A falls above C M drawn at right angles 
 to C K, the slide will not open the port until after the main valve has 
 closed. But if A falls to the left of C M, a new eccentric must be taken. 
 
 Given, the same data, viz.: lap, lead and travel of main valve, and lap 
 and travel of expansion valve; to find position for expansion eccentric to 
 produce a given cut-off. 
 
 In Fig. 33, draw C F and C Kas before, and on C /, the given position 
 of crank at cut-off, set off C A, the negative lap of the expansion valve. 
 From A draw A J at right angles to C A, and, with the half travel of the 
 expansion valve, define from E on A Jthe point L. C M, equal and par- 
 allel to L F, is the position sought for the expansion eccentric arm. 
 
52 
 
 STEAM USING; OK, STEAM ENGINE PRACTICE. 
 
 Other problems will readily be solved if there be sufficient data, but 
 we think we have given the most important, and enough to show the flex- 
 ibility and power of the method. 
 
 When such valves as described are used, the engine is said to have an 
 expansion valve. When the cut-off can be changed while running or when 
 still, the engine is said to have variable expansion. When the cut-off is 
 changed by the action of the engine itself, owing to change of speed, it is 
 called automatic expansion. 
 
 FIG. 33. 
 
 When the cylinder is so long that a single slide becomes inconvenient, 
 two exhaust ports are used, and the valve is divided into two portions, one 
 for each end of the cylinder. Care must here be taken that the exhaust 
 port is not open to the steam by the valve having too much travel. 
 
 Sometimes a piston valve is used, which consists of two pistons on the 
 valve stem, either arranged to give steam space between the pistons and 
 the exhaust connections at the end spaces, or through the hollow piston; 
 or with the exhaust port between the pistons and the steam space at the 
 ends. Examples of these arrangements are given on page 55. The 
 graphical method heretofore used will answer for all these varieties. 
 
 In large engines and in many paddle-wheel steamers four valves are 
 used, generally "Equilibrium poppets," or the older "Cornish equilibrium:" 
 these were introduced at a very early period in the history of the steam 
 engine. The single unbalanced poppet is still used in small engines on 
 the Mississippi, and a very common arrangement is a "relief valve," or a 
 small poppet on the top of a large one, the small one being lifted first and 
 its continued movement raising the large one. By the use of moveable 
 seats and poppets set on the valve stem the "balance" may be carried to 
 
OV VM.VK 
 
 53 
 
 any desired extent. The ordi- 
 nary forms, as also the Cornish, 
 require one end of the valve to 
 pass through the seat for the 
 other end, thus limiting the 
 degree of closeness of the 
 agreement of areas or the "bal- 
 ancing. " 
 
 The stems of these poppets 
 are usually moved by levers, 
 worked by cams on one or more 
 auxilliary rock shafts placed 
 near the cylinder. The valves 
 are always moved vertically. 
 The movement of the rock 
 shafts is usually effected by an 
 eccentric on the main shaft. 
 On the Mississippi river, in- 
 stead of an eccentric, a cam is 
 used. One cam is used for full 
 stroke in either direction, and 
 the reversing is done by hook- 
 ing to either one of a pair of 
 arms on a rock shaft; a second 
 cam is used for cutting off for 
 the steam valves when running 
 ahead only, the exhaust being 
 moved by the full stroke cam. 
 The forms of the cams are such 
 as to give very rapid move- 
 ments of opening and closing, 
 as will be seen from the indi- 
 cator diagrams taken from the 
 steamer Phil. Chappel, shown 
 in Chapter IV. The shape of 
 cams and arrangement of valve 
 gear is shown in the drawings 
 of the engines of the steamer 
 "Montana," as applied to the 
 Mississippi boats, and also their 
 application to Engine No. 1, 
 High Service, St. Louis Water- 
 works, which is an example of 
 the application usual near New 
 York, and on the North Eiver 
 class of boats. 
 
 When poppet jvalves [are 
 
54 
 
 STEAM USING; OK, STEAM ENGINE PRACTICE. 
 
 used with a drop cut-off, there is a shock which causes rapid wear on the 
 valve and seat unless a dash pot is used to prevent it. At a speed of more 
 than 30 revolutions per minute, poppet valves do not appear to give 
 entire satisfaction, but with a small number of revolutions they work well. 
 We have seen them used up to 70 revolutions, but have generally found 
 upon enquiry that very frequent grinding was required. 
 
 FIG. 35. DOUBLE VALVE. 
 
 From the variations in steam pressure and w^ork on stationary engines 
 there necessarily resulted variations in speed, which for many reasons is 
 exceedingly undesirable, and we find that Watt very soon produced his 
 centrifugal governor, applied to the throttle valve in the well known man- 
 ner. Under various forms this arrangement is still the most common one, 
 and for small variations of speed it is perhaps as good as anything yet 
 devised. 
 
 For large variations in speed it was found that with light loads the 
 engine was so throttled that the initial pressure in the cylinder was much 
 below the boiler pressure, and a manifest waste of steam resulted, the 
 steam not yielding the amount of work which might be obtained there- 
 from, and a class of engines 
 with four valves and a "drop 
 cut-off," regulated by the 
 governor, was introduced by 
 Mr. Geo. H. Corliss. 
 
 The engine introduced 
 by Mr. Corliss was in many 
 respects a very great im- 
 provement. The valves were 
 placed close to the cylinder 
 and were rotary instead of 
 sliding. The clearance space 
 was very much reduced and 
 
 the engines became very successful. The drop cut-off regulated by the 
 governor kept the initial pressure of steam in the cylinder well up to the 
 boiler pressure, and changes of speed were followed so closely by changes 
 of cut-off, that in engines well proportioned to the work an economy 
 never before attained was reached. 
 
 In the United States the term Corliss is applied only to engines with 
 the rotative Corliss valve, but in Europe it has been used for any engine 
 with cut-off regulated by the governor, as this was first successfully 
 
 FIG. 36. TKICK'S VALVE. 
 
OX YAI.VK <;EAH. 
 
 55 
 
 VALVE SEAT. 
 
 FIG. 37. 
 
 Exhaust 
 
 PISTON VALVES. 
 
 SECTION THROUGH 
 VALVE SEAT. 
 
 applied by Mr. Corliss. A host of imitators soon followed, each with a 
 variation in the "let-off" gear, and with slide and poppet valves moved 
 by one or two eccentrics. Since the expiration of the Corliss patents a 
 crop of designs has come forward, all distinguished by the appendix 
 "Corliss." Of these we shall illustrate one built by Messrs. E. P. Allis & 
 Co., of Milwaukee, Wis., from the designs of their manager, Mr.' Edwin F. 
 Reynolds, and placed in the St. Louis Cotton Mill. 
 
 FIG. 38. DOUBLE POPPET VALVE. 
 
56 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 A modification of the Corliss engine is manufactured by Mr. Jerome 
 Wheelock, of Worcester, Mass., in which, with only two ports and two 
 main valves, the steam admission is closed by two other valves adjacent to 
 the main valves and worked from them by clutch links; all four valves are 
 of the Corliss type. As built, the only objection to these engines is the 
 difficulty of arranging the cushion to the varied requirements of practice. 
 
 The production of an engine with cut off regulated by the governor, 
 but without the let-off, or drop gear, which, by the way, is not adapted to 
 higher speeds than 120 revolutions, has been secured in the Buckeye and 
 the Porter- Allen engines. In the former, two ports, a balanced slide valve 
 and an expansion slide are used, the cut-off being changed by the gover- 
 nor, and that by turning the expansion eccentric round the shaft by the 
 centrifugal force of two weights held back by springs in a manner easily 
 understood from the drawings. In the Porter-Allen engine there are four 
 ports and the steam valves are balanced slides moved by an expansion 
 link, already explained, and the exhaust is moved from the end of the 
 same link. We shall give illustrations of these in Chapter IV which will 
 more fully explain their principles. 
 
 In large engines using slide valves the ports are often made double, 
 the valves having passages cast in them connecting the steam and exhaust 
 ports. This in no way affects the valve diagram constructions already 
 given. 
 
 When the cylinders are over 12 inches in diameter it is well to connect 
 the metal across the ports by bridges, so that the heads shall not be ren- 
 dered weak, or the pull on the bolts concentrated on those next the ends 
 of the ports. 
 
 In some excellent examples of engines with four ports, gridiron slides 
 are used, working transversely to the cylinders. These slides are moved 
 usually by cams on a "lay shaft," and the cut-off is changed by shifting 
 the cams along the shaft, bringing different portions of the cam face with 
 different angular forces into action. A good example of this is found in 
 the "Howard" engine. By this means a very sharp opening and closing- 
 can be given the valves, but in small engines a larger clearance is required 
 than is desirable. Each valve can be adjusted by itself, which is a de- 
 sirable feature. 
 
 Valve gear of this class has been employed with great success by Mr. 
 E. D. Leavitt, Jr., in his celebrated pumping engines. 
 
CHAPTER III. 
 THE QUANTITY OF STEAM WHICH MIGHT BE AND WHICH IS USED. 
 
 Although we have no such engines as that which was described in the 
 preceding chapter, it is desirable to assume, for the purposes of computa- 
 tion, and in order to obtain a view of complex operations in detail, that we 
 use the steam in a non-conducting vessel whose volume can be varied: in 
 other words, in a cylinder constructed of material which cannot conduct 
 heat. But we must be very careful to remember that no such engine at 
 present exists, and that we shall have to adapt the deductions made from 
 such a case to real engines. Serious disappointment and useless expense 
 have resulted from ignoring this fact in designing engines, and much dis- 
 cussion on the subject has arisen. 
 
 In the case of engines using steam non- expansively, or without "cut- 
 ting- off," we should have very little trouble in computing the work done 
 by a pound weight of steam, or the number of pounds of steam used to 
 obtain a horse-power of work in the cylinder, as we shall very easily see. 
 
 We know that the, equivalent of a unit of heat is 772 foot-pounds of 
 work, and that 33,000 foot-pounds of work per minute is the standard 
 horse-power: 
 
 33,000 75 
 
 Hence, ^- = 42 100 = 
 
 the number of heat units per minute which have to be expended to 
 obtain a horse-power, provided, we had any means of transforming heat 
 into work without waste. And 
 
 75 
 42 100 x 6ft = 2,565 heat units, the equivalent of a horse-power per 
 
 hour. 
 
 In a full stroke engine with non-conducting cylinder, and without 
 other losses, AVC should only have to divide 2,565 by the number of heat- 
 units given in the Table of the "Properties of Saturated Steam, " Chapter 
 I, as heat expended in doing external work, or external heat, for one 
 pound weight of steam at the boiler pressure, to obtain the number of 
 pounds of water which must be boiled per hour to furnish, or to exert, one 
 horse- power per hour for the work of the steam in the cylinder. 
 
 There are many reasons why the amount of steam used per hour, and 
 per horse-power, should vary from this: 
 
 1. The steam passages may be too small to maintain the steam supply 
 at the boiler pressure. 
 
 2. The boiler may be so small in steam room that the pressure in 
 both boiler and cylinder may fall during the stroke. 
 
58 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 3. The clearance, or waste space, in the cylinder will be filled with 
 steam at boiler pressure before the piston starts, which will be discharged 
 at exhaust without having performed any work. 
 
 4. The fact that the cylinder and piston receive heat when the steam 
 comes in and give it out while the steam goes out. More will be said on 
 this point later. 
 
 The power exerted by the steam in the cylinder is subjected to the 
 following further losses before it can be utilized: 
 
 1. The vapor, or steam, on the exhaust side of the piston has to be 
 pushed out of the cylinder by the advancing piston against the back 
 pressure. 
 
 2. Mechanical work is absorbed ( in friction by the piston slides and 
 connections of non-rotative engines, and by the main bearings of rotative 
 engines. 
 
 These several points give rise to the measurement of horse-power in 
 three ways: 
 
 1. The number of horse-power really exerted by the steam called 
 in English the total, and in French the absolute, horse-power. 
 
 2. The number of horse-power exerted by the steam after deducting 
 that expended in sweeping out the exhaust. This is given by the ordinary 
 indicator measurement, and is called the Indicated Horse Power: it is 
 what is usually understood by the phrase Horse-power, without qualifica- 
 tion, in the United States. 
 
 3. The power which may be taken from the shaft by a dynamometer, 
 or by a belt, which represents the power of the engine to do useful work 
 outside of itself. This is called the Net, or Effective Horse Power. 
 
 As these three quantities are for any given measurement produced by 
 the same amount of water boiled, we find that, dividing the quantity of 
 water used per hour by the number of horse-power, we have also three 
 quotients, viz.: The number of pounds used per hour for Total, for Indi- 
 cated, and for Net Horse Power. These may be considered as the "stu- 
 dent's," the "ordinary," and the "commercial" standards of measurement. 
 Engineers are obliged to retain all three of these quantities, for: In any 
 given engine with given speed and circumstances, the work expended on 
 the exhaust and friction is nearly constant, and hence, the greater the 
 power of the steam used the less is the percentage of that uselessly 
 expended. 
 
 In illustration of the process, referred to, we give the following 
 examples: 
 
 Example I. What is the minimum steam consumption per hour for 
 an engine using steam at full stroke, at 25 Ibs. pressure above the atmos- 
 phere, with a vacuum of 12 pounds, and indicating 200 horse-power for a 
 non-conducting cylinder? Clearance 8 per cent. 
 
 Steam pressure = 25 + 15 = 40 Ibs. above zero. 
 Back pressure = 15 12 = 3 Ibs above zero. 
 Pressure expended in obtaining the indicated horse- power is: 
 40 - 3 = 25 + 12 = 37 Ibs. 
 
QUANTITY OF STKAU f'SKfl. ETC. 59 
 
 From the Table of "Properties of Steam" we find that the external work 
 of 1 pound weight of steam at 25 Ibs. pressure is equal to 77 heat units, 
 and we should have to use to obtain a Total Horse-power, 2,565, the 
 equivalent of heat units for a horse-power per hour, divided by 77 as 
 above, which would give the number of pounds of water per hour, 
 neglecting clearance. 
 
 To obtain the indicated horse -power: we have ~ 7 y x ^ = pounds 
 
 per hour. 
 
 If the friction of the engine absorbs as usual 2 Ibs. pressure, we shall 
 
 have 77 x 35 = Ibs. of water per net horse -power per hour. 
 
 40 
 The number of Total Horse-power is 200 x 3 _ . 
 
 37 
 The number of Net Horse -power is 200 x ^ . 
 
 The water used per hour neglecting clearance is: 
 
 2565 40 
 77 < 37 X 200 ' 
 
 and including clearance 
 
 2 | 7 5 x 3 x 200 x 1.08. 
 
 Working out these figures, we shall have: 
 Water per Total Horse-power per hour = 33.3. 
 Water per Indicated Horse-power per hour = 36.0. 
 Water per Net Horse-power per hour = 38.0. 
 Water per hour = 200 x 1.08 x 36 = 7,720 pounds. 
 When the cylinder is not non-conducting a much larger quantity of 
 water will be consumed. 
 
 Example II. The following figures illustrate about as bad a case in 
 actual practice, as ever came within the author's experience: 
 Steam pressure 60 Ibs. above atmosphere. 
 Back 45 tt 
 
 Steam pressure 75 Ibs. above zero. 
 Back 60 " 
 
 Acting pressure on piston 75 60 = 15 Ibs. 
 External heat for 60 Ibs. = 79 heat units. 
 
 Water per Total Horse-power - = 32.4 Ibs. 
 
 2565 75 
 Water per Indicated Horse-power 79 x - = 162 Ibs. 
 
60 STEAM USING; OH, STEAM ENGINE PRACTICE. 
 
 Example III. 
 
 Steam pressure 205 Ibs. above atmosphere. 
 
 Back 2 " 
 
 Steam " 220 " above zero. 
 
 Back 17 " 
 
 Acting " 183 " on piston. 
 
 External work for 205 Ibs. = 85 heat units. 
 
 Water per hour for Total Horse-power g5 = 30.2. 
 
 " for Indicated " -gg- x 2Q3 = 32.3. 
 
 We see from examples I and III that, although in the intrinsic work 
 of the steam there is not much gained by the use of high pressure steam 
 in full stroke engines, 
 
 33.3 30.2 3 
 ~30~ =9^ per cent., 
 
 yet that high pressure non- condensing engines may be practically more 
 economical than low pressure condensing ones; for supposing, as we did, 
 that 3 pounds moved the engine only, we have: 
 
 38.0 33.2 6 
 
 38.0 = 12 f<) P er cent " 
 
 as the practical gain in the cost of a Net horse-power: and there are the 
 further advantages that the loss by the use of a conducting cylinder is 
 less, and the machinery is far less bulky and complex. 
 
 This is a very fair comparison of the early steamboat engines when 
 working full stroke, as used on the Atlantic ocean and the Mississippi 
 river. The use of steam at a very high pressure, not only requiring 
 cheaper machinery and of less weight, but actually using less steam to do 
 the work, when worked at or near full stroke. 
 
 The process of computing the amount of steam used by expanding 
 engines with non-conducting cylinders is much less easy to explain, al- 
 though the work of computation is but little more difficult. 
 
 In order to study the action of expanding gases it is convenient to 
 represent the volume and pressure for any given state and quantity by 
 distances set off at right angles to each other, on convenient scales. Thus, 
 by the distance of a point above a line we measure the pressure, and by 
 the distance of a point to the right of a line, we measure the volume. For 
 a different state of volume and pressure we have a different point; for 
 states of volume and pressure, differing little, we have points near to- 
 gether, and for changing states of volume and pressure we have a moving 
 point, which may be considered to trace a line. When a series of changes 
 brings the gas back to its original state, the line traced will return into 
 itself. 
 
 The product of a force into a distance through which it moves is 
 known as energy exerted, and it is equal to the product of the resistance 
 
QUANTITY OF STKAM USED, ETC. 61 
 
 by the distance through which it is moved; or, what is known as the work 
 done; and, as in an engine cylinder, the moving force on the piston is the 
 pressure times the area of the piston, and, as for a small movement the 
 pressure remains nearly the same, the energy expended is the product of 
 the mean pressure, by the piston area, by the distance moved; or the 
 pressure times the change in volume as the product of the piston area by 
 the distance moved by the piston is the change in volume occupied. The 
 number of cylinders in which such changes of volume take place is 
 immaterial, as the total energy expended must be the sum of the ener- 
 gies expended in all such cylinders; hence, the statement, that the 
 energy expended is the product of the mean pressure by the change in 
 volume. Hence, also, the diagram, just explained, furnishes the data 
 for finding the energy expended, as the latter must be represented by 
 the area between the horizontal line of "no pressure," the two verti- 
 cals at the end of the change of volume, and the line traced by the moving 
 point. 
 
 In any series of changes in which the original state of volume and 
 pressure is again reached, the energy effectively exerted is, of course, the 
 difference between that expended on the piston and that exerted by the 
 piston; or, that which is represented by the enclosed figure. 
 
 The first thing to be determined is the curve of expansion, or the 
 curve which would be drawn by an indicator attached to such an expand- 
 ing engine with a non-conducting cylinder; but as no such cylinders ex- 
 ist, we have to approximate the curve from the table of the "Properties of 
 Steam." 
 
 If we set out with the volume of one pound weight of steam at a given 
 pressure, and then assume it to expand by any defined law until its pres- 
 sure has been lowered to a given amount, we can compute the external 
 work done in changing the volume; then taking this heat from the total 
 internal heat contained in the steam, if we find that the change in inter- 
 nal heat is less than that required to do the external work, as by assump- 
 tion we find it in a non-conducting cylinder, some of the heat must be 
 supplied by the condensation of the steam itself, and that, therefore, the 
 volume must be reduced, and with it the external work, and, of course, 
 also the heat required to do this external work. By a little care in ap- 
 proximating, we shall arrive at such a condensation that the change in in- 
 ternal heat added to the heat given by condensation will just balance the 
 heat absorbed by the external work of expansion. Thus, a point on the 
 expansion curve for the indicator diagram of steam expanding in a non- 
 conducting cylinder is obtained. 
 
 The assumed law of expansion may be anything, but perhaps the best 
 for the purpose is that given by the volumes and pressures of saturated 
 steam, or, what is known as the "steam line." A common assumption is 
 that, the product of the pressures and volumes is constant, though there 
 is no basis for such an assumption. 
 
 The computation of the work done by the expansion can be perform- 
 ed in various ways, but for our purposes we shall plot the steam line and 
 
62 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 compare the area of the diagrams by any of the methods or instruments 
 used for measuring indicator cards. 
 
 8 10 
 
 NUMBER OF EXPANSIONS. 
 FIG. 39. 
 
 Such a steam line or expansion curve for dry steam having been drawn, 
 see Fig. 39, and the areas measured: first, of the rectangle o a b 1 and next, 
 for the area under the curve, we find the quotient of the area under the 
 curve divided by the rectangle area for different expansions to be as fol- 
 lows: 
 
 
 Area under 
 
 Area under 
 curve. 
 
 
 curve. 
 
 Rectangle. 
 
 1 . . . 
 
 
 o 
 
 2 
 
 be 2 
 
 68 
 
 3 
 
 b d 3 
 
 1 12 
 
 4 
 
 6 e 4 
 
 1 33 
 
 5... 
 
 6/5 
 
 1 53- 
 
 8 
 
 I) n 8 
 
 1.97 
 
 
 
 b h 10 
 
 2.14 
 
 .... 
 
 & i 20 
 
 2.73 
 
 
 
 
 Now, assuming 85 pounds as an average steam pressure, or, 15 + 85 = 
 100 pounds above zero, we have for the external work of evaporation, 81 
 units, and for the total internal work 1,100 units. Now, taking the rect- 
 angle o a b 1 in the figure as representing the work of 81 units, it is clear 
 that the work of expansion, in units of heat, will be obtained by multiply- 
 
QUANTITY OF STEAM USED, ETC. 
 
 63 
 
 ing the 81 units by the number given in the quotients above, which will 
 give values as follows: 
 
 Heat units in 
 work of ex- 
 pansion. 
 
 
 55 
 91 
 
 108 
 124 
 160 
 173 
 221 
 
 The terminal pressures are found by division and subtraction : 
 85 + 15 _ 15 
 
 r 
 
 The corresponding internal heats for dry steam are given in the fol- 
 lowing table: 
 
 ; 
 
 Terminal 
 pressure. 
 
 Internal 
 heat. 
 
 Heat units in 
 work of 
 expansion . 
 
 Heat units left af- 
 ter deducting work 
 of expansion from 
 1,100 units. 
 
 1 
 
 85 
 
 1,100 
 
 
 
 1,100 
 
 2 
 
 33 
 
 1,090 
 
 55 
 
 1 045 
 
 3 
 
 16 
 
 1,083 
 
 91 
 
 1,009 
 
 4 
 
 8 
 
 1 079 
 
 108 
 
 992 
 
 5 
 
 3 
 
 1, 7 9 
 
 124 
 
 976 
 
 8 
 
 I 
 
 1,071 
 
 160 
 
 940 
 
 10 
 
 6 
 
 1,069 
 
 173 
 
 927 
 
 20 
 
 11 
 
 1,060 
 
 221 
 
 879 
 
 
 
 
 
 
 Now the internal heat at the end of expansion is less than that given 
 for dry steam at the same pressure; therefore, a part of the heat for expan- 
 sion must have been furnished by the condensation of a portion of the 
 steam. But such condensation reduces the work of expansion itself and 
 hence the condensation supplies heat again and so on. The correct values 
 for the curve in a non-conducting cylinder are approximated as follows: 
 
 r 
 
 Heat in dry 
 steam. 
 
 Heat at end of 
 expansion. 
 
 Excess re- 
 quired from 
 condensation. 
 
 Change in 
 Units. 
 
 Change in 
 Excess. 
 
 1 
 
 1,100 
 
 1,100 
 
 
 
 o 
 
 
 2 
 
 3 
 
 1,089 
 1 084 
 
 1,045 
 1 009 
 
 44 
 75 
 
 2 
 
 42 
 
 70 
 
 4 
 
 1 080 
 
 992 
 
 88 
 
 g 
 
 79 
 
 5 
 
 1,079 
 
 976 
 
 103 
 
 12 
 
 91 
 
 ,M 
 
 1 071 
 
 MO 
 
 131 
 
 20 
 
 111 
 
 10 
 
 1,069 
 
 927 
 
 142 
 
 23 
 
 119 
 
 20 
 
 1 060 
 
 879 
 
 181 
 
 38 
 
 143 
 
 
 
 
 
 
 
64 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 But these values are now in error, for the one is assumed on the basis 
 of more external work than has been done, and the other on less. The 
 true value lies between them, and the assumption that the true value 
 divides the difference in the proportion of the values found gives us a 
 close approximation. 
 
 T 
 
 Per cent, 
 of 
 change. 
 
 Change in 
 units. 
 
 Excess required from 
 condensation. 
 
 Work of 
 expansion . 
 
 In units. 
 
 Per cent. 
 
 Gain. 
 
 I 
 
 
 
 
 
 
 2 , 
 
 95 
 91 
 90 
 
 88 
 85 
 84 
 79 
 
 2 
 6 
 8 
 11 
 17 
 19 
 30 
 
 44 
 
 78 
 87 
 102 
 127 
 139 
 182 
 
 4 
 
 J 
 
 o 
 
 9 
 12 
 12 
 17 
 
 0.67 
 1.11 
 1.31 
 1.50 
 1.90 
 2.05 
 2.55 
 
 3 
 
 4 
 
 5 
 
 8 
 
 10 ..... 
 
 20 
 
 
 From the curve already drawn for Dry steam, we draw the new curve 
 by changing the points for volume, reducing it by deducting the per- 
 centage for condensation. The points so found are marked on Fig. 39, 
 with the same letter accentuated. A comparison of the areas under 
 the new curve, between the same volumes, will give the last column of 
 figures. 
 
 A similar curve drawn for steam at 25 pounds pressure, does not differ 
 in the second place of decimals, and for steam at 185 pounds it only differs 
 by 1 in the second decimals for ten expansions. The approximation made 
 is within one heat unit and we may therefore use these results hereafter 
 with confidence, remembering that we must obtain the last column for 
 corresponding values of r for volumes not pressures. 
 
 The curve of expansion of steam in a non conducting cylinder has 
 been called by Prof. Kankine, an "Adiabatic" curve, and by Prof. Clausius 
 an "Isentropic" curve, and these terms may be met with in all works on 
 steam, frequently adding to the difficulties of the student. They will, 
 therefore, not be used, nor referred to, again in this work. 
 
 The amount of steam required to do the work in a non-conducting 
 cylinder is found by dividing the amount per total horse power per hour 
 for a non-expanding cylinder by 1, plus the ratio given in the last column 
 of the preceding table, or, the ratio of the total work done compared with 
 that done at boiler pressure. 
 
 From a large diagram constructed by the process described we have 
 tabulated the following values: 
 
'jl'.\ XT1TY OF STEAM USED, ETC. 
 
 65 
 
 PROPERTIES OF THE CURVE OF EXPANSION IN A NON-CONDUCTING 
 
 CYLINDER. 
 
 No. of Expansions. 
 
 Ratio of Total Area to 
 Area of Rectangle. 
 
 Ratio of Mean Total Ratio of Total Ter- 
 to Initial Total Pres- minal to Initial 
 sure. Pressure. 
 
 1 
 
 .1 1.090 
 
 0.991 
 
 0.90 
 
 .2 1.180 
 
 .983 
 
 .82 
 
 .3 1.261 
 
 .970 
 
 .75 
 
 M 1.333 
 
 .952 
 
 .69 
 
 .5 
 
 1.396 
 
 .931 
 
 .64 
 
 .6 
 
 1.459 
 
 .912 
 
 .59 
 
 .8 
 
 1.576 
 
 .866 
 
 .52 
 
 .0 
 
 1.666 
 
 .833 
 
 .46 
 
 .5 
 
 1.873 
 
 .749 
 
 .36 , 
 
 .0 
 
 2.035 
 
 .678 
 
 .30 
 
 .0 
 
 2.278 
 
 .568 
 
 .21 
 
 .0 
 
 2.476 
 
 .493 
 
 .17 
 
 .0 
 
 2.629 
 
 .436 
 
 .14 
 
 7.0 
 
 2.746 
 
 .393 
 
 .12 
 
 8.0 
 
 2.854 
 
 .357 
 
 .10 
 
 9.0 2.953 
 
 .328. 
 
 .087 
 
 10.0 3.034 
 
 .303 
 
 .077 
 
 11.0 3.106 
 
 .282 
 
 .070 
 
 12.0 3.169 
 
 .264 
 
 .063 
 
 13.0 3.232 
 
 .248 
 
 058 
 
 U.O 3.286 
 
 .235 
 
 .053 
 
 16.0 3.385 
 
 .211 
 
 .046 
 
 20.0 3.547 
 
 ,177 
 
 .036 
 
 25.0 3.709 
 
 .148 
 
 .028 
 
 30.0 3.835 
 
 .128 
 
 .023 
 
 From the preceding table, and the table of external work of steam, 
 the following table is computed. The first line, or steam used for one 
 expansion, or full stroke, is obtained, as already explained, by dividing 
 2,565, the number of units of heat for a horse-power per hour, by the 
 number of units of heat for external work; and the other quantities by 
 dividing these numbers by the gain by expansion, or ratio of total area of 
 figure to area of rectangle. This has also been explained. From the ratio 
 of total mean pressure to initial pressure, the three next tables are com- 
 puted, by assuming that a back pressure of 4 pounds per square inch 
 exists in the condensing, and 16 pounds, or 1.3 pounds above the atmos- 
 phere, in the non-condensing engine. In each case the mean total pres- 
 sure is found and the back pressure deducted. The quantity of water per 
 total horse-power is then multiplied by the ratio of the mean total to the 
 mean indicated pressure, or the pressure left to act on the piston after 
 deducting the back pressure, 4 or 16 pounds respectively, to obtain the 
 consumption of water per indicated horse-power per hour in a non- 
 conducting cylinder. 
 
66 
 
 STEAM L T SING: OR, STEAM K \G1XK PRACTICE. 
 
 COST PER HOUR PER TOTAL HORSE-POWER, IN POUNDS, IN DRY STEAM IF 
 WORKED IN NON-CONDUCTING CYLINDERS. 
 
 BOILER PRESSURE IN POUNDS PER SQUARE INCH. 
 
 No. of 
 
 Expan- 
 
 
 
 
 
 
 
 sions. Atmos. 
 
 
 
 
 
 
 
 
 
 20 
 
 40 60 
 
 80 
 
 100 : 120 140 j 180 220 
 
 1 
 
 35.7 
 
 33.7 
 
 32.9 32.5 
 
 32.1 
 
 31.7 31.3 
 
 30.9 110.5 30.2 
 
 1.1 32.7 
 
 31.0 
 
 30.9 29.8 
 
 29.4 
 
 29.0 28.7 
 
 28.3 , 28.0 27.7 
 
 1.2 
 
 30.2 
 
 28.6 
 
 27.9 27.5 
 
 27.2 
 
 26.8 
 
 26.5 
 
 26.2 
 
 25.9 
 
 25.5 
 
 .3 
 
 27. G 
 
 26.2 25.5 25.2 
 
 24.8 
 
 24.5 
 
 24.2 
 
 23.9 
 
 23.7 
 
 23.4 
 
 .4 
 
 26.1 
 
 24.7 
 
 24.1 23.8 
 
 23.5 
 
 23.2 
 
 22.9 
 
 22.6 
 
 22.4 
 
 22.1 
 
 .5 
 
 24.9 
 
 23.6 
 
 23.0 
 
 22.7 
 
 22.4 
 
 22.2 
 
 21.9 
 
 21.6 
 
 21.4 
 
 21.1 
 
 .6 
 
 24.4 
 
 23.1 
 
 22.5 22.2 
 
 22.0 
 
 21.6 
 
 21.4 i 21.2 
 
 20.9 
 
 20.7 
 
 .8 
 
 22.9 
 
 21.7 
 
 22.1 
 
 20.8 
 
 20.6 
 
 20.3 
 
 20.1 19.8 
 
 19.6 
 
 19.4 
 
 2.0 
 
 21.4 
 
 20.3 
 
 19.7 
 
 19.5 
 
 19.2 
 
 19.0 
 
 18.8 
 
 18.5 
 
 18.3 
 
 18.1 
 
 2.5 
 
 19.0 
 
 18 
 
 17.6 17.3 
 
 17.0 
 
 16.9 
 
 16.7 16.5 
 
 16.3 
 
 16.1 
 
 3.0 
 
 17-5 
 
 16.6 
 
 16.2 
 
 16.0 
 
 15.8 
 
 15.6 
 
 15.4 
 
 15.2 
 
 15.0 
 
 14.8 
 
 4.0 
 
 15.7 
 
 14.9 
 
 14.5 
 
 14.3 
 
 14.1 ! 13.9 13.8 
 
 13.6 
 
 13.4 13.3 
 
 5.0 
 
 14.4 
 
 13.6 
 
 13.3 
 
 13.1 13.0 i 12.8 12.6 
 
 12.5 
 
 12.3 i 12.2 
 
 6.0 
 
 13.6 
 
 12.9 12.5 
 
 12.4 i 12 2 12.1 11.9 
 
 11.8 
 
 11.6 ! 11.5 
 
 7.0 
 
 13.0 
 
 12.3 
 
 11.9 
 
 11.8 ! 11.7 11.5 11.4 
 
 11.2 11.1 11.0 
 
 8.0 
 
 12.5 
 
 11.8 
 
 11.5 
 
 11.4 : 11.2 i 11.1 ; 11.0 
 
 10.8 10.7 10.6 
 
 9.0 
 
 12.0 
 
 11.4 
 
 11.1 
 
 11.0 i 10.8 
 
 10.7 
 
 10.6 
 
 10.4 
 
 10.3 10.2 
 
 10.0 
 
 11.7 
 
 11.1 
 
 10.8 
 
 10.7 10.6 10.4 
 
 10.3 
 
 10.2 
 
 10.1 10.0 
 
 11.0 
 
 11.6 
 
 10.9 
 
 10.6 
 
 10.4 i 10.3 10.2 
 
 10.1 
 
 10.0 
 
 9.8 9.7 
 
 12.0 
 
 11.3 
 
 10.7 
 
 10.4 
 
 10.3 10.1 
 
 10.0 
 
 9.9 
 
 9.8 
 
 9.6 
 
 9.5 
 
 13.0 
 
 11.1 
 
 10.5 
 
 10.2 
 
 10.1 i 9.9 
 
 9.8 
 
 9.7 
 
 9 6 9.5 9.4 
 
 14.0 
 
 10.8 
 
 10.3 
 
 10.0 
 
 9.9 
 
 9.8 
 
 9.6 
 
 9.5 
 
 9,4 
 
 9.3 9.2 
 
 16.0 
 
 10.5 
 
 10.0 
 
 9.7 
 
 9.6 
 
 9.5 9.4 
 
 9.3 
 
 9.1 
 
 9.0 8.9 
 
 20.0 
 
 10.0 
 
 9.5 
 
 9.3 
 
 9.2 
 
 9.0 | 8.9 
 
 8.8 
 
 8.7 
 
 8.6 
 
 8.5 
 
 25.0 
 
 9.6 
 
 9.1 
 
 8.9 
 
 8.7 
 
 8.6 8.5 
 
 8.4 
 
 8.3 
 
 8.2 
 
 8.1 
 
 30.0 9.1 
 
 8.8 
 
 8.6 
 
 8.5 
 
 8.4 8.3 
 
 8.2 
 
 8.1 8.0 
 
 7.9 
 
 MEAN TOTAL PRESSURE IN POUNDS PER SQUARE INCH IN NON-CONDUCTING 
 
 CYLINDERS. 
 
 No. of 
 
 Expan 
 
 sions. 
 
 INITIAL PRESSURE ABOVE ATMOSPHERE. 
 
 4d 60 80 100 120 i 141 
 
 220 
 
 .1 
 
 14.9 
 
 34.7 
 
 54.5 
 
 74.3 
 
 94.1 
 
 114 
 
 134 
 
 154 
 
 193 
 
 233 
 
 .2 
 
 14.7 
 
 34.4 
 
 54.1 
 
 73.7 
 
 93.4 
 
 113 
 
 133 
 
 152 
 
 192 
 
 231 
 
 .3 
 
 14.5 
 
 33.9 
 
 53.4 
 
 72.8 
 
 92.1 
 
 112 
 
 131 
 
 150 
 
 189 
 
 228 
 
 .4 
 
 14.2 
 
 33.3 
 
 52.4 
 
 71.6 
 
 9H.5 
 
 109 
 
 129 
 
 148 
 
 186 
 
 224 
 
 .5 
 
 14.0 
 
 32.6 
 
 51.2 
 
 69.7 
 
 88.4 
 
 107 
 
 126 
 
 144 
 
 181 
 
 219 
 
 1.6 
 
 13.7 
 
 31.9 
 
 50.2 
 
 C8.4 
 
 86.6 
 
 105 
 
 123 
 
 141 
 
 178 
 
 214 
 
 1.8 
 
 13.0 
 
 30.3 
 
 47.6 
 
 94.9 
 
 82.2 
 
 99.5 
 
 117 
 
 134 
 
 169 
 
 203 
 
 2.0 
 
 12.5 
 
 29.2 
 
 45.8 
 
 63.9 
 
 79.1 
 
 95.8 
 
 112 
 
 126 
 
 162 
 
 196 
 
 2.5 
 
 11.2 
 
 26.2 
 
 41.2 
 
 56.2 
 
 71.2 
 
 86.2 
 
 101 
 
 116 
 
 146 
 
 176 
 
 3.0 
 
 10.2 
 
 23.7 
 
 37.3 
 
 50.9 
 
 64.4 
 
 78.0 
 
 91-6 
 
 105 
 
 132 
 
 159 
 
 4.0 
 
 8.5 
 
 19.9 
 
 31.2 
 
 42.6 
 
 54.0 
 
 65-3 
 
 76.7 
 
 88.0 
 
 111 
 
 133 
 
 5.0 
 
 7.4 
 
 17.3 
 
 27.2 
 
 37.1 
 
 47.0 
 
 56.9 
 
 66.8 
 
 76.7 
 
 96.5 
 
 116 
 
 6.0 
 
 6.6 
 
 15.3 
 
 24.0 
 
 32.7 
 
 41.5 
 
 50-2 
 
 58.9 
 
 67.7 
 
 85.1 
 
 103 
 
 7.0 
 
 5.9 
 
 13.7 
 
 21.6 
 
 30,1 
 
 37.3 
 
 46.2 
 
 52.9 
 
 60.8 
 
 76.5 
 
 92.2 
 
 8.0 , 
 
 5.3 
 
 12.5 
 
 19.6 
 
 26.7 
 
 33.9 
 
 41.0 , 
 
 48.1 
 
 55.3 
 
 69.5 
 
 83 8 
 
 9.0 
 
 4.9 
 
 11.5 
 
 18.0 
 
 24.6 
 
 31.2 
 
 37.7 
 
 44.3 
 
 50.9 
 
 64.0 
 
 77.1 
 
 10.0 
 
 4.6 
 
 10.6 
 
 16.7 
 
 22.8 
 
 28.8 
 
 34.9 
 
 41.0 
 
 47.0 
 
 59.2 
 
 71.3 
 
 11.0 
 
 4.2 
 
 10.1 
 
 15.5 
 
 21.2 
 
 26.8 
 
 32.5 
 
 38.1 
 
 43.8 
 
 55.0 
 
 66.4 
 
 12.0 
 
 4.0 
 
 9.2 
 
 14.5 
 
 19.8 
 
 25.0 
 
 30.3 
 
 35.6 
 
 41.8 
 
 51.0 
 
 62.0 
 
 13.0 
 
 3.7 
 
 8.7 
 
 13.6 
 
 18.6 
 
 23.7 
 
 29.2 
 
 33.5 
 
 38.4 
 
 48.4 
 
 58.3 
 
 14.0 
 
 3.5 
 
 8.2 
 
 12.9 
 
 17.6 
 
 22.3 
 
 27.0 
 
 31.7 
 
 36.4 
 
 45.8 
 
 55,2 
 
 16.0 
 
 3.2 
 
 7.4 
 
 11.6 
 
 15.9 
 
 20.1 
 
 24.3 
 
 28.6 
 
 32.8 
 
 41.2 
 
 49.7 
 
 20.0 
 
 2.7 
 
 6.2 
 
 9.8 
 
 13.3 
 
 16.8 
 
 20.4 
 
 23.9 
 
 27.5 
 
 34.5 
 
 41.7 
 
 25.11 
 
 2.2 
 
 5.2 
 
 8 2 
 
 11.1 
 
 14.1 
 
 17.1 
 
 20.0 
 
 23. U 
 
 28.9 
 
 34.9 
 
 30.0 
 
 1.9 
 
 4.5 
 
 7_o 
 
 y.o 
 
 12.1 
 
 14.7 
 
 17.3 
 
 19.8 
 
 24.9 
 
 30.0 
 
QUANTITY OF STEAM USED, ETC. 
 
 67 
 
 MEAN EFFECTIVE PRESSURE IN CONDENSING ENGINES, NON-CONDUCTING 
 
 CYLINDERS. 
 
 No. of 
 Expan- 
 sions. 
 
 INITIAL PBESSUBE ABOVE ATMOSPHEBE, IN POUNDS 
 
 PEE SQUABE INCH. 
 
 
 
 20 
 
 i 
 
 i 
 
 60 80 
 
 100 
 
 120 
 
 140 
 
 180 220 
 
 I 
 
 1.1 
 1.2 
 1.3 
 1.4 
 1.5 
 1.6 
 1.8 
 2.0 
 2 5 
 3.0 
 4.0 
 5.0 
 6.0 
 7.0 
 8.0 
 9.0 
 10.0 
 11.0 
 12.0 
 13.0 
 14.0 
 16.0 
 20 
 25.0 
 30.0 
 
 10.9 
 10.7 
 10.5 
 
 10.2 
 10.0 
 
 9.7 
 9.0 
 8.5 
 7.2 
 6 2 
 4.5 
 3.4 
 2.6 
 1.9 
 1.3 
 0.9 
 0.6 
 0.2 
 
 
 30.7 
 30.4 
 29.9 
 29.3 
 28.6 
 27.9 
 26 3 
 25.2 
 22.2 
 19.7 
 15.9 
 13.3 
 11.3 
 9.7 
 8.5 
 7.5 
 6.6 
 6-1 
 5.2 
 4.7 
 4.2 
 3.4 
 2.2 
 1.2 
 0.5 
 
 50.5 
 50.1 
 49.4 
 48.4 
 ; 47.2 
 46.2 
 43.6 
 41.8 
 37.2 
 33.3 
 27.2 
 23.2 
 20.0 
 17.6 
 15.6 
 14.0 
 12.7 
 11.5 
 10.5 
 9.6 
 8.9 
 7.6 
 5.8 
 4.2 
 3.0 
 
 70.3 
 69.7 
 68.8 
 67.6 
 65.7 
 64.4 
 60.9 
 59.9 
 52.2 
 46.9 
 38.6 
 33.1 
 28.7 
 26.1 
 22.7 
 20.6 
 18.8 
 17.2 
 15.8 
 14.6 
 13.6 
 11.9 
 9.3 
 7.1 
 5.6 
 
 90.1 
 89.4 
 88.1 
 86.5 
 84.4 
 82.6 
 78.2 
 75.1 
 67.2 
 60.4 
 50.0 
 43.0 
 37.5 
 33.3 
 29.9 
 27.2 
 24.8 
 22.8 
 21.0 
 19.7 
 18.3 
 16.1 
 12.8 
 10.1 
 8.1 
 
 110 
 109 
 108 
 105 
 103 
 101 
 95.5 
 91.8 
 82.2 
 74.0 
 61.3 
 52.9 
 46.2 
 42.2 
 37.0 
 33.7 
 30.9 
 28.5 
 26.3 
 25.2 
 23.0 
 20.3 
 16.4 
 13.1 
 lo.7 
 
 130 
 129 
 127 
 125 
 122 
 119 
 113 
 108 
 97 
 87.6 
 72.7 
 62.8 
 54.9 
 48.9 
 44.1 
 40.3 
 37.0 
 34.1 
 31.6 
 29.5 
 27.7 
 24.6 
 19.9 
 16.0 
 13.3 
 
 150 
 148 
 146 
 144 
 140 
 137 
 130 
 122 
 112 
 101 
 84.0 
 72.7 
 63.7 
 56.8 
 51.3 
 46.9 
 43.0 
 39.8 
 37.8 
 34.4 
 32.4 
 28.8 
 23.5 
 19.0 
 15.8 
 
 189 
 188 
 185 
 182 
 177 
 174 
 165 
 ; 158 
 142 
 128 
 107 
 I 92.5 
 81.1 
 72.5 
 65.5 
 64.0 
 55.2 
 51.0 
 47.0 
 44.4 
 41.8 
 37.2 
 30.5 
 24.9 
 20.9 
 
 m 
 
 227 
 224 
 220 
 215 
 210 
 199 
 192 
 172 
 155 
 129 
 112 
 99 
 88-2 
 79.8 
 73.1 
 67.3 
 62.4 
 58.0 
 54.3 
 51.2 
 45.7 
 37.7 
 30.9 
 26.0 
 
 
 
 
 
 
 
 MEAN EFFECTIVE PRESSURE, IN POUNDS PER SQUARE INCH, IN NON- 
 CONDENSING ENGINES, NON-CONDUCTING CYLINDERS. 
 
 No. of 
 Expan 
 sions. 
 
 INITIAL PRESSURE ABOVE ATMOSPHEBE IN 
 
 POUNDS 
 
 PEB SQUABE INCH. 
 
 
 
 20 
 
 40 
 
 60 
 
 80 
 
 100 
 
 120 
 
 140 
 
 180 220 
 
 .1 
 .2 
 
 .4 
 .5 
 1.6 
 1.8 
 2.0 
 2.5 
 3.0 
 4.0 
 5.0 
 6.0 
 7 
 8.0 
 9.0 
 10.0 
 11 
 12.0 
 13.0 
 14.0 
 16.0 
 20.0 
 25.0 
 30.0 
 
 16.7 
 16.4 
 15.9 
 15.3 
 14.6 
 13.9 
 12.3 
 11.2 
 8.2 
 5.7 
 1.9 
 
 36.5 
 36.1 
 35.4 
 34.4 
 33.2 
 32.2 
 29.6 
 27.8 
 23.2 
 19.3 
 13.2 
 9.2 
 6.0 
 3.6 
 1.6 
 
 : 
 
 56.3 
 55.7 
 54.8 
 53.6' 
 51.7 
 50.4 
 46.9 
 45.9 
 38.2 
 32.9 
 24.6 
 19.1 
 14.7 
 12.1 
 8.7 
 6.6 
 4.8 
 3.2 
 1.8 
 0.6 
 
 76.1 
 75.4 
 74.1 
 72.5 
 70.4 
 68.6 
 64.6 
 61.1 
 53.2 
 46.4 
 36.0 
 29.0 
 23.5 
 19.3 
 15.9 
 13.2 
 10.8 
 8.8 
 7.0 
 5.7 
 4.3 
 2.1 
 
 96 
 95 
 94 
 91 
 89 
 87 
 81.5 
 77.8 
 68.2 
 60.0 
 47.3 
 38.9 
 32.2 
 28.2 
 23.0 
 19.7 
 16.9 
 14.5 
 12.3 
 11.2 
 9.0 
 6.3 
 2.4 
 
 116 
 115 
 113 
 111 
 108 
 105 
 99 
 
 73 6 
 58-7 
 48.8 
 40.9 i 
 34.9 
 30.1 
 26.3 
 23.0 
 
 15.5 \ 
 13.7 
 10.6 
 5.9 ' 
 2.0 
 
 136 
 134 
 132 
 130 
 126 
 123 
 116 . 
 108 
 98 
 87 
 70.0 
 5S.7 
 49.7 
 42.8 
 37.3 
 32.9 
 19.0 
 25.8 
 23.8 
 20.4 
 18.4 
 14.8 
 9.5 
 5.0 
 1.8 
 
 175 
 174 
 171 
 168 
 163 
 160 
 151 
 144 
 128 
 114 
 93 
 78.5 
 67.1 
 58.5 
 51.5 
 46.0 
 41.2 
 27.0 
 33.0 
 30.4 
 27.8 
 23.2 
 16.5 
 10.9 
 6.9 
 
 215 
 213 
 210 
 206 
 201 
 196 
 185 
 178 
 158 
 141 
 115 
 98 
 85 
 74.2 
 65.8 
 59.1 
 53.3 
 48.4 
 44.0 
 40.3 
 37.2 
 31.7 
 23.7 
 16.9 
 12.0 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
68 STEAM r.s'/AV;,- OH, STEAK ENGINE PRACTICE. 
 
 DATA FURNISHED BY EXPERIMENT. 
 
 In comparing the accompanying tables with the performance of actual 
 engines the back pressure may be found to vary from the 4 or 16 pounds 
 per square inch mentioned, so that the consumption of water is only 
 tabulated per total horse-power at present, and the comparisons made on 
 this basis are not affected by the back pressure. 
 
 In comparing experiments made upon the performance of actual en- 
 gines, the fact must not be forgotten that the value of the results depend 
 on the data which have been used and the skill of the experimenters; 
 hence, it will differ. The most valuable data are those in which both the 
 heat received and the heat rejected by the engine have been measured. 
 These require measurements, preferably by weight, of the feed water fur- 
 nished to the boiler, the pressure and temperature of evaporation, and 
 the dryness of the steam near the engine; the work done in the cylinder, 
 the quantity of injection water and its rise in temperature; the difference 
 between the heat delivered added to the work done by the engine; while 
 the heat received furnishes an important check. To appreciate the value 
 of this check one should examine some of the first experiments in which 
 this measurement was attempted, and which may be found in the Bulletin 
 de Societe' Industrielle de Mulhouse for 1857; and the record of Hirn's ex- 
 periments show the difficulties he overcame. 
 
 First, reliable experimental data can only be obtained from Stationary 
 Condensing Engines, on account of the impossibility of measuring con- 
 densing water: the difficulty of measuring the feed water precludes the 
 use of marine engines for such a purpose. 
 
 Next in point of value are experiments where the heat received and 
 delivered are carefully ascertained, either by measurements of the feed and 
 priming, or the water of condensation, and its rise in temperature. The 
 latter is the easiest measurement to be taken, but, for the most reliable 
 data is restricted to stationary condensing engines. Measurements of feed 
 water and priming can be made in all classes of engines, with the excep- 
 tion of large marine engines at sea, where the difficulty of getting at the 
 quantity of feed water has not yet been overcome, though it perhaps might 
 be by the use of a water meter. 
 
 Third in point of value are long- continued experiments made on en- 
 gines in which the feed water only is noted, but in which the boilers are so 
 large that the priming may be neglected. Such are most of the experi- 
 ments made by the United States Naval authorities. 
 
 Next to the latter in point of scientific value, but first in practical in- 
 terest, are the records of the performances of the large ocean steamships 
 as to fuel used and power developed in long voyages; and again, the records 
 of the duty of pumping engines. In such records it is impossible to sep- 
 arate the performances of engine and boiler, but the results are compre- 
 hensive and of great value. 
 
 Last in point of value are short experiments in which the fuel or the 
 feed-water is measured, in some indirect manner, and the engine "indi- 
 cated" only, as, of course it must be. 
 
We give, in the following table, some engine trials made by various 
 authorities, classing them under the three first summaries of value given 
 above, as A, B and C. This table might be greatly extended, but only by 
 the admission of experiments which are either isolated, improbable, or 
 deficient in the required data. 
 
 A very casual examination of the table shows us that economy of 
 steam is promoted by high pressure and high speed, while the cost in 
 steam of a total horse-power is lessened by large expansion, and the cost 
 of a net horse-power is least with moderate expansion. 
 
 Another fact brought prominently to the front is that the actual use of 
 steam is very far in excess of that given by our tables for a non-conduct- 
 ing cylinder. This excess is due to several causes: 
 
 1. To the use of wet steam. 
 
 2. To the loss by clearance space. 
 
 3. To external radiation and loss of heat. 
 
 4. To internal radiation and the transfer of heat between the iron of 
 the cylinder and its contained steam. 
 
 For our purpose we shall not individually consider these four causes 
 of loss, but we may take up our table of experimental data and compute 
 from the tables already given the amount of steam that would be used by 
 a non-conducting cylinder, working with steam of the same initial 
 pressure and expansion. Deducting this quantity from the quantity 
 actually used we will examine the excess to ascertain what law, if any, can 
 be found to account for it. 
 
70 
 
 STEAM USING; Off, STEAM ENGINE PRACTICE. 
 
 
 I 
 
 S * 
 
 S 
 
 o 
 
 S 
 
 
 
 
 > rt xj- 
 
 S. 
 
 OQ" " 
 
 
 
 50 2 S O 
 
 <j o i m 
 
 CO 
 
 
 
 
 <J W 
 as- s 2 5 ! " 3 " *" " " " " " " 
 
 
 
 
 1 
 
 r5 ^ 
 
 O O 
 
 5 
 
 
 1 
 
 iii i 1 
 
 III - s 1 !' i . 1 . 
 
 1 
 
 g 
 
 H 
 
 
 o EB* - "S * rs ** " "2" " 
 c a oj k,)^ 2 a r9 TJ o 
 
 g^ ^ o g- 
 
 PQQQD >- co O O 00 
 
 "5" 
 
 fi 
 
 g -MMOJ 
 
 5 -9SJOH iaN d 
 
 
 g^^5 
 
 1 
 
 WAVOJ-9SJOH 
 
 SSSS8SS8S",H V^-T^T 
 
 o oi * o x t- c 
 
 w 
 
 TgrjQ T JO^ 
 
 t^XHHlOt^iOtfisO t^ <^> t^GiCCOii-lO 
 
 10 o as o 01 o <M 
 
 rH* 
 
 OS 
 
 
 .. ^.CSCOCSCC-. 
 
 53 
 
 w 
 
 J9M.OJ-9SJOH 
 pg^BOtpUI J9<J 
 
 Ss'sSSrt'Sri 3 S^' 3S2S^'^^'^ 
 
 OS X rH C^ to O rH 
 
 1 
 
 
 
 K> CO CO CO PS W ^ 
 
 uoisuBdxg jo 
 
 O IO C^- t^ CO 1C CC 
 -* 1-1 1C rH rH O rH W 
 
 i rH rH rH fi W O 00 
 
 SSgtggggg g ^S gs s : s s 3FSI " 
 
 CO <DO D(M_OOCCOO rHr-'^^Xia'-' 
 rH rH l-i rH rH rH rH rH rH CO 5 t- C) IO CS IM IO 55 1M CJ rH rH ^ 
 
 M t- OS <N rH 
 
 9HO.I1S 
 
 5 55 
 
 rH S^ICO X- 
 
 ij I 
 

 
 I 
 
 tri 
 
 Q I 'A A 77 T } ' n F N TE. 1 -V I '>' E 1>. E T< '. 
 
 71 
 
 (4- 
 
 od.5 
 
 :22 22 
 
 |= . s -. ,,,-. , ,..,.. : -.. , -.-.-. , .... -. ...... ;:-. -. , 
 
 O ___ 
 
 II 1 
 I \\^\ a 
 
 "~* -JJ ^ *3 fl - ~ _j ~ _3 -2 
 
 "S C fl fl S 5 9" * 
 
 JJ S ,iS > Jd O O O O 
 
 ^2:22 22 3,2 2 22 -O O 'O ofc 'g-'S "S* " " ' 'v* S ' ~ 
 
 P QQ ' ' O l- ^ >1 1^ 
 
 MCn-^; -H . Xl.1^-l i-IOTHr-l -i(5:C*e5!-l- t-O-*->J<t- CSOXXO SSC5 i-IC;'*C 
 
 ! S5 5 5! 2555 
 
 i c-. ~ x a x it- c os si =: c; ?! =; r; <M CMM t- cs '= r: - x ss o c o o t- -* 10 <* <N so * o o o t- o 
 
 ' ' ' 8S8R8 a 55 si 8* ssassa aasss asss^ 88?' ' 
 
 " gg 
 
 ^33,^i IggSSS 3S3S 32SiS iSiS gggSS SS5 ^3333 3^'g 
 
 132 2 2 2 2 2 SS 2 2222 2 2 2 2 2 2 22222 22222 22222 22222 222 
 ?,: 2 2 2 2 2 ^2 : 2 2 ^2 ; 2 ; 2 2 2 22222 22222 22222 22222 222 
 
 H2222 2'. 222 2222 222222 22222 22222 22222 22222 222 
 
 O 
 
 38S8S $3883 
 
72 
 
 STEAM rWAV/,- oii, 8TEAM ENGI&E PRAVTIUE. 
 
 
 i ! 1 
 
 A*-- r/i '*----- 
 
 l| 
 
 
 >** P 
 
 H^ 
 
 
 ' 0_ 
 
 o 6 
 
 .>>,.., . 
 
 n ENGINES. 
 
 > 
 
 I 
 
 1 g 
 3 s s - -8 s 8 * S s ; 
 
 1 1 
 
 -B I-S 
 
 g- = - - sgfelils 
 
 " d- r H 1 
 
 III 
 
 s 
 
 P5 MaAVOfT o o <o N e* IH cc x * t- co 
 
 oo o co oo e?> o CM <M co -* '. 
 
 g 
 
 -aSJtOH ^N Md S co S 3c3*c3 
 
 3S?38 S8SS8 : 
 
 t) ^ 
 
 
 
 w 1 
 3 s 
 
 
 OO5-^<MO O 5O 00 -it O SI 
 CO CN CO CO CO (M <N CN CN CN i-H 
 
 
 
 
 CO O 
 
 
 co co co co <M ^i n c^ CM CN 
 
 ^ 
 
 
 t o o <o ** OOCM^-^ 
 
 COt-COXO t-OOO10O?C LO 
 
 
 
 
 
 OCCOCOC-OOJ Wt-t-^CO'* 
 
 
 EH 
 
 -said ra^a^s I^Wini JoSS^SS ^SSS5 
 
 ----- ----- s 
 
 s 
 
 ja AA o j o*t-i-a5THt- ocoowcoio 
 
 J.0v^cl_ OOCOCOrHC^CO ^^OjJ5 
 
 -* Oi CO CN O X -H t- t- CO 1C 
 O CN t OO i < i (O5OOIO O 
 r-KMCOCN-* rSCOCS^^ t- 
 
 
 9<vn iHOt-t-COO OJSOIM IOCO 
 
 oicoo-^oi ooono -* 
 
 
 -UTH jad suo W nioA9a gggg^^S ^g^SSS 
 
 Iff O 00 t- OJ C O5 O O5 O5 i-l 
 
 
 00 
 Ol 
 
 
 
 
 
 M 
 
 
 
 S^ 
 
 oo 
 
 
 1* 
 
 
 
 1 t 
 
 1 s 
 
 i i 
 
 
 5 H 
 g 
 
 d 
 
 
 aoaaiajan ^oj -o X SSSSSgffegEigS 
 
 COOOXOCOO XXXXX Si 
 

 
 s 
 
 CD 
 OB 
 
 i o 
 
 S a 
 
 ^ - d 
 
 " a^ * -4 
 
 =3 23 - 
 
 "I 111 
 
 O S PQ 
 
 ! 
 
 ^ oc 
 
 t-t <U 
 
 1 1 * 
 if 
 
 S3 x 
 
 f 
 
 
 
 
 ^ 
 
 pq 
 
 
 
 3 
 
 o 
 
 O 
 
 
 cc 
 
 , 
 
 2 2 
 
 18. I 
 
 I &2 a 
 
 ss * as i *!^-S|L 
 g SSsSS:* 
 
 f gsfelsl 2 0c<0 3ci 
 - -s 
 
 I .1 1. 
 
 s" - i &" 
 
 4* i2 S 
 
 PQ K co 
 
 t " S J 2 ; 
 
 ^ 
 
 
 
 
 
 | 
 
 -^ " 
 
 
 SS^SS S?3S?5S S^22S 3?g8 
 
 
 H 
 
 rH 00 
 
 ss 
 
 t- r-l * -* 10 05 10 t- t- 
 
 e-j jo' * * * -* id -^ *' H! * 
 
 t-COJOCOrH H"~OrHr-l OOt-OOi r-ilOJOCO-*( 
 
 id id id -ii id id i-' id o o' ci ci oi jo' co" s o oo * id 
 
 rH rH r-l rH i-H rH rH rH i-H rH rH rH rH rH rH rH i-H rH rH rH 
 
 CN o oo co o oo 
 
 t-' IO OS rH r-I -H4 
 
 z 
 
 
 
 
 
 1 
 
 06 <N' 
 
 o' o 06 cc o" t^ t-1 o' o' o t-' 
 
 o e rn' 05" oc' oc'cdx'o'o' ooo't-'x c'^'^'xd 
 
 CO rH OC ' IQ rH 
 
 o 
 
 iH C 
 
 
 
 
 H 
 
 -1 = 
 
 CM i-l <M t- t- OO O C TH 
 
 
 
 r 
 
 O HI 
 
 e rn o t- .00 ooxojoooooo 
 
 
 00 CO CO CO <M <M 
 
 
 
 * t- rH 00 10 d t- 
 
 X t- O O t- O IO CO rH CO 
 
 t-lOrHOIOC5 
 
 
 !o 
 
 ototSS .; goTigiS 
 
 ?2ogo? ?ooio g||Sc2 33?^S5 
 
 ^S??Si 
 
 
 ot- 
 
 coocot-o ^05^H.oo 
 
 3SSS2 ^5!5;S SIS8SS 3S2SS 
 
 So?13 
 
 
 CMrH 
 
 H.H.H.HiHICO COCOCOCOCOCO 
 
 JOCNrHrHCS rH rH rH rH 
 
 
 
 ooio 
 
 
 
 05050 1000 
 
 
 
 333333 333333 
 
 88838 3333 SI2? iSr1Sa?5 
 
 IO t~ C^ O5 rH t 
 
 ^< os c-i oc os * 
 
 
 S: 
 
 , ::::S = : .. 
 
 22s s : s s s s ^ S^ s s : s : s s 3 
 
 33335: 
 
 
 a 
 
 1 s 
 
 S 
 
 5 
 -o -e 
 C;::: fl s ^ s s : : 
 
 C8 03 
 c5 ^ 
 
 ,, i ; 
 
 05 
 
 . I 
 
 1 1...... ....I... 
 
 88S8S sassa sssss sssss 
 
74 
 
 XTKAM 
 
 Xf,'; Oil, STEAM KXaiXK PRACTICE. 
 
 
 | 
 
 Is 1. 
 
 H oa -5 
 $ - o 
 
 s 
 
 * 5* * 
 
 CO J 
 
 Remarks. 
 
 ! 
 
 1 
 
 o o o' " o "S 
 
 1 ' S If 
 
 a s fl o 
 ! t-s r5 OO 
 
 
 o 6 
 
 00 
 
 3 
 
 pq<j pq jj 
 
 03 ,. 00- - 05- - - 
 
 
 
 5 
 
 o! 5 5 
 
 
 
 W 
 
 tO iH W t- * W O O O Cl 
 
 
 O3 
 
 g H 
 
 oicoinod t> t- t- 05 os 
 
 
 i 
 
 1 & 
 
 
 1 
 
 
 
 % *' 
 
 g w 
 
 Og.OOC, ^gggjg 
 
 s 
 
 s 
 
 
 p? 
 
 S a|| 
 
 SSS SSS3?! 
 
 p 
 
 g 'd'H'N *: <=. "*. t 1H . O<N^OOO 
 
 ^ lag 
 
 S8SSSSB gg^S 
 
 i 
 
 
 1 s ll 
 
 PM^ 
 
 
 -e 
 
 3 
 
 1 'd'H'i S52S52 w rt - cc M - 
 
 S - 
 S o^s 
 
 occ^SS 333 
 
 1 
 
 pq " - 
 &" 1 'd 'H 'I O * O o o <N o oo oo 
 
 1 II 
 
 005100505 t-r-JrHM 
 
 1 
 
 t rH,H.-l<N<NiH<N<NN 
 
 PH ! ^ 
 
 
 a 
 
 'UOISUBd os o > o IN Td o a <N co M 
 
 PM 
 
 
 5 
 
 H 
 
 " X 3 JO OI^'B^J ' c^'os'os'icwo o <o >o co 05" 05* 
 8Jn888J<J 
 
 
 O ^5 
 
 s^^iS iS^s 
 
 ENGINE 
 
 r "Tj *T *OkT i . .?^ 
 
 a . i "jv S e8 88 8 S eiooi-icoc; 
 
 ^ S 
 
 
 
 ^3 
 
 S.i 
 
 sssesssssa 
 
 
 a^^UTpjf J9d ift^eoiMcsift cor-it-iMrj^ 
 
 ^ a 
 
 3^S??S ftSc?5S 
 
 
 suoi(jn[OAaH: ^g^i^S S!&ggS 
 
 i 
 
 
 
 8 
 
 
 
 o 
 
 B 
 
 g s ^ g$ ! ><8 
 
 
 1 
 
 ^^ 
 
 a a l 
 
 gxxxxx ^ w^, 
 !8S5gg8 j$ v^ 
 
 
 . oc 
 
 g cc 
 
 S 
 
 M \^i e8 
 
 
 JO J9^9m < ei(T Ov" - - 22^:^2 
 
 flg 
 
 1^ 
 
 
 
 hj 
 
 c 
 
 o 
 
 1 
 
 
 1 1 
 
 1 1 
 
 3, 
 
 J 
 
 
 |3 
 
 <j W 
 
 
 
 s| 
 
 33305 ggggg 
 
 
 -aaSJ^x ! SSSS8S sllSS^i 
 
 I I i 
 
 
'./TAXTITY OF STEAM USED, ETC. 
 
 75 
 
 SINGLE UN JACKETED CYLINDERS. 
 
 To continue our investigation, we may set forth our results, obtained 
 from the foregoing table, as follows: 
 
 U. S. S. S. "MICHIGAN." 
 
 Steam used in Non- 
 Steam used per to- conducting Cylin- 
 tal horse-power per der per tot'l horse- 
 hour, power per hour. 
 
 35 
 
 30.9 
 
 29.4 
 
 30.6 
 
 29.8 
 
 30.7 
 
 32.0 
 
 Excess per total 
 
 horse-power 
 
 per hour. 
 
 23.3 
 20.0 
 16.3 
 15.4 
 13.5 
 11.7 
 
 6 
 
 7.6 
 9.4 
 14.3 
 14.4 
 17.2 
 20.3 
 
 Excess per hour 
 used. 
 
 2,065 
 1,709 
 2,160 
 2,200 
 2,071 
 1,552 
 1,782 
 
 -T- 7 = 1,934 ponuds, mean. 
 
 13,539 
 
 The excess found clearly follows no law connected with the expansion, 
 and we shall hereafter be justified in taking the mean value as that to be 
 followed. 
 
 U. S. B. S. "DALLAS." 
 
 Steam used per to- 
 tal horse-power per 
 hour. 
 
 Steam used in Non- 
 conducting Cylin- 
 der per tot'l horse- 
 power per hour. 
 
 Excess per total 
 horse-power 
 per hour. 
 
 Excess per 
 used. 
 
 hour 
 
 8 23.0 
 9 23.3 
 10 23.1 
 11 24.8 
 12 26.5 
 
 13.2 
 16.0 
 16.2 
 16.3 
 19.5 
 
 9.8 
 7.3 
 6.9 
 8 5 
 7.0 
 
 1,562 
 
 1,584 
 1,778 
 2,400 
 1,935 
 
 
 9,259 -i- 5 = 1,852 pounds, mean. 
 
 9,259 
 
 U. S. R. S. "DEXTER. 1 
 
 21.7 
 21.9 
 21.8 
 21.6 
 
 13.6 
 14.7 
 15.0 
 16.4 
 
 8.1 
 7.2 
 6.8 
 5.2 
 
 -f- 4 = 1,650 pounds, mean. 
 
 25.6 
 25.9 
 28.2 
 
 15.9 
 18.0 
 19.7 
 
 9.7 
 7.9 
 
 4,660 -i- 3 = 1,553 pounds, mean. 
 
 1,360 
 1.430 
 1,870 
 
 4,660 
 
76 
 
 STEAM USING; OR, STEAM ENGINE PRACT WE. 
 
 SINGLE UNJACKETED CYLINDEKS. Continued. 
 U. S. R. S. "GALLATIN." 
 
 
 
 Steam used in Non- 
 
 
 Steam used per to- conducting Cylin- Excess per total Excess per hour 
 tal horse-power per; der per tot'l horse- horse-power used, 
 hour. power per hour. per hour. 
 
 
 1 
 
 
 19.2 
 
 13.8 
 
 5.4 
 
 1,430 
 
 
 19 4 
 
 15.1 
 
 4.3 
 
 1,330 
 
 
 21.2 
 
 17.2 4.0 
 
 1,420 
 
 
 
 4,180 -T- 3 = 1,393 pounds, mean. 
 
 4,180 
 
 23 
 
 19.8 
 
 13.1 6.7 
 
 2,120 
 
 24 
 
 20.5 
 
 14.3 
 
 6.2 
 
 1,990- 
 
 25 
 
 21.5 
 
 13.6 
 
 7.9 2,410 
 
 26 
 
 21.3 
 
 11.3 
 
 10.0 2,170 
 
 27 
 
 20.6 
 
 12.4 
 
 8.2 
 
 2,200 
 
 28 
 
 19.6 
 
 13.1 
 
 6.5 
 
 2,030 
 
 29 
 
 21.5 
 
 14.0 
 
 7.5 
 
 2,360 
 
 
 
 15,280 - 7 = 2,183 pounds, mean. 
 
 15,280 
 
 00 
 
 22.4 
 
 12.5 9.9 
 
 1,450 
 
 31 
 
 23.0 
 
 13.2 9.8 
 
 1,440 
 
 32 
 
 21.3 
 
 14.9 
 
 6.4 
 
 1,310 
 
 33 
 
 21.8 
 
 16.9 
 
 4.9 
 
 1,310 
 
 34 
 
 23.4 
 
 18.7 4.7 
 
 1,340 
 
 
 
 6,850 -r 5 1,370 pounds, mean. 
 
 6,850 
 
 
 MILLERS' EXHIBITION AT CINCINNATI. 
 
 35 
 36 
 37 
 
 1 - 
 
 11.5 
 
 5.6 
 
 1,058 
 
 40 
 41 
 42 
 43 
 44! 
 
 17.24 
 18.61 
 
 36.4 
 29.2 
 31.1 
 32.1 
 30.3 
 
 HIRN'S ENGINE. 
 
 11.8 
 14.4 
 
 5.4 
 4.2 
 
 1,310 -r- 2 = 655 pounds, mean. 
 This will be increased on account of the 
 French units used, and will equal 733 
 pounds, mean. 
 
 U. S. S. "EUTAW." 
 
 15.4 
 17.5 
 21.5 
 21.6 
 23.0 
 
 21.0 
 11.7 
 
 9.6 
 10.5 
 
 7.5 
 
 16,670 -f- 5 = 3,337 pounds, mean. 
 
 650 
 660 
 
 1.310 
 
 3,520 
 2,810 
 3,760 
 3,370 
 3,210 
 
 16,670 
 
 HIGH SERVICE PUMPING ENGINE, NO. 1, ST. LOUIS WATER WORKS. 
 1 | 15.3 4.3 4,120 
 
 MASSACHUSETTS INSTITUTE OF TECHNOLOGY. 
 
 46 2 , . 7 
 
 14.4 
 
 13.3 
 
 173 
 
 47 
 
 29.0 
 
 19.6 
 
 9.5 
 
 175 
 
 48 
 
 33.5 
 
 24.0 
 
 9.5 
 
 197 
 
 
 
 545 -4- 3 = 182 r 
 
 ounds. mean. 
 
 545 
 
QUA XT/TV <>r STA'.I.V 1'SED, ETC. 77 
 
 It will be observed that we have summed up the experiments, made 
 at the Millers' Exhibition, at Cincinnati, on three different competing 
 engines. These showed so little difference that the results of our investi- 
 gation on one may be taken to represent the three. 
 
 We have also summed up the following data from the table: 
 
 U. S. STEAMER "MICHIGAN." Seven experiments: Cylinder 
 36" x 96". Initial steam pressure, 20 Ibs. above atmosphere. 
 Back pressure, say 5 Ibs., equal to 10 Ibs. below atmosphere. 
 Mean excess of 'water per hour over that required in a non- 
 conducting cylinder 1,934 Ibs. 
 
 U. S. KEVENUE STEAMER "DALLAS." Six experiments: Cylin- 
 der 36" x 30". Initial steam pressure, say 30 Ibs. above 
 atmosphere. Mean back pressure, say 5 Ibs. above zero. 
 Mean excess of water per hour above that required by a 
 non-conducting cylinder 1,852 Ibs. 
 
 U. S. KEVENUE STEAMER "DEXTER." Four experiments: Cylin- 
 der 26" x 30". Mean back pressure, say, 5 Ibs. above zero. 
 Initial steam pressure above atmosphere, 68 Ibs. Excess of 
 water per hour above that required by a non conducting 
 cylinder '. 1,650 Ibs. 
 
 Three experiments: Cylinder 26" x 30" Mean back pressure, 
 say, 5 Ibs. above zero. Initial steam pressure above atmos- 
 phere, 40 Ibs. Excess of water per hour above that required 
 by a non-conducting cylinder 1,553 Ibs. 
 
 U. S. KEVENUE STEAMER "GALLATIN." Mean of seven experi- 
 ments: Cylinder 34.1" x 30". Mean back pressure, say, 5 
 Ibs. above zero. Initial steam pressure 70 Ibs. above atmos- 
 phere. Excess of water per hour above that required by a 
 non-conducting cylinder 2,183 Ibs. 
 
 Mean of five experiments: Cylinder 34.1" x 30". Mean back 
 pressure, say, 5 Ibs. above zero. Initial steam pressure 40 
 Ibs. above atmosphere. Excess of water per hour above 
 that required by a non-conducting cylinder 1,370 Ibs. 
 
 Mean of three experiments: Cylinder 34.1" x 30". Mean back 
 pressure, 2 Ibs. above atmosphere. Initial steam pressure,, 
 70 Ibs. above atmosphere. Excess of water per hour above 
 that required by a non-conducting cylinder 1,393 Ibs. 
 
 MILLER'S EXHIBITION AT CINCINNATI. Mean of three engines: 
 Cylinder 18" x 48". Initial steam pressure, above atmos- 
 phere, 82 Ibs. Mean back pressure, say 4 Ibs. Excess of 
 water per hour above that required by a non-conducting 
 cylinder 1,058 Ibs. 
 
78 STEAM USING; 07?, STEAM ENGINE PRACTICE. 
 
 HIBN'S ENGINE. Mean of two experiments: Cylinder 24" x 
 78". Initial steam pressure above atmosphere 54 Ibs. Mean 
 back pressure, 2 Ibs. Excess of water, per hour, above that 
 required by a non-conducting cylinder 733 Ibs. 
 
 U. S. STEAMEB "EUTAW." Mean of five experiments: Cylinder 
 58" x 105". Initial steam pressure, 25 Ibs. above atmos- 
 phere. Mean back pressure, 4 Ibs. Excess of water, per 
 hour, above that required by a non-conducting cylinder. . . 3,337 Ibs. 
 
 MASSACHUSETTS INSTITUTE or TECHNOLOGY. Mean of three 
 experiments with Corliss engine: Cylinder 8" x 24". Initial 
 steam pressure, 50 Ibs. above atmosphere. Excess of water 
 per hour above that required by a non-conducting cylinder. 182 Ibs. 
 
 HIGH SEBVICE PUMPING ENGINE, No. 1, ST. Louis WATEE 
 WOBKS. One experiment: Cylinder 85" x 120". Initial 
 steam pressure, 40 Ibs. above atmosphere. Excess of water 
 per hour above that required by a non-conducting cyl- 
 inder 4,120 Ibs. 
 
 Of the four causes of excess in steam used over that required by a 
 non-conducting cylinder, which we have already mentioned, the first or 
 that of steam entering with water caused by foaming or priming, we shall 
 neglect here, as, in some of our experiments, to -wit, the last two, this has 
 been eliminated, and in others can scarcely be very large. The third, or 
 that of external radiation, is usually very small and can not exceed that of 
 a steam-heating coil of the same area as the external surface of the cylin- 
 der. The second is a loss which may, in determining the cost of a total 
 horse-power, be considered to vary with the volume of clearance space, 
 but which ranges from less than 1 to 15 per cent, of the piston displace- 
 ment, and is, on an average, about 8 per cent, thereof. This, in a cylinder 
 full of steam at the terminal pressure, is not a great loss in itself, but 
 there is a loss during the expansion also. In any case it is not large com- 
 paratively, is nearly proportional to the volume, and, consequently, varies 
 with the piston area and a percentage of the stroke. 
 
 The fourth source of loss is by far the largest, and is due to an action 
 first mentioned by D. K. Clark in his "Eailway Machinery, " in 1851, and 
 afterwards elaborated by M. G. A. Him, in 1854 and again in 1857, in the 
 Bulletins de la Societe Industrielle de Mulhouse, and therein repeated at 
 frequent intervals up to the present date. This action was subsequently 
 noted by Isherwood in the second volume of his "Experimental Kesearches 
 in Steam Engineering," and it was rediscovered by Mr. G. B. Dixwell, of 
 Boston, and communicated by him to the Society of Arts of that city. 
 
 The internal radiation, or the action of the internal surfaces of the cyl- 
 inder upon the within contained steam, may be explained as follows: 
 Steam enters the cylinder from the boiler at ,a temperature corresponding 
 
QUANTITY OF STEAM USED, ETC. 79 
 
 to the pressure, and leaves the cylinder at a lower temperature correspond- 
 ing to the lower pressure. The metal of the cylinder being a very good con- 
 ductor of heat, receives heat from the incoming and delivers heat to the 
 outgoing steam at every revolution. In detail the action is thus: When 
 the steam-valve opens there is admitted from the boiler hot steam which 
 first fills the clearance space, coming in contact with the cool surfaces 
 which have just been open to the exhaust surfaces, and which are from 
 100 to 200 Fahr. lower in temperature than the incoming steam. The 
 amount in weight of the steam is small and the amount in surface of 
 the enclosing metal being large, the result is naturally that the steam con- 
 denses until heat enough has been given to the metal to raise its surface 
 to the temperature of the steam. The piston moves and the condensation 
 continues up to the cut-off. During the expansion, as the pressure falls, 
 the warmed surface begins to give out heat to the steam as the pressure 
 and temperature of the steam fall, while, as the piston moves, the metal 
 which has been exposed to the exhaust is opened to the steam and the ac- 
 tion is the reverse of that going on at other portions of the surface, while 
 during exhaust the action continues to transfer heat from the metal to the 
 steam which is swept out of the cylinder. 
 
 There is experimental reason to believe that the temperature of what 
 may be called the skin of the metal scarcely varies from that of the steam, 
 while the depth to whjch the influence extends, or what may be called the 
 thickness of this skin, depends upon the intensity and rapidity of the 
 changes of temperature to which it is subjected. 
 
 The experimental evidence is as follows: A metallic pyrometer must 
 be so made that a thin sheet of metal can be exposed to the steam in the 
 cylinder, or connected to the indicator fittings, having a needle so adjusted 
 and arranged as to show changes of length in the sheet of metal. The 
 instrument must be rated by exposure to steam free from air at atmos- 
 pheric pressure and to water of a known temperature. On exposing such 
 an instrument to the action of the steam in the cylinder, a change of tem- 
 perature will be noted at each stroke. 
 
 If the shell be made of iron 0.03 inch thick, and the piston have a 
 speed due to 60 revolutions of the crank per minute, nearly the whole 
 change of temperature due to the change in pressure, and the needle, will 
 remain stationary during nine -tenths of the exhaust stroke. 
 
 If the instrument be filled with mercury so that heat may be trans- 
 mitted to the interior through the skin, while the freedom of movement 
 of the skin, by which alone the change of temperature can be observed, 
 is not interfered with, at a piston speed due to 85 revolutions per minute, 
 the same change has been observed in the action of the instrument as be- 
 fore the introduction of the mercury. 
 
 If the number of revolutions per minute be increased beyond, say 100, 
 the indications of the instrument decrease, and are, approximately, in- 
 versely proportional to the number of revolutions. 
 
 The problem of the transfer of heat to and from the steam, in an en- 
 gine cylinder, although complex, is probably within the compass of pure 
 
80 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 mathematics, but we shall not attempt to analyse it here, for it would be 
 foreign to the spirit of this work. 
 
 The fourth loss might be considered to be proportional to: 1. The 
 whole internal surface of the cylinder: 2. To the area of the cylinder 
 and piston heads and a fraction of the barrel: 3. To the area of the piston: 
 and, 4. To the diameter of piston. It may also be considered to vary with 
 the difference of temperature between initial pressure and that of the 
 condenser or exhaust pipe. 
 
 A careful examination of the table of experiments shows: that neglect- 
 ing priming and external radiation the whole excess of water used per 
 hour over that required in a non-conducting cylinder is rudely propor- 
 tional to the difference of temperature between the incoming and outgoing 
 steam, and to the diameter of the piston; and that such excess is nearly 
 constant for the great range of piston speed and revolutions therein 
 found, and moreover is entirely independent of the expansion. 
 
 We give some of the figures connected with the experiments in refer- 
 ence to the above points: 
 
 TJ. S. STEAMER "MICHIGAN." 
 
 Seven experiments. Diameter of piston, 3 feet. Excess of water in 
 pounds per hour over that required by a non-conducting cylinder = 1,934. 
 
 Temperature of steam 259 
 
 Temperature of condenser 104 
 
 259 104 = 155 = change of temperature: 
 
 155 x 3 = diameter of piston = 465 = product of change of temperature x 
 diameter of piston. 1,934 -=- 465 = 4 16 = pounds of water in excess per 
 hour, per foot diameter of piston, per degree Fahr. of change in tem- 
 perature. 
 
 TJ. S. STEAMER "DALLAS." 
 
 Diameter of piston, 3 feet. Excess of water in pounds per hour over 
 that required by a non-conducting cylinder = 1,852. 
 
 Temperature of steam 274 
 
 Temperature of condenser 104 
 
 274 104 = 170 = change of temperature: 
 
 170 x 3 = 510 = product of change of temperature x diameter of piston. 
 1,852 -f- 510 = 3.63 = pounds of water in excess per hour, per foot diameter 
 of piston, per degree Fahr. of change of temperature. 
 
 U. S. STEAMER "DEXTER." 
 
 Diameter of piston 2.17 feet. Excess of water in pounds per hour 
 over that required by a non-conducting cylinder = 1,650. 
 
 Temperature of steam 315 
 
 Temperature of condenser 104 
 
 315 104 = 211 = change of temperature: 
 
 211 x 2.17 = say 458 = product of change of temperature x diameter of 
 piston. 1,650 -=- 458 = 3.60 = pounds of water in excess per hour, per foot 
 of piston diameter, per degree Fahr. of change of temperature. 
 
QUANTITY OF STEAM USED, ETC. 81 
 
 Excess of water in pounds per hour over that required by a non- 
 conducting cylinder = 1,553. 
 
 Temperature of steam 287 
 
 Temperature of condenser 104 
 
 287 104 = 183 = change of temperature: 
 
 183 x 2.17 = 397 = product of change of temperature x diameter of pis- 
 ton. 1,553 -4- 397 = 3.91 = pounds of water in excess per hour, per foot 
 diameter of piston, per degree Fahr. of change of temperature. 
 
 U. S. REVENUE STEAMER "GALLATIN." 
 
 Mean of seven experiments: Diameter of piston 2.84 feet. Excess of 
 water in pounds per hour over that required by a non-conducting cylin- 
 der = 2,183. 
 
 Temperature of steam 316 
 
 Temperature of condenser 104 
 
 316 104 = 212 = change of temperature: 
 
 212 X 2.84 = say 602 = product of change of temperature x diameter of 
 piston. 2,183 -4- 602 = 3.62 = pounds of water in excess per hour, per foot 
 diameter of piston, per degree Fahr. of change of temperature. 
 
 Mean of five experiments: Excess of water in pounds per hour over 
 that required by a non-conducting cylinder = 1,427. 
 
 Temperature of steam 287 
 
 Temperature of condenser 104 
 
 287 104 = 183 = change of temperature: 
 
 183 x 2.84 = say 520 = product of change of temperature x diameter of 
 piston. 1,370 -=- 520 = 2.63 = pounds of water in excess per hour, per foot 
 diameter of piston, per degree Fahr. of change of temperature. 
 
 Mean of three experiments: Excess of water in pounds per hour over 
 that required by a non-conducting cylinder = 1,380. 
 
 Temperature of steam 316 
 
 Temperature of condenser 212 
 
 316 212 = 104 = change of temperature: 
 
 104 x 2.84 = 295 = product of change of temperature x diameter of piston. 
 1,393 -=- 295 = 4.72 = pounds of water in excess per hour, per foot diometer 
 of piston, per degree Fahr. of change of temperature. 
 
 MILLER'S EXHIBITION AT CINCINNATI. 
 
 Mean of three engines: Diameter of piston 1.5 feet. Excess of water 
 in pounds per hour over that required by a non-conducting cylinder = 
 1,058. 
 
 Temperature of steam 326 
 
 Temperature of condenser 104 
 
 326 104 = 222 = change of temperature: 
 
 222 x 1.5 = 333 = product of change of temperature x diameter of piston. 
 1,058 -f- 333 = 3.17 = pounds of water in excess per hour, per foot diameter 
 of piston, per degree Fahr. of change of temperature. 
 
82 STEAM rSING; OR, STEAM E\<;1XK PRACTICE. 
 
 HIKN'S ENGINE. 
 
 Diameter of piston 2 feet. Excess of water in pounds per hour over 
 that required by a non-conducting cylinder = 733. 
 
 Temperature of steam , 302 
 
 Temperature of condenser 104 
 
 302 104 198 = change of temperature: 
 
 198 x 2 = 396 = product of change of temperature x diameter of piston. 
 733 -4- 396 = 1.85 = pounds of water in excess per hour, per foot diame- 
 ter of piston, per degree Fahr. of change of temperature. 
 
 U. S. STEAMEK "EUTAW." 
 
 Diameter of piston = 4.83 feet. Excess of water in pounds per hour 
 over that required by a non-conducting cylinder 3,334. 
 
 Temperature of steam 267 
 
 Temperature of condenser 104 
 
 267 104 = 163 = change of temperature: 
 
 163 x 4.83 = 787 = product of change of temperature x diameter of piston. 
 3,337 -r- 787 = 4.24 = pounds of water in excess per hour, per foot diame- 
 ter of piston, per degree Fahr. of change of temperature. 
 
 MASSACHUSETTS INSTITUTE OF TECHNOLOGY. 
 
 Diameter of piston = 0.67 foot. Excess of water in pounds per hour 
 over that required by a non-conducting cylinder = 182. 
 
 Temperature of steam 298 
 
 Temperature of condenser 212 
 
 298 212 = 86 = change of temperature: 
 
 86 x 0.67 = 58 = product of change of temperature x diameter of piston. 
 182 ~- 58 = 3.14 = pounds of water in excess per hour, per foot diameter 
 of piston, per degree Fahr. of change of temperature. 
 
 HIGH SEBVICE PUMPING ENGINE, NO. 1, ST. LOUIS WATER WORKS. 
 
 Diameter of piston = 7.08 feet. Excess of water in pounds per hour 
 over that required by a non-conducting cylinder = 4,120. 
 
 Temperature of steam 287 
 
 Temperature of condenser ^20 
 
 287 120 = 167 = change of temperature. 
 
 167 x 7.08 = 1,182 = product of change of temperature x diameter of pis- 
 ton. 4,120 -f- 1.182 = 3.48 pounds of water in excess per hour, per foot 
 diameter of piston, per degree Fahr. of change of temperature. 
 
QUANTITY OF STEAM USED, ETC. 
 
 83 
 
 SUMMARY. 
 
 IL 
 
 If 
 
 LOCATION. 
 
 
 20 
 
 36 
 
 7 
 
 U. S. Steamer "Michigan" .. .. 
 
 4 16 
 
 29 12 
 
 30 
 
 36 
 
 6 
 
 U. S. Steamer "Dallas" 
 
 3.63 
 
 21.78 
 
 70 
 
 26 
 
 4 
 
 U. S. Steamer "Dexter" 
 
 3.60 
 
 14 40 
 
 40 
 
 
 3 
 
 
 3 91 
 
 11 73 
 
 70 
 
 34 
 
 
 U. S Steamer "Gallatin" 
 
 3 62 
 
 25.34 
 
 40 
 
 
 5 
 
 
 2.63 
 
 13.15 
 
 70 
 
 no vacuum 
 
 3 
 
 m m n 
 
 4.72 
 
 14.16 
 
 25 
 
 58 
 
 5 
 
 U. S. Steamer "Eutaw" 
 
 4 24 
 
 21 20 
 
 80 
 
 18 
 
 3 
 
 Miller's Exhibition 
 
 3.17 
 
 9.51 
 
 40 
 54 
 
 85 
 24 
 
 1 
 9 
 
 St. Louis Water Works, No. 1 H. S. . 
 Him 
 
 3.48 
 1 85 
 
 3.48 
 3.70 
 
 50 
 
 8 
 
 3 
 
 Mass. Institute Technology 
 
 3.14 
 
 9.42 
 
 
 
 
 
 
 
 
 
 49 
 
 
 
 176.99 
 
 176.99 -r- 49 = say 3.6 = mean pounds of water in excess per hour, per 
 foot of piston diameter, per degree Fahr. difference of temperature. 
 
 In the 49 experiments, above recorded, we find a certain variation in 
 the resulting excess per foot of piston diameter per degree of change of 
 temperature, but in connection with this we must remember that we have 
 not taken into account the difference of clearance between the different 
 engines. For instance, the lowest value given above is that for Hirn's 
 engine, and this has the least clearance, while the condition of the steam 
 and the amount of cushion are in all the cases neglected. Furthermore, 
 while the results are widely different, yet the error in per cent, of the 
 whole steam used is a much smaller one, and we shall find that adding to 
 the steam used in a non-conducting cylinder the excess found above, we 
 shall arrive at a close approximation to the steam actually used. 
 
 When we examine the cases of single -jacketed cylinders, we find, as a 
 whole, a less excess in the use of steam above that used in a non-conduct- 
 ing cylinder, but the gain so made does not appear to be reduced to any 
 such simple law as that found for unjacketed engines, and, in fact, the use 
 of larger expansion, and consequent loss by back pressure work, very 
 often neutralizes the gain achieved. 
 
 The compound engines, in our table, give very little better results 
 than the simple engines, with the same steam pressures and expansions, in 
 the cost of steam per total horse power; while per net horse power the 
 larger amount of back pressure work, and the actual friction of two pis- 
 tons, with their rods and set-off connections, go very far to neutralize any 
 very great gain in such types. 
 
84 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 We find that the data are not sufficient to give an empirical formula 
 for the excess of water over that used in a non-conducting cylinder; but 
 we see that it is not very far from that of a single-jacketed cylinder of the 
 same size as the large one. We are obliged, therefore, to await further 
 experiments. 
 
 We do not claim, even, for the single un jacketed cylinder that our 
 method of investigation is either final, exhaustive, or rational, but that 
 the results are all that our present knowledge of the subject will give us, 
 will, we think, be admitted. What is required is a great number of 
 experiments under the conditions of class A, upon all kinds and sizes of 
 engines; we can then hope to frame a much more accurate and rational 
 theory than the crude one we have given. 
 
 We add a few tables, the application of which will be readily seen. 
 
 NUMBER OF POUNDS OF WATER USED PER SQUARE FOOT OF PISTON PER 
 HOUR, FOR A PISTON SPEED OF ONE FOOT PER MINUTE IN A NON- 
 CONDUCTING CYLINDER. 
 
 INITIAL PEESSUBE IN POUNDS PEE SQUAEE INCH ABOVE ATMOSPHEEE. 
 
 No. of 
 
 
 
 
 
 
 
 
 
 
 
 Expan- 
 sions. 
 
 
 
 20 
 
 40 
 
 60 
 
 80 
 
 100 
 
 120 
 
 140 
 
 180 
 
 220 
 
 1.1 
 
 2.12 
 
 4.69 
 
 7.35 
 
 9.66 
 
 12.1 
 
 14.4 
 
 16.8 
 
 19.0 
 
 23.6 
 
 28.2 
 
 1.2 
 
 .92 
 
 4.30 
 
 6.64 
 
 8.86 
 
 11- 1 
 
 13.2 
 
 15.4 
 
 17.4 
 
 21.6 
 
 32.5 
 
 1.3 
 
 .75 
 
 3.88 
 
 5.93 
 
 7.99 
 
 9.99 
 
 11.9 
 
 13.9 
 
 15.7 
 
 19.5 
 
 23.3 
 
 1.4 
 
 .63 
 
 3.60 
 
 5.51 
 
 7.42 
 
 S.28 
 
 11.1 
 
 12.9 
 
 14.6 
 
 18.1 
 
 21.6 
 
 1.5 
 
 .52 
 
 3.36 
 
 5.14 
 
 6.91 
 
 8.66 
 
 10.4 
 
 12.0 
 
 13.6 
 
 16.9 
 
 20.2 
 
 1.6 
 
 .46 
 
 3.22 
 
 4.93 
 
 6.64 
 
 8.31 
 
 9.91 
 
 11.5 
 
 13.1 
 
 16.2 
 
 19.3 
 
 1.8 
 
 .30 
 
 2.86 
 
 4.38 
 
 5.90 
 
 7.38 
 
 8.83 
 
 10.2 
 
 11.6 
 
 14.4 
 
 17.2 
 
 2.0 
 
 .17 
 
 2.58 
 
 3.95 
 
 5.44 
 
 6.65 
 
 7.95 
 
 9.21 
 
 10.5 
 
 13.3 
 
 15.5 
 
 2.5 
 
 0.933 
 
 2.06 
 
 3.16 
 
 4.25 
 
 5.29 
 
 6.36 
 
 7.37 
 
 8.36 
 
 10 4 
 
 12.4 
 
 3.0 
 
 0.777 
 
 1.72 
 
 2.63 
 
 3.54 
 
 4.43 
 
 5.30 
 
 6.14 
 
 6.97 
 
 8.66 
 
 10.3 
 
 4.0 
 
 0.583 
 
 1.26 
 
 1.97 
 
 2.66 
 
 3.32 
 
 3.97 
 
 4.58 
 
 5.23 
 
 6.49 
 
 7.74 
 
 5.0 
 
 0.466 
 
 1.03 
 
 1.58 
 
 2.17 
 
 2.66 
 
 3.19 
 
 3.69 
 
 4.18 
 
 4.96 
 
 6 19 
 
 6.0 
 
 0.389 
 
 0.859 
 
 1.32 
 
 1.77 
 
 2.07 
 
 2.65 
 
 3.07 
 
 3.48 
 
 4.32 
 
 5.16 
 
 7.0 
 
 0.333 
 
 0.736 
 
 1.12 
 
 1.55 
 
 1.90 
 
 2.32 
 
 2.63 
 
 2.99 
 
 3.71 
 
 4.42 
 
 8.0 
 
 0.292 
 
 0.644 
 
 0.987 
 
 1.33 
 
 1.66 
 
 1.98 
 
 2.30 
 
 2.61 
 
 3.25 
 
 3.87 
 
 9.0 
 
 0.258 
 
 0.571 
 
 0.875 
 
 1.18 
 
 1.47 
 
 1.76 
 
 2.04 
 
 2.32 
 
 2.88 
 
 3.43 
 
 10.0 
 
 0.233 
 
 0.516 
 
 0.789 
 
 1.06 
 
 1.33 
 
 1.59 
 
 1.84 
 
 2-09 
 
 2.60 
 
 3.09 
 
 11.0 
 
 0.212 
 
 0.235 
 
 0.717 
 
 0.966 
 
 1.21 
 
 1.44 
 
 1.68 
 
 1.90 
 
 2.36 
 
 2.81 
 
 12.0 
 
 0.194 
 
 0.430 
 
 0.658 
 
 0.886 
 
 1.11 
 
 1.32 
 
 1.54 
 
 1.78 
 
 2.16 
 
 2.58 
 
 13.0 
 
 0.180 
 
 0.397 
 
 0.593 
 
 0.818 
 
 1.03 
 
 1.25 
 
 1.39 
 
 1.61 
 
 2.00 
 
 2.38 
 
 14.0 
 
 0.167 
 
 0.369 
 
 0.564 
 
 0.759 
 
 0.949 
 
 1.14 
 
 1.32 
 
 1.49 
 
 1.85 
 
 2.21 
 
 16.0 
 
 0.146 
 
 0.322 
 
 0.494 
 
 0.664 
 
 0.831 
 
 0.993 
 
 1.15 
 
 1.31 
 
 1.62 
 
 1.93 
 
 20.0 
 
 0.117 
 
 0.258 
 
 0.395 
 
 0.531 
 
 0.665 
 
 0.775 
 
 0.921 
 
 1.04 
 
 1.30 
 
 1.54 
 
 25.0 
 
 0.0933 
 
 0.206 
 
 0.316 
 
 0.425 
 
 0.532 
 
 0.636 
 
 0.737 
 
 0.836 
 
 1.04 
 
 1.21 
 
 30.0 
 
 0.0777 
 
 0.172 
 
 0.263 
 
 0.354 
 
 0.443 
 
 0.526 
 
 0.587 
 
 0.697 
 
 0.866 
 
 1.03 
 
QUANTITY OF STEAM USED, ETC. 
 
 85 
 
 NUMBER OF TOTAL HORSE-POWER FOR EACH CUBIC FOOT OF SPACE 
 SWEPT BY PISTON PER MINUTE. 
 
 INITIAL PBESSUBE IN POUNDS PEE SQUABE INCH ABOVE ATMOSPHEBE. 
 
 No. of 
 
 
 
 
 
 
 
 
 
 
 
 Expan- 
 sions. 
 
 
 
 20 
 
 40 
 
 60 
 
 80 
 
 100 
 
 120 
 
 140 
 
 180 
 
 220 
 
 .1 
 
 0.0649 
 
 0.151 
 
 0.238 
 
 0.324 
 
 0.410 
 
 0.497 
 
 0.584 
 
 0.670 
 
 0.843 
 
 1.020 
 
 .2 
 
 0.0644 
 
 0.150 
 
 0.236 
 
 0.322 
 
 0.408 
 
 0.494 
 
 0.579 
 
 0.665 
 
 0.837 
 
 1.010 
 
 .3 
 
 0.0635 
 
 0.148 
 
 0.233 
 
 0.317 
 
 0.402 
 
 0.487 
 
 0.572 
 
 0.656 
 
 0.825 
 
 0.995 
 
 .4 
 
 0.0624 
 
 0.145 
 
 0.229 
 
 0.312 
 
 0.395 
 
 0.478 
 
 0.561 
 
 0.644 
 
 0.810 
 
 0.977 
 
 .5 
 
 0.0609 
 
 0.142 
 
 0.224 
 
 0.304 
 
 0.386 
 
 0.467 
 
 0.548 
 
 0.629 
 
 0.792 
 
 0.955 
 
 1.6 
 
 0.0597 
 
 0.139 
 
 0.219 
 
 0.299 
 
 0.378 
 
 0.458 
 
 0.537 
 
 0.617 
 
 0.776 
 
 0.935 
 
 1-8 
 
 0.0567 
 
 0.132 
 
 0.208 
 
 0.283 
 
 0.359 
 
 0.435 
 
 0.510 
 
 0.585 
 
 0.737 
 
 0.888 
 
 2.0 
 
 0.0545 
 
 0.127 
 
 0.200 
 
 0.279 
 
 0.345 
 
 0.418 
 
 0.491 
 
 0.563 
 
 0.725 
 
 0.854 
 
 2 5 
 
 0.0490 
 
 0.114 
 
 0.180 
 
 0.245 
 
 0.311 
 
 0.376 
 
 0.441 
 
 0.507 
 
 0.638 
 
 0.768 
 
 3.0 
 
 0444 
 
 0.103 
 
 0.163 
 
 0.222 
 
 0.281 
 
 0.340 
 
 0.400 
 
 0.459 
 
 0.577 
 
 0.696 
 
 4.0 
 
 0.0372 
 
 0.0868 
 
 0.136 
 
 0.186 
 
 0.235 
 
 0.285 
 
 0.335 
 
 0.384 
 
 0.483 
 
 0.583 
 
 5.0 
 
 0.0324 
 
 0.0756 
 
 0.119 
 
 0.162 
 
 0.205 
 
 0.248 
 
 0.292 
 
 0.335 
 
 0.421 
 
 0.508 
 
 6.0 
 
 0.0286 
 
 0.0667 
 
 0.105 
 
 0.143 
 
 0.181 
 
 0.219 
 
 0.257 
 
 0.295 
 
 0.371 
 
 0.448 
 
 7.0 
 
 0.0257 
 
 0.0599 
 
 0.0942 
 
 0.131 
 
 0.163 
 
 0.201 
 
 0.231 
 
 0.265 
 
 0.334 
 
 0.402 
 
 8.0 
 
 0.0233 
 
 0.0545 
 
 0.0856 
 
 0.117 
 
 0.148 
 
 0.179 
 
 0.210 
 
 0.241 
 
 0.303 
 
 0.366 
 
 9.0 
 
 0.0215 
 
 0.0501 
 
 0.0788 
 
 0.107 
 
 0.137 
 
 0.165 
 
 0.193 
 
 0.222 
 
 0.279 
 
 0.337 
 
 10.0 
 
 0.0199 
 
 0.0463 
 
 0.0728 
 
 C.0994 
 
 0.126 
 
 0.152 
 
 0.179 
 
 0.205 
 
 0.258 
 
 0.311 
 
 11.0 
 
 0.0185 
 
 0.0441 
 
 0.0678 
 
 0.0924 
 
 0.117 
 
 0.142 
 
 0.166 
 
 0.191 
 
 0.240 
 
 0.290 
 
 12.0 
 
 0.0173 
 
 0.0403 
 
 0.0633 
 
 0.0864 
 
 0.109 
 
 0.132 
 
 0.155 
 
 0.183 
 
 0.224 
 
 0.271 
 
 13.0 
 
 0.0162 
 
 0.0379 
 
 0.0595 
 
 0.0811 
 
 0.103 
 
 0.127 
 
 0.146 
 
 0.168 
 
 Q.21'2 
 
 0.254 
 
 14.0 
 
 0.0154 
 
 0.0359 
 
 0.0563 
 
 0.0768 
 
 0.0973 
 
 0.118 
 
 0.138 
 
 0.159 
 
 0.200 
 
 0.241 
 
 16.0 
 
 0.0138 
 
 0.0323 
 
 0.0507 
 
 0.0692 
 
 0.0877 
 
 106 
 
 0.125 
 
 0.143 
 
 0.180 
 
 0.217 
 
 20 
 
 0.0116 
 
 0.0271 
 
 0.0426 
 
 0.0580 
 
 0735 
 
 0.089 
 
 0.104 
 
 0.120 
 
 0.151 
 
 0.182 
 
 25.0 
 
 0.0097 
 
 0.0227 
 
 0.0356 
 
 0.0486 
 
 0.0615 
 
 0.0745 
 
 0.0874 
 
 0.101 
 
 0.126 
 
 0.152 
 
 30.0 
 
 0084 
 
 0.0195 
 
 0.0307 
 
 0.0418 
 
 0.0530 
 
 0.0638 
 
 0.0753 
 
 0.0865 
 
 0.109 
 
 0.131 
 
 NUMBER OF INDICATED HORSE-POWER FOR EACH CUBIC FOOT SWEPT BY 
 PISTON PER MINUTE FOR CONDENSING ENGINES. 
 
 INITIAL PBESSUBE IN POUNDS PEB SQUABE INCH ABOVE ATMOSPHEBE. 
 
 No. of 
 
 
 
 
 
 
 
 
 
 
 Expan- 
 sions. 
 
 
 
 20 40 
 
 60 
 
 80 
 
 100 
 
 120 
 
 140 
 
 180 
 
 220 
 
 1.1 
 
 .0475 
 
 0.184 0.221 
 
 0.307 
 
 0.393 
 
 480 
 
 0.567 
 
 0.653 
 
 0.826 
 
 1.003 
 
 .2 
 
 .0470 
 
 .133 .219 
 
 .305 
 
 .391 
 
 .477 
 
 .562 
 
 .648 
 
 .820 
 
 0.993 
 
 .3 
 
 .0461 
 
 .131 .216 
 
 .300 
 
 .385 
 
 .470 
 
 .555 
 
 .639 
 
 .808 
 
 .978 
 
 .4 
 
 .0450 
 
 .128 .212 
 
 .295 
 
 .378 
 
 .461 
 
 .544 
 
 .627 
 
 .793 
 
 .960 
 
 .5 
 
 .0435 
 
 .125 ] .207 
 
 .287 
 
 .369 
 
 .450 
 
 .531 
 
 .612 
 
 .775 
 
 .938 
 
 .6 
 
 .0423 
 
 .122 i .202 
 
 .282 
 
 .361 
 
 .441 
 
 .520 
 
 .600 
 
 .759 
 
 .918 
 
 .8 
 
 .0393 
 
 .115 | .191 
 
 .266 
 
 .342 
 
 .418 
 
 .493 
 
 .568 
 
 .720 
 
 .861 
 
 2.0 
 
 .0371 
 
 .110 .183 
 
 .262 
 
 .328 
 
 .411 
 
 .474 
 
 .546 
 
 .708 
 
 .837 
 
 2.5 
 
 .0316 
 
 .097 I .163 
 
 .228 
 
 .294 
 
 .359 
 
 .424 
 
 .490 
 
 ,621 
 
 .751 
 
 3.0 
 
 .0270 
 
 .086 .146 
 
 .205 
 
 .264 
 
 .323 
 
 .383 
 
 .442 
 
 .560 
 
 .679 
 
 4.0 
 
 .0198 
 
 .0694 .119 
 
 .169 
 
 .218 
 
 .268 
 
 .318 
 
 .367 
 
 .466 
 
 .566 
 
 5.0 
 
 .0150 
 
 .0582 .102 
 
 .145 
 
 .188 
 
 .231 
 
 .275 
 
 .318 
 
 .404 
 
 .491 
 
 6.0 
 
 .0112 
 
 .0493 .088 
 
 .126 
 
 .164 
 
 .212 
 
 .240 
 
 .278 
 
 .354 
 
 .431 
 
 7.0 
 
 .0083 
 
 .0425 .0768 
 
 .114 
 
 .146 
 
 .184 
 
 .214 
 
 .248 
 
 .317 
 
 .3fc5 
 
 8.0 
 
 .0059 
 
 .0371 .0682 
 
 .100 
 
 .131 
 
 .162 
 
 .193 
 
 .224 
 
 .286 
 
 .349 
 
 9.0 
 
 .0041 
 
 .0327; .0614 
 
 .090 
 
 .120 
 
 .148 
 
 .176 
 
 .205 
 
 .262 
 
 .320 
 
 10.0 
 
 .0025 
 
 .0289 .0554 
 
 .0820 
 
 .109 
 
 .135 
 
 .162 
 
 .188 
 
 .241 
 
 .294 
 
 11.0 
 
 .0011 
 
 .0264 .0504 
 
 .0747 
 
 .100 
 
 .125 
 
 .149 
 
 .124 
 
 .223 
 
 .273 
 
 12.0 
 
 
 .0229 .0459 
 
 .0690 
 
 .092 
 
 .115 
 
 .138 
 
 .166 
 
 .207 
 
 .254 
 
 13.0 
 
 
 .0205 .0421 
 
 .0637 
 
 .086 
 
 .110 
 
 .129 
 
 .151 
 
 .195 
 
 .237 
 
 14.0 
 
 
 . 0185 . 0389 
 
 .0594 
 
 .0799 
 
 .101 
 
 !l21 
 
 .142 
 
 'l83 
 
 .224 
 
 16.0 
 
 
 . 0149 . 0323 
 
 .0518 
 
 .0703 
 
 .089 
 
 .108 
 
 .126 
 
 163 
 
 .200 
 
 20.0 
 
 
 .0097 .0252 
 
 .0406 
 
 .0561 
 
 .0716 
 
 .087 
 
 .103 
 
 .134 
 
 .165 
 
 25.0 
 
 
 
 .0053 !oi82 
 
 .0312 
 
 .0441 
 
 .0571 
 
 .0700 
 
 .084 
 
 109 
 
 .135 
 
 30.0 
 
 ::::."" 
 
 .0021 .0133 
 
 .0244 
 
 .0356 
 
 .0464 
 
 !o579 
 
 .0691 
 
 092 
 
 .114 
 
 
 
 
 
 
 
 
 
 
 
STEAM USING; 07?, STEAM ENGINE PRACTICE. 
 
 NUMBER OF INDICATED HORSE -POWER FOR EACH CUBIC FOOT SWEPT BY 
 PISTON PER MINUTE IN NON-CONDENSING ENGINES. 
 
 INITIAL PBESSUBE IN POUNDS PER SQUARE INCH ABOVE ATMOSPHERE. 
 
 No. of 
 
 
 
 
 
 
 
 
 
 
 
 Expan 
 sions. 
 
 
 
 20 
 
 40 
 
 60 
 
 80 
 
 100 
 
 120 
 
 140 
 
 160 
 
 220 
 
 1 
 
 
 073 
 
 160 
 
 246 
 
 0.332 
 
 419 
 
 0.506 
 
 592 
 
 765 
 
 942 
 
 2 
 
 
 0.072 
 
 0.158 
 
 0.244 
 
 330 
 
 0.416 
 
 501 
 
 0.587 
 
 759 
 
 932 
 
 .3 
 
 
 0.070 
 
 0.155 
 
 0.239 
 
 324 
 
 0.409 
 
 494 
 
 0.578 
 
 0.747 
 
 0.917 
 
 .4 
 
 
 0.067 
 
 0.151 
 
 0.234 
 
 0.317 
 
 0.400 
 
 0.483 
 
 0.566 
 
 0.732 
 
 0.899 
 
 .5 
 
 
 0.064 
 
 146 
 
 226 
 
 308 
 
 0.389 
 
 0.470 
 
 551 
 
 0.714 
 
 0.877 
 
 .6 
 
 8 
 
 
 0.061 
 054 
 
 0.141 
 0.130 
 
 0.221 
 0.205 
 
 0.300 
 281 
 
 0.380 
 357 
 
 0.459 
 0.432 
 
 0.539 
 507 
 
 0.698 
 0.659 
 
 0.857 
 810 
 
 2 
 
 
 0.049 
 
 0.122 
 
 0.201 
 
 0.267 
 
 0.340 
 
 0.413 
 
 0.485 
 
 0.647 
 
 0.776 
 
 2.5 
 
 
 0.036 
 
 0.102 
 
 0.167 
 
 0.233 
 
 0.298 
 
 0.363 
 
 0.439 
 
 0.560 
 
 0.690 
 
 3 
 
 
 0.025 
 
 085 
 
 0.144 
 
 203 
 
 0.262 
 
 322 
 
 381 
 
 0.499 
 
 618 
 
 4.0 
 
 
 0.009 
 
 0.058 
 
 0.108 
 
 0.157 
 
 0.207 
 
 0.257 
 
 0.306 
 
 0.405 
 
 0.505 
 
 5 
 
 
 
 0.041 
 
 0.084 
 
 0.127 
 
 170 
 
 214 
 
 257 
 
 0.343 
 
 430 
 
 6 
 
 
 
 0.027 
 
 0.065 
 
 0.103 
 
 0.141 
 
 0.179 
 
 0.217 
 
 0.293 
 
 0.370 
 
 7 
 
 
 
 0.016 
 
 0.053 
 
 0.085 
 
 0.123 
 
 0.153 
 
 0.187 
 
 0.256 
 
 0.324 
 
 8 
 
 
 
 009 
 
 0.039 
 
 0.070 
 
 101 
 
 132 
 
 0.163 
 
 0.225 
 
 0.288 
 
 9 
 
 
 
 
 0.029 
 
 0.059 
 
 0.087 
 
 0.115 
 
 0.144 
 
 0.201 
 
 0.259 
 
 10 
 
 
 
 
 021 
 
 0.048 
 
 0.074 
 
 0.101 
 
 0.127 
 
 180 
 
 0.233 
 
 11.0 
 
 
 
 
 014 
 
 0.039 
 
 0.064 
 
 0.088 
 
 0.113 
 
 0.162 
 
 0.212 
 
 12 
 
 
 
 
 009 
 
 031 
 
 0.054 
 
 0.077 
 
 105 
 
 146 
 
 193 
 
 13 
 
 
 
 
 
 025 
 
 0.049 
 
 068 
 
 090 
 
 0.134 
 
 176 
 
 14.0 
 
 
 
 
 
 019 
 
 0.040 
 
 0.060 
 
 0.081 
 
 0.122 
 
 0.163 
 
 16 
 
 
 
 
 
 010 
 
 028 
 
 0.047 
 
 0.065 
 
 0.102 
 
 139 
 
 20 
 
 
 
 
 
 
 0.011 
 
 086 
 
 
 0.073 
 
 0.104 
 
 25.0 
 
 
 
 
 
 
 
 0.009 
 
 
 0.048 
 
 0.074 
 
 30.0 
 
 
 
 
 
 
 
 
 
 031 
 
 0.053 
 
 
 
 
 
 
 
 
 
 
 
 
CHAPTER IV. 
 
 ON THE INDICATOR, THE INDICATOR DIAGRAM, AND THE DIFFERENT 
 CLASSES OF ENGINES. 
 
 James Watt constructed an instrument for observing the volume and 
 pressure of the steam in the engine cylinder to which he gave the name 
 of "Indicator." It consisted of two parts: One, a rectangular frame moving 
 in guides backward and forward and actuated by the engine, to the beam 
 of which it was connected by a cord, with a weight or spring attached 
 to keep the cord stretched. In this way the frame moved with the piston, 
 stopping when that member stopped, and corresponding with it com- 
 pletely. The other part carried a small cylinder with piston, piston rod, 
 and a spring in connection therewith. Steam being admitted to this cyl- 
 inder from the engine cylinder, the pressure would vary and the piston 
 would consequently rise and fall within the cylinder against the spring. 
 
 By attaching a pencil to the top of the piston rod. and with a piece 
 of paper secured to the frame, a diagram is obtained in which the position 
 of a point, vertically, in relation to a line traced when the steam is shut off 
 from the instrument, is governed by the steam pressure; while its position 
 horizontally is proportioned to the movement of the piston of the engine, 
 or the volume occupied by the steam. The diagram thus produced is pro- 
 portioned to the pressure in height and to the volume occupied in length, 
 and the area is therefore proportional to the work done in the engine 
 cylinder per stroke. The mean pressure in pounds per square inch is 
 found, either, by measuring at ten equi-distant places, or by measuring 
 the area of the figure and dividing by the length, of course taking into 
 account the stiffness, or scale, of the spring. The mean pressure times the 
 piston area in square inches multiplied by the stroke in feet gives the 
 number of foot pounds per stroke of engine. 
 
 By using the piston speed in feet per minute, in place of the stroke, 
 and dividing this by 33,000, the number of horse-power exerted is deter- 
 mined. 
 
 By using a mean pressure of two pounds less, the number of net 
 horse-power may be estimated. By setting off the line of no pressure, 
 14.7 pounds below the atmospheric line, and drawing verticals to the ends 
 of the diagram, and by using this as the diagram of an engine without back 
 pressure, we can find the mean total pressure and the number of total 
 horse-power. 
 
 In such an instrument, however, there are many imperfections. The 
 moving springs are subjected to sudden changes of pressure, or tension, 
 and vibrations in the springs are set up by the inertia of the moving parts. 
 This is true for both the pencil and paper movements. The motion of the 
 
88 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 paper from the engine motion is subject to errors both geometric and 
 mechanical: the former can only be reduced at the expense of the latter. 
 
 The remedy for vibrations of spring is to reduce the weight of the 
 moving parts, to decrease the amount of motion therein, and to increase 
 the stiffness of the spring itself. 
 
 The mechanical errors of connection are: the variations in length of 
 cords, and vibrations, and the slack in the bars used for reducing the mo- 
 tions. There is also a loss in pressure caused by the leakage of the indi- 
 cator piston, and for engines with small cylinders an error is introduced 
 by cylinder condensation in the indicator. 
 
 In large, slow moving engines these errors are all small and may be 
 neglected, but they increase in importance with the number of revolu- 
 tions, and, inversely, with the size of the engine. 
 
 The indicator was first improved by McNaught, with Kichards, Thomp- 
 son, Crosby, Tabor and Professor Sweet, following, in their efforts to re- 
 duce the weight of the moving parts, all using, however, a separate in- 
 strument. We give an illustration of Thompson's Indicator as a good 
 example. 
 
 THOMPSON'S STEAM ENGINE INDICATOR. 
 
 OUTSIDE VIEW. 
 
 INSIDE VIEW. 
 
ON THE INDICATOR, ETC. 89 
 
 M. G. A. Hirn employed the beam of the engine, the deflection being 
 properly multiplied, for the spring opposing the steam pressure; and Mr. 
 William E. Worthen has proposed the use of the spring of the cylinder 
 head of the engine for the same purpose. This would give very satisfac- 
 tory results with proper means for multiplying the motion. 
 
 In France a device has been tried which can be applied to engines 
 running with nearly constant load, but which is not well suited for en- 
 gines with quick acting automatic cut-off gear. An adjustable yoke con- 
 fines the movement of the pencil within narrow limits, and by changing 
 this adjustment the card may be taken piecemeal. An elegant adaptation 
 of this method was used on the South Eastern Railway of France, where 
 the adjustment was made by compressed air, and the diagrams from the 
 cylinders taken in the dynamometer car. 
 
 Engines may be classified as single acting and double acting, accord- 
 ing as the working steam is used in one or both ends of the cylinder; and 
 condensing and non- condensing, according as the steam from the cylinder 
 is cooled by water and condensed, or is exhausted directly into the air. 
 
 There are two kinds of single-acting engines. The one, the older, is 
 mainly employed in the pumping of water; and the other, the modern type, 
 maintains a high speed of revolution on a shaft, with the connecting rods 
 kept under stress of one kind, usually compression, so that the shock on 
 the boxes, due to change of force from compression to tension, and the 
 effects of wear, are avoided. These two kinds of engines are entirely 
 different in design and construction. 
 
 The oldest form of engine was worked by admitting steam from a 
 boiler to the cylinder, below the piston, which was connected with a beam 
 and pump rod. The weight of the pump rod forced the water up from 
 the pump and also carried the piston to the top of the cylinder. A jet 
 of water was then thrown into the cylinder condensing the therein 
 contained steam, which had been previously shut off from the boiler. 
 The pressure of the atmosphere now drove the piston down, at the 
 same time lifting the pump rod, when after it had reached the bottom 
 the water was discharged from the cylinder and the process was re- 
 peated. 
 
 James Watt devised the separate condenser and covered the cylinder 
 so as to introduce steam at its upper end, thereby lifting the pump rod. 
 When the piston reached the bottom of the cylinder the steam was cut off 
 from the boiler. An arrangement, called the equilibrium valve, opens 
 communication between the two ends of the cylinder, and the steam 
 now pressing equally above and below the piston, the weight of the 
 pump rod carries the piston to the top of the cylinder, driving the steam 
 which was above the piston to its lower end, and doing the pumping at 
 the same time. The steam is now admitted above the piston as before, 
 but the bottom of the cylinder is opened to the condenser and a vacuum 
 is produced below the piston at the same time as the steam exerts its 
 pressure above it. 
 
 These engines were introduced into the mines of Cornwall, and into 
 
90 STEAM USING; OR, STEAM ENGINE PE ACTIVE. 
 
 deep shafts, and with ample boilers and high pressure steam they became 
 famous as the "Cornish Pumping Engines." With the quick admission 
 of high pressure steam a sudden pull was exerted on the pump rods, 
 which, being constructed of wood and of long length, readily absorbed 
 this jerk and began to rise. An early cut-off allowed the rapidly falling 
 steam pressure in the cylinder to be helped out by the inertia of the 
 weight lifted, which came slowly to rest, and then reacted upon the 
 column of water, commencing gradually, conditions very favorable for 
 the pumping part of the work. The boilers used gave very high evapora- 
 tion by reason of the very moderate manner in which they were worked, 
 and the duty, or number of foot pounds of water raised per pound, or per 
 bushel of coal, was also, usually very high. In these mines systematic 
 record was kept and published monthly, and competition was thus 
 induced among the men in charge of the engines. High pressure steam 
 and high expansion here received its first practical confirmation. 
 
 When, however, these engines were applied to pumping water for 
 water works, it was found that without the elasticity of the long and 
 heavy pump rods used in the mines, the pump was apt to jump, if high 
 pressure steam was used; that is, the plunger would rise without the pump 
 filling and a very hard shock was the result. In consequence the use of 
 high pressure steam, and high expansion, was abandoned in such engines 
 built for waterworks purposes in the United States, and it is safe to say 
 that for such purposes no more of this class of engine will ever be built in 
 this country. In no case where used for water works has any such duty 
 been reached as was obtained by these engines in the mines. The indicator 
 diagrams for this class of engines are to be placed one above the other; any 
 difference between the exhaust line of the steam end and the admission 
 line of the exhaust end being so much lost pressure and work. 
 
 In the class of Cornish engines, introduced by Captain Bull, the beam 
 was dispensed with and the cylinder placed directly over the pump, the 
 steam being introduced at the lower and the exhaust taken from the 
 upper end. A pair of such engines are used at the Kiver Pumping Station 
 of the St. Louis Water Works, but, as suggested above, with steam at a 
 low pressure. The diagrams given herewith show the limited expansion 
 possible. 
 
 The other class of single acting engines was introduced in order to at- 
 tain a higher speed of rotation than had been found convenient in the or- 
 dinary double acting engines, where the inertia of the rods and the change 
 from thrust to tension brings first one side of the box and then the other 
 into bearing. Unless this change is a gradual one it is accompanied by a 
 shock more or less disastrous to the machine. By keeping, say, a thrust 
 continually on the rods the boxes are always in bearing on one side and no 
 such shock occurs. The steam during admission and expansion causes 
 a pressure in one direction on the piston which is not changed during 
 exhaust and compression. The irregularity of action may be remedied by 
 a heavy fly-wheel or by the use of one or more cylinders. As examples: 
 for two cylinders engines we select the Westinghouse; for engines with 
 
OJV THE INDICATOR, ETC. 
 
 91 
 
 three cylinders, Brotherhood's, of London; and with six cylinders, the 
 Colt Disc Engine, from West's and Darkin's patents, manufactured by the 
 Colt's Fire Arms Company, of Hartford, Conn. These machines have, at 
 the time of writing, been used more particularly with electric apparatus, 
 and in small boats, but have attained much popularity in other directions. 
 The exhaust is taken into the chamber where it lubricates the main 
 
 DIAGRAM FROM CORNISH ENGINE, No. 1, Low SERVICE, 
 ST. Louis, Mo., WATER WORKS. 
 
 56" X 138" X 9 double strokes per minute. 
 
 30-i 
 
 20- 
 
 10- 
 
 DIAGRAM FROM ENGINE No. 1, HIGH SERVICE, ST. Louis, Mo., 
 WATER WORKS. 
 
 85" X 120" X 11% revolutions per minute. I. H. P. 705. 
 
92 STEAM USING; OK, STEAM ENGINE PRACTICE. 
 
 THE WESTINGHOUSE ENGINE. 
 
 
 FRONT VIEW. 160 H. P. 
 
VARIETIES OF ENGINES. 
 
 THE WESTINGHOUSE ENGINE. 
 
 REAR VIEW. 160 H. P. (One fly-wheel removed.) 
 
94 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 THE WESTINGHOUSE ENGINE. 
 SECTION THROUGH SHAFT. 
 
 9 
 
 A A, Cylinders. B, Valve Chamber. 6', Bed, or Crank Case. 1) 1), Pistons. F F, 
 Connecting Rods. G G, Cranks. H H, Crank Shaft. /, Eccentric. ,7, Valve Guide. 
 K, Centre Bearing. M, Steam Connection. N, Exhaust Connection. 7?, Oil Pipe. V, 
 Valve. W, Wiper. Y, Fly-Wheel. Z, Pulley, a a, Cylinder Heads, b b, Steel Wrist 
 Pin. c, Crank Case Head, d d, Crank Shaft bearings, d' Cover, e, Oil Passage. //, Oil 
 Cups. Q g, Bolts, h, Bonnet, j j, Spider Heads, k k, Rings. I, Hollow Valve Bolt. 
 , Syphon Overflow, o, Hole in Funnel Head, r, Eccentric Rod. 1 1, Collar Washers. 
 ?', Lead Washer, x x. Bobs on Crank. 
 
96 
 
 STEAM USING-: OR, STEAM ENGINE PRACTICE. 
 
 a E 
 
 53 ^ 
 
 h 
 
 H fl 
 
 *| 
 
 u1 
 
 
 1 
 
 ^ 
 
 03 E* 
 
 
 1 
 o 
 
 : ? 
 
 
 a 
 
 S ^ 
 
 
 
 'a . 
 
 
 
 QQ fQ 
 
 w 
 
 OQ 
 
 O 
 
 :NE ENGIN 
 
 s* 
 
 ol 
 
 s 
 
 *3 
 
 o 
 
 V '_g 
 
 n 
 
 L> O 
 
 S3 
 
 s 
 
 o ^ 
 
 s 
 
 1 
 
 g" 
 
 
 es 
 > 
 
 *! 
 
 
 ^ 
 
 ^ (2 
 
 
 
 . a 
 
 
 
 QJ S 
 
 
 2 
 
 > s 
 ga 
 
 
 5 
 
 . . a, 
 
 
 o 
 
 ^ a 
 
 
 0) 
 
 
 
 > 
 
 I-H 
 
 9 a 
 
 sq 
 
 O 02 
 
 "II 
 
 s? 
 
i : i /,'//; 77 /-:.s OF EX(; i .v/-;>. 
 
 97 
 
STEAM USING; OR, STEAM EN&IXE PRACTICE. 
 
 THE BROTHERHOOD THREE-CYLINDER ENGINE. 
 
 PERSPECTIVE ELEVATION. 
 
 bearing with the wet steam and the oil introduced by the steam pipe. The 
 Brotherhood engines are also used with compressed air for driving tor- 
 pedoes, and we have seen one, driven by a Westinghouse Air Compressor 
 used for drilling in a locomotive repair shop. 
 
 It appears to us that these engines are well adapted for constant work, 
 but for intermittent use the temptation to run them without cleaning must 
 be so great as to render them liable to a rapid deterioration. 
 
 When indicator diagrams from large, slow moving engines are exam- 
 ined, and the weight of steam present at any two points of the stroke, such 
 as at cut-off, and release, is calculated by the aid of the table of the "Pro- 
 perties of Steam," and the volume, pressure, and density thereof, we shall 
 rarely find any kind of agreement in the two results. And, if there has 
 been also a careful measurement of the quantity of feed water consumed 
 per stroke, the amount will be found to be much in excess of that given by 
 
VARIETIX8 OF 
 
 99 
 
 THE COLT DISC ENGINE. 
 West's and Parkin's Patents. 
 
 PERSPECTIVE ELEVATION. 
 
 computation. This difference, in an engine with a tight piston, is only to 
 be accounted for by the action of the metal of the cylinder which transfers 
 heat to and from the within contained steam. This has already been ex- 
 plained in Chapter III. We shall consider the effect of this action upon 
 the indicator diagram, and the loss of work which occurs compared with 
 that which should be done in a non-conducting cylinder; or, in other 
 words, the consequent increase in the quantity of steam used for a given 
 work shown. 
 
 Let us take for example the data of the experiment conducted by Mr. 
 J. W. Hill at the Miller's Exhibition, held at Cincinnati. The experiment 
 from which we obtain the following was made upon one of three Corliss 
 engines, built by different makers, but all having the same general dimen- 
 sions, viz: 
 
100 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 THE COLT DISC ENGINE. 
 
 LONGITUDINAL SECTION. 
 
 CROSS SECTION, SHOWING CIRCULAR VALVE, PORTS, ETC. 
 
VAltTETIE* OF EX'r/XES. 
 THE COLT DISC ENGINE. 
 
 101 
 
 CROSS SECTION, SHOWING INTERIOR OF ENGINE, STEAM PORTS AND 
 EXHAUST PASSAGES 
 
 The main body of the engine consists of one casting, containing six cylinders, 
 arranged in a circle, and parallel with one another. The pistons A take the form of a 
 hollow plunger, one end terminating in a blunt cone which bears continuously against 
 the periphery of the disc B. They are single acting, being subject to steam pressure 
 upon the flat end only. Steam is admitted successively to the six cylinders from the 
 steam chest C, three pistons being constantly in action at different points of the stroke, 
 thereby imparting a uniform rolling motion to the conical disc B, which is steadied at 
 its center by the ball and socket joint D, and rolls upon the conical surface of the back 
 plate , which receives the full thrust of the pistons, and protects the ball and socket 
 joint 1) from strain. The crank pin F is securely fixed in the centre of the conical disc 
 /, the rolling motion of the disc causing the pin to describe a circle, and by means of 
 the crank G, imparting a rotary motion to the shaft H. The shaft Jf passes through the 
 centre of the steam chest and carries an eccentric giving motion to the circular valve K. 
 The valve K is a flat circular ring which slides steam tight but perfectly freely between 
 the port face and a balance plate. The steam is admitted to and fills the annular space 
 f ' left in the steam chest outside the circumference of the valve ring K, the eccentric 
 motion of which alternately opens and closes all the steam ports, successively admitting 
 steam to the cylinders, from which it again escapes to the exhaust chamber M formed 
 by the inside of the valve ring, and thence through openings into the body of the engine, 
 and is finally discharged by the exhaust pipe 3", 
 
102 
 
 STEAM VSIXG; OJt, STEAM ENGINE PRACTICE. 
 
ON THE INDICA TOE, ETC. 1O3 
 
 Cylinder, 18" x 48"; number of revolutions, 75.4; steam at 80 pounds 
 pressure, cutting off at 13 per cent, of the stroke for the Allis Engine: cut- 
 off pressure, 84 pounds; piston area = 1.7 square feet; piston displace- 
 ment = 6.8 cubic feet; 
 
 13 per cent, of 6.8 = 0.884 cubic feet. 
 
 Clearance volume = 0.2 cubic feet; volume occupied by steam at cut-off, 
 1.084 cubic feet; density of steam at 84 pounds = 0.231; 
 
 0.231 x 1.084 = weight of steam at cut-off = 0.250 pounds; 
 weight of feed water per hour = 3422.6 pounds; weight of feed water per 
 stroke = 0.378 pounds: 
 
 0.378 0.250 = 0.128 pounds condensed at cut-off; 
 
 or, 34 per cent, of all the fluid present is water; or, the water present 
 equals in weight 50 per cent, of the steam. 
 
 Real expansion = -~- = 6.4. 
 
 Now taking the length of the diagram to represent 6.8 cubic feet, and 
 adding the clearance = 0.2 cubic feet, or, say 3 per cent, of the stroke, we 
 set off at the cut-off pressure a volume 50 per cent, greater than the cut-off 
 
 loor 
 
 DlAGEAM FROM REYNOLD'S CORLISS ENGINE, 
 Cylinders 18" X 48" X 75 revolutions per minute, at Millers' Exhibition, Cincinnati, O. 
 
 volume; and if from this point we draw a curve of expansion for steam in a 
 non-conducting cylinder till we reach the pressure of release, and thence 
 draw a vertical line to the back pressure, we have the diagram which the 
 water boiled should have given per stroke in a non-conducting cylinder; 
 and the difference between the area of this diagram and the real diagram 
 represents the loss which has taken place by the action of the sides of the 
 cylinder. Measuring the area of the diagram by the planimeter we have 
 4.40 square inches, and for the area of the new diagram, 5.91 square inches, 
 which makes the latter 0.34 per cent, larger. The indicated horse-power 
 
104 STEAM USING; Oil, STEAM ENGINE PRACTICE. 
 
 is 152.7, which in a non-conducting cylinder would be nearly 204, so that 
 a loss of 50 indicated horse-power, or, say 25 per cent., is caused by the 
 transfer of heat in the sides of the cylinder. 
 
 This experiment is a good one by which to examine other matters, so 
 that we may see what can be obtained from an indicator card. 
 
 In the above experiment the quantity and rise of temperature of the 
 injection water were measured, and the steam supplied was found to be 
 dry by calorimeter tests. We abstract from the report of the trial the 
 following: 
 
 Dry steam supplied per hour for ten hours .......... 3422.6 pounds. 
 
 Dry steam per stroke .................................. 0.3782 pounds. 
 
 The injection water = 30.88 times the feed and rose from 72 to 102 
 Fahr., which gives 0.3782 x 30.88 x 30 = 351 heat units rejected per stroke. 
 Mean indicated pressure ...................... 34.26 pounds. 
 
 Mean total pressure ........................... 40.5 pounds. 
 
 Area of piston ................................ 253 square inches. 
 
 Total work per stroke = _ 4 = 53 units. 
 
 Heat per stroke, feed at 32 = 0.3782 x 1181 .................. 446 units. 
 
 Deduct for feed at 102 32 = 0.3782 x 70 .................. 26 " 
 
 Heat given to engine per stroke ............................ 420 " 
 
 Heat expended in work per stroke ............ . .......... 53 
 
 Heat in external radiation, say .......................... 4 57 
 
 363 " 
 Heat found in rise in injection water ....................... 351 
 
 12 
 
 12 
 Error in measurement = , or, say 3 per cent. 
 
 This error in the experiment probably arose in the measurement of the 
 injection water, and perhaps, from not obtaining the average of the hot 
 well accurately. 
 
 At cut-off we observed that of the amount of 0.378 pound of feed 
 water evaporated per stroke, 0.25 pound was steam, while the balance, 
 equal to 0.1282 pound, had condensed. 
 
 Units. 
 
 At this time the internal heat of the steam present = 1100 x 0.25 = 275 
 
 water " =...298X0.1282= 38 
 275 38 = 313, or, 
 
 Heat in fluid present 313 
 
 Heat received from boiler as given above 446 
 
 Heat absorbed by cylinder = 446 313 = 133 
 
OX THE INDICATOR, ETC. 1O5 
 
 At the end of the stroke the volume occupied is 7 cubic feet, with a 
 total pressure of 12 pounds; density = 0.031: 
 
 Weight of steam present, 7 x 0.031 0.217 pounds. 
 
 Weight of water present, 0.3782 0.217 0.1612 pound. 
 
 Internal heat in steam present, 0.217 x 1072 . . .233 units. 
 
 Heat in water present, 0.1612 x 170 27 " 
 
 Heat in fluid present 260 " 
 
 Heat used in expansion 36 " 
 
 u 
 
 Heat accounted for 296 " 
 
 Heat in fluid at cut-off 313 " 
 
 Heat added to iron during expansion, 313 296 17 " 
 
 Heat in iron at end of stroke, 133 4-17 150 " 
 
 And this latter amount is necessarily thrown away during the exhaust, 
 from the iron of the cylinder into the condenser, and is the cooling effect 
 of the latter upon the former. 
 
 The computation at the end of the stroke is never so satisfactory as at 
 cut-off, because the changes of volume are much greater for small 
 changes of pressure with low pressure steam than with high. Suppose 
 for instance the terminal pressure be taken as 2 pounds below the atmos- 
 phere instead of 3. The density will equal 0.034, and the weight of steam 
 0.238 pounds, while its internal heat will equal 0.238 x 1072 = 255 units. 
 
 94- 
 
 The water will equal 0.1402 pounds and its heat = . 
 
 279 
 
 This will make 19 units more accounted for in the fluid present, and 
 will give 131 units in the iron at the end of the stroke, instead of 150, by 
 reason of the cooling effect of the condenser. If the agreement between 
 the heat rejected and that given from the boiler had been closer, we 
 should have found a check upon this cooling effect of the condenser, as 
 follows: 
 
 Heat found in condenser. 351 units. 
 
 Heat given by back pressure work 7 
 
 Fluid heat at end of stroke, 260 or 279 units, according to the terminal 
 pressure taken. But this is too large, as the temperature of the water in 
 the condenser is 102 and not 32, and we have 0.3782 x 70 = 26 units to de- 
 duct as before, which leaves 234 or 253 units received into the condenser 
 from the fluid; or from the fluid and work of exhaust 241 or 260 units. 
 Deducting these amounts from 351 units, we have 110 or 91 units as the heat 
 received from the iron; but if we had found 13 units more in the condenser 
 the values would be 123 and 104 units, thus rendering the lower value of 
 terminal pressure the more probable. A very slight amount of water in the 
 steam supplied would account for much of the inconsistency of these 
 results. We shall give in Chapter V examples of such computations by 
 M. O. Hallauer, showing more consistent results. 
 
 With the use of high pressure steam, and high expansion, as employed 
 by Trevithick, in Cornwall, the shock and change of pressure in the cyl- 
 
1O6 
 
 STEAM USING: OE, 8TEAM ENGINE PRACTICE. 
 
 inder between the beginning and end of the stroke, of course, became 
 considerable, and it was suggested by Hornblower, in England, and long 
 afterwards by Woolf, in Germany, that the steam should be first introduced 
 into a small cylinder, whereby the strain, produced on the connections by 
 the sudden influx of high pressure steam, might be reduced, and that after 
 acting for more or less of the stroke at boiler pressure, the steam should 
 be put in communication with a larger cylinder in which the expansion 
 should be completed. But while pushing the large piston forward there 
 is a tendency to retard the motion of the small one. The following dia- 
 gram, Fig. 40, will explain the indicated work and the total work: 
 
 FIG. 40. 
 
 Let a b represent the volume of the small cylinder, b c the clearance 
 between the small and large cylinder and a b + c d the volume of the 
 large cylinder. 
 
 The area a g f e b is the total work done in the small cylinder and the 
 area a n i h j k d is the total work done in the large cylinder. The indi- 
 cated work of the small cylinder is n g f e and the back pressure work is a 
 neb. The indicated work of the large cylinder is n i h j k n, while the 
 back pressure work is a n k d. We see that the work a nm b has been 
 
VARIETIES OF XX'rlXJ-s. 1O7 
 
 counted twice, which is once too many; and we also see that there are two 
 ways of representing the portion n m i, which may be considered as part 
 of the back work of the small cylinder while it is forward work for the 
 large cylinder. Or we take it as part of the forward work of the small 
 cylinder, giving a g fe b as the forward work, while a n m b is the back 
 pressure work in that cylinder, leaving n gfe m as the indicated work; 
 while, for the large cylinder we have chjd for the total, I hj k for the 
 indicated and elk d for the back pressure work. 
 
 If the receiver between the cylinders be small, the pressure at the be- 
 ginning of the stroke of the large cylinder is nearly equal to that at the 
 end of the stroke of the small cylinder. With a large receiver the press- 
 ure is apt to be much less, while it can be raised by a cut- off on the large 
 cylinder. 
 
 The amount of clearance volume between the cylinders affects the 
 form of the curves e n hj and i n; and when very large, regarding c d, the 
 work done in each cylinder is measured by itself without any deductions 
 and the diagrams are placed one under the other, as in figure 41, where: 
 a b c d e a is the high pressure diagram, and a f g h k the low pressure dia- 
 gram. The enclosed areas are the indicated work; the area j k hi the 
 back pressure work on the large piston, and a e m j that of the small 
 piston. 
 
 When the question of indicated power only is under discussion the 
 measurement of the diagram by the ordinates or by the planimeter is suf- 
 ficient. Each piston by itself is taken and the work done on it found and 
 added. Much misconception of this subject appears to have prevailed, 
 and one modern writer claimed, in 1883, a generalization and geomet- 
 rical construction which is given in Rankine's Steam Engine, in the first 
 edition. 
 
 Many forms of compound engines are to be met with, some of which 
 may be briefly described as follows: 
 
 1 . Engine with two cylinders of equal stroke acting on one crank pin, 
 the steam from one end of the small cylinder passing to the farther end 
 of the large one. The cylinders are usually in line, or "tandem." 
 
 2. Engine with two unequal cylinders attached to the same end of a 
 beam, the steam passing from one end of the small cylinder to the further 
 end of the large one. This form was the original one introduced by Woolf . 
 
 3. Engine with two cylinders of the same diameter, set side by side, 
 with two cranks at 180, the steam from the small cylinder passing across 
 to the near end of the large one. 
 
 4. Engine with two cylinders attached to opposite ends of a beam, the 
 steam from the top of the small cylinder passing to the top of the large. 
 This form has been largely used by McNaught and lately with horizontal 
 cylinders. 
 
 5. Engine with two cylinders attached to two cranks at an angle of 
 90 with each other and a more or less intermediate receiver. With verti- 
 cal cylinders they have been more used than any other form for marine 
 engines. 
 
1O8 STEAM rslXf,',- (HI, STEAM ENGINE PRACTICE. 
 
 6. Engines with three cylinders attached to cranks at 120 with each 
 other, one of them being used for high pressure and the two others, of the 
 same, or greater diameter, for low pressure. This class has been success- 
 fully used on the very largest marine engines. Those of the steamships 
 Arizona, Alaska, Servia, Aurania and Oregon may be instanced. 
 
 7. Engine with four equal cylinders attached to cranks at 45, using 
 one as high pressure and the other three as low pressure. 
 
 8. Two pairs of tandem engines driving two cranks at 90. 
 
 9. Three pairs of tandem engines driving three cranks at 120. 
 
 10. Three unequal cylinders driving three cranks at 120, the steam 
 passing through all three cylinders. 
 
 11. Two pairs of tandem engines driving two cranks at 90, the steam 
 passing through all four cylinders and two interheaters. 
 
 The use of receivers of comparatively large volume between two cyl- 
 inders was first suggested by Ernest Woolf, and the great progress made 
 in marine engines by its use, under the name of compound engines, has 
 been mainly due to its adoption by the engineers of Glasgow. 
 
 For marine service the compound engine with high pressure steam 
 has developed the best practical results. The most usual type adopted is 
 that with two cylinders, the smaller, or high pressure, and the larger or 
 low pressure, being coupled by cranks at 90. The most usual type of 
 screw engine being vertical, we find the small cylinder occasionally placed 
 on top of, and in line with, the large cylinder, both pistons working on 
 the same rod. Two or three such engines are used with cranks at 90 with 
 two cylinders, and at 120 when three cylinders are employed. 
 
 When the use of two cylinders would require the low pressure cylin- 
 der to be excessively large, two low pressure cylinders are used in connec- 
 tion with one high pressure cylinder, coupled to cranks, generally at 120. 
 
 The use of three cylinders of different sizes has been tried with steam at 
 125 and 140 pounds pressure, the higher pressure was adopted on the Steamer 
 Propontis. Although the boilers in this instance had to be abandoned, the 
 results were very good while the high pressure could be maintained. 
 
 The following table prepared with great care by Mr. Marshall gives 
 the results of the progress in marine engine practice. While the economy 
 in the use of high pressure steam and the compound engine, taken to- 
 gether, is evident, it must be confessed that it is not easy to separate the 
 effects of the two causes; and while the mechanical advantages of com- 
 pounding for marine engines of large size, for which it is especially useful, 
 are well known, yet it may be stated that with steam of moderate pressure 
 no advantage in economy has been found when the compound engine is 
 compared with single cylinder engines of good design and construction, 
 and working steam at the same pressure. 
 
 It is also readily conceded that higher rates of expansion may be used 
 with the compound engine than with single cylinder engines, using steam 
 at the same pressure; but it may be doubted whether any practical econ- 
 omy has resulted therefrom, as may be seen from an examination of the 
 table of engine trials and the paper by Hallauer in Chapter V. 
 
VARIETIES UF 
 
 109 
 
 THE AVERAGE CONSUMPTION OF COAL PEE I. H. P. PER HOUR, 
 
 By Steamships using Compound Engines in long sea voyages. A table reduced from 
 one accompanying a paper read before the Institution of Mechanical Engineers, Lon- 
 don, in 1881, by F. C. Marshall. 
 
 s 
 
 I 
 
 Cylinders. 
 
 ^ Receivers at 
 
 o 90. 
 
 6 
 
 | 
 
 Piston g 
 Speed. g 
 
 s 
 
 o 
 
 lifi i 
 
 cc *< 4 
 
 Ms 1 
 
 s -as o 
 
 .5 < 
 
 otal Heating 
 Surface. 
 
 rate Area. 
 
 idicated Horse- 
 Power. 
 
 oal in Pounds 
 per I. H. P. per 
 hour. 
 
 X. 
 
 a 
 
 Si 
 
 H 
 
 -i - 
 a^ 
 
 fe 
 
 o 
 
 2 
 
 H 
 
 i i o 
 
 H 
 
 
 
 
 
 
 
 
 Ft. sq. ft. 
 
 ft. in. 
 
 -. 
 
 sq. ft. sq ft. 
 
 
 
 1 34&61X45 
 
 450 2466 
 
 15 3 70 5 
 
 > 4216 140 
 
 900 1.5 
 
 14.00 
 
 '1 4 '2 A: SOX 48 
 
 552 
 
 72 5 c i 
 
 > 6000 
 
 1881 l.t? 
 
 32 25 
 
 3 36*70X48 
 
 400 2400 
 
 17 90 5 
 
 ! 4440 160 
 
 1200 1.63 
 
 2l'" 
 
 4 46*87X57 
 
 484 5000 
 
 19 80 ' 
 
 t 7803 25U 
 
 2200 1.66 
 
 40 
 
 5 22*44X30 
 
 360 705 
 
 11 100 ] 
 
 Lj 1402 49.5 
 
 920 1.67 
 
 7.5 
 
 6 50*86X54 
 
 540 4865 
 
 17 6 72 4 
 
 t 7722 273 
 
 2673 1.67 
 
 48 
 
 7 35 & 70X48 
 
 424 2000 
 
 17 90 J 
 
 5 4774 150 
 
 1200 1.69 
 
 21.5 
 
 8 54*94X60 
 
 530 7420 
 
 18 3% 75 ( 
 
 \ 10839 313 
 
 2207 1.70 
 
 40.3 
 
 9 54*94 X 60 
 
 486 7422 
 
 18 82.5 ( 
 
 5 11340 324 
 
 1801 1 70 
 
 32.8 
 
 in Hn*5SX39 
 
 400 1513 
 
 14 2 80 ! 
 
 !j 2608 69.7 
 
 650 1.72 
 
 12 
 
 11 29& 56X33 
 
 350 1250 
 
 13 3 70 J 
 
 2 2379 66 
 
 500 1.76 
 
 9.5 
 
 12 34*66X42 
 
 406 1700 
 
 15 6 80 
 
 >; 3474 107.2 
 
 875 1.76 
 
 16.5 
 
 13 36*68X42 
 
 434 1821 
 
 16 3 77 i 
 
 \- 3714 110 
 
 854 1.76 
 
 16.75 
 
 14 54*97X60 
 
 480 7427 
 
 18 10% 70 ( 
 
 5 11045 329 
 
 2000 1.80 
 
 38.6 
 
 ir> 51*88X60 
 
 590 5000 
 
 17 6 75 
 
 [' 9248 332.5 
 
 2745 1.83 
 
 54. 
 
 16 28*53X38 
 
 380 1560 
 
 14 n 75 ! 
 
 > 2433 78 
 
 560 1.84 
 
 11 
 
 17 50&86X54 
 
 540 5500 
 
 17 6 70 
 
 I 7525 273 
 
 2422 1.85 
 
 48 
 
 18 38*70X48 
 
 416 2600 
 
 17 9 80 
 
 4864 65 
 
 1160 1.85 
 
 23 
 
 19 35&70X48 
 
 408 2005 
 
 17 90 5 
 
 J 4826 150 
 
 1099 1.87 
 
 22 
 
 2035*70X48 
 
 440 2000 
 
 20 9 90 
 
 J 4396 136 
 
 1135 1.89 
 
 23 
 
 21 34 1 &64X42 
 
 560 1647 
 
 15 4 80 
 
 > 2950 106 
 
 880 1.90 
 
 18 
 
 22 48&84X60 
 
 550 4468 
 
 19 70 I 
 
 { 8200 340 
 
 2300 1.90 
 
 47 
 
 23 50*86X54 
 
 510 4842 
 
 17 9 70 ( 
 
 5 9839 310 
 
 2213 1.90 
 
 43.5 
 
 24 54&94X60 
 
 441 7420 
 
 17 9 70 i 
 
 *, 11750 312 
 
 2400 1.90 
 
 48.9 
 
 25 56&97X54 
 
 504 5000 
 
 18 6 7d 
 
 ; 8215 292 
 
 2500 1.93 
 
 51.9 
 
 26 30*60X36 
 
 372 1600 
 
 13 90 
 
 J 2753 115 
 
 600 1.94 
 
 12.5 
 
 27 36*70X45 
 
 560 2900 
 
 13 75 
 
 I 4622 166 
 
 1600 2.00 
 
 32 
 
 28 36*64X36 
 
 450 2059 
 
 13 70 
 
 I 2854 106 
 
 1020 2.12 
 
 22 
 
 29 36&68X42 
 
 530 2500 
 
 13 70 
 
 > 3462 130 
 
 1250 2.25 
 
 29 
 
 3036*67X42 
 
 530 2400 
 
 13 70 , 
 
 I 3451 129 
 
 1230 2.25 
 
 28 
 
 
 mean 467 
 
 mean 77.4 
 
 
 mean 1.828 
 
 
 Tandem 
 
 
 
 
 
 
 31 48*83 X60d'ble. 
 
 523 9000 
 
 23 6 70 
 
 3 19104 624 
 
 4900 1.77 
 
 93 
 
 32 26*58X45 Sin. 
 
 444 1700 
 
 15 80 
 
 2 3160 99 
 
 820 1-90 
 
 16.75 
 
 33 27*56X52 " 
 
 395 1730 
 
 15 9 80 
 
 I 3244 102 
 
 730 1.92 
 
 15.1 
 
 34 27*56X52 " 
 
 412 1651 
 
 15 9 75 - 
 
 I 3570 102 
 
 771 1.93 
 
 16 
 
 :sr> 2s*60X54d'ble. 
 
 504 4H K) 
 
 20 90 
 
 3 7400 300 
 
 1900 1.96 
 
 40 
 
 36 28*60X54 " 
 
 522 4100 
 
 20 80 
 
 J 7413 302 
 
 1850 2.01 
 
 40 
 
 37 16*34X30 " 
 
 360 768 
 
 12 6 68 
 
 L 1350 47.2 
 
 270 2.25 
 
 7 
 
 38 26*52X42 " 
 
 336 2400 
 
 , 17 3 70 
 
 I 3650 132 
 
 900 2.47 
 
 24 
 
 
 mean 437 
 
 mean 76. 7 
 
 
 mean 2.026 
 
 
 39 60& | JJj j- X66 
 
 605 
 
 21 90 
 
 1 19500 780 
 
 6300 1.63 
 
 110 
 
110 STEAM USING; OK, STEAM ENGINE PRACTICE. 
 
 While Mr. Marshall's table is of great value, it must not be forgotten 
 that the gain and rapid advance is due to the increase of pressure and the 
 higher expansion used; and must not be confounded with the great gain 
 which resulted from the introduction of the compound engine following 
 the surface condenser. Of this a portion was a mechanical one, due to 
 smoother action, and the remainder, the saving of the heat lost by con- 
 stant blowing off when salt water was fed to the boilers. 
 
 Mr. Marshall compares with a paper read by Mr., now, Sir F. J. Bram- 
 well, in 1872, in which Mr. Bramwell gave particulars of twenty-eight 
 steamships. 
 
 The average consumption of coal per indicated horse-power per hour, 
 from nineteen of these vessels, was 2.11 pounds, the steam pressure rang- 
 ing from 45 to 65 pounds above the atmosphere. The steam averaged, say. 
 52.5 pounds, and the piston speed was 376 feet per minute. 
 
 We see that the average steam pressure, was, in 1881, 77.4 pounds, the 
 average piston speed 467 feet, while the coal consumed per indicated 
 horse-power per hour, was 1.83 pounds, or 13.4 per cent, less; and the boiler 
 surface is also less per horse-power. This gain is shown to be that theo- 
 retically due to the higher initial pressure, the same terminal and back 
 pressure being assumed, the number of expansions being in the one case 
 5.15, and in the other 7.05. 
 
 Now with steam of 60 pounds total working pressure, and with 4.57 ex- 
 pansions, the theoretical steam is 21.93 pounds per indicated horse -power 
 per hour; for 6 expansions 19.09 pounds, and for 8 expansions 18.8 pounds. 
 The difference between 21.9 and 18.8 = 3.1, which shows the gain by increas- 
 ing from 4.57 to 8 expansions. 
 
 The gain from 6 to 8 expansions, or from 19.1 to 18.8 pounds, of steam 
 is only 0.3, or, for 7 expansions say 18.9 pounds; so that we find that most 
 of the gain must come from the increase in pressure and very little from 
 the increase in expansion. 
 
 The value of jacketting and compounding is still on open question. 
 For slow moving engines, both appear to add to the economy; but for 
 high speed engines little gain is observed. In Chapter V, in the Alsatian 
 experiments, we shall find a very strong argument in favor of the single 
 cylinder unjacketed engine using superheated steam. In comparing the 
 economy of the engines of the "Leila," using compound unjacketed cylin- 
 ders, with those of the "Siesta," we find for the cases of maximum power 
 that little was added by the superheating but the smaller engine was 
 driven the harder. Looking at the Miller's Exhibition engine we find 
 that with the cylinder of the same size it use^d but little more steam per net 
 horse -power, steam of the same pressure. 
 
 We present here two diagrams from Porter-Allen engines. One 
 taken at a trial using superheated steam, made at the American Institute 
 Fair in New York several years ago, and interesting mainly as showing an 
 expansion curve similar to that of steam in a non-conducting cylinder. 
 The other is from a compound engine, jacketed, and with interheater. 
 The diagrams then taken are given, and also the combined diagram; the 
 
o.V THE INDICATOR, ETC. Ill 
 
 latter shows two curves of expansion for a non-conducting cylinder, the 
 inner one for the steam passing through the engine and the outer one for 
 all the steam used. A comparison of the enclosed areas with that of the 
 outer curve shows the loss. We also give illustrations, and detailed draw- 
 ings of some of the principal features of the Porter- Allen engine which re- 
 
 DlAGBAMS FROM PORTER-Al/LBN ENGINE. 
 
 16" X 30", 125 revolutions per minute. 
 
 Scale 
 
 Record, 1 horse-power per 25.8 lt>s. of water per hour 
 
 quire no explanation, see pages 167 to 187. They are given as an example 
 of the most skillful design which has been attended with the best work- 
 manship ever used in engine building. 
 
 We also give illustrations and many details of the Reynolds Corliss 
 
112 STEAM USING; OR, 8TEAM ENGINE PRACTICE. 
 
 Scale ,'G. 
 
 TO- 
 
 .TO 
 
 40- 
 
 3C 
 
 10 
 
 DIAGRAM FROM PORTER-ALLEN COMPOUND ENGINE. 
 
 12" and 21" X 24", at 180 revolutions per minute. 
 
 COMBINED DIAGRAM FROM PORTER- ALLEN. 
 ENGINE. 
 
 Upper curve shows expansion for all steam used, 
 if in a non-conducting cylinder. Second curve 
 shows expansion for the steam passing through 
 the cylinders. 
 
 Cylinders 12" and 21" X 24" X 184 revolutions 
 per minute. 
 
 JL\Cul.Ft, 
 
ON THE INDICA TOR, ETC. 
 
 113 
 
 engine biiilt by Messrs. E. P. Allis & Co., of Milwaukee, Wis., which were 
 kindly furnished by Mr. Edwin Reynolds, and these require no further 
 explanation, see pages 153 to 162. 
 
 On pages 163 to 166 we also give some illustrations of details of the 
 Lambert ville, N. J., Iron Works, Automatic Cut-off Engine. 
 
 The class of engines with poppet valves we illustrate by some details 
 of the engines of the Mississippi River Steamboat, "Montana," pages 119 to 
 121, and an elevation of the High Service Engine, No. 1, of the St. Louis 
 Water works, page 114. 
 
 We illustrate herewith the action of river engines by diagrams taken 
 from the Steamboat "Phil. E. Chappel" and the "James Howard." A dia- 
 
 150- 
 
 100- 
 
 DlAGRAM FROM LARGE CYLINDER OP ENGINES OF RlVER 
 
 STEAMER "PHIL. E. CHAPPEL." 
 
 22 revolutions per minute. 
 
 Mean pressure 
 90 and 99 Ibs. 
 
 120 
 
 -100 
 
 80 
 
 60 
 
 40 
 
 -20 
 
 
 
 DIAGRAM FROM ENGINES OF RIVER STEAMER "JAS. HOWARD. 
 
 34&" X 120" X 11M" with condenser, 12M" without condenser. 
 1247 I. H. P., both engines with condenser. 
 1268 I. H. P., both engines without condenser. 
 
114 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 HIGH SERVICE PUMPING ENGINE, 
 
 No. I, St. Louis, Mo., Water Works. 
 
 Steam cylinder 85" X 120". 
 
LAWRENCE, MASS., WATER WORKS. 
 May 4, 1876. 
 
 INDICATOR DIAGRAMS, UPPER END. 
 
 High pressure cylinder, 18 in. diameter, 96 in. stroke. 
 Piston rod, 3.5 in. diameter. Volume, 23508 cu. in. = 13.6041 
 cu. ft. Clearance, 601 cu. in. = 2.56 per cent. 
 
 Low-pressure cylinder, 38 in^ diameter, 96 in. stroke. 
 Piston rod, 4 in. diameter. Volume, 107674 cu. in. = 62.1313 
 cu. ft. Clearance, 1135 cu. in. = 1.54 per cent. 
 
 Difference of volume of cylinders, 107674 23508 = 84,166 
 cu. in. 
 
 Clearance of low-pressure cylinder, and difference of 
 volume of cylinders, expressed in terms of volume of high- 
 pressure cylinder, as follows: 
 
 1135 -r- 23508 = .048281 = 4 83 per cent. 
 
 84166 -7- 23508 = 3.5803. 
 
 Isothermal curve, p * v. J 
 
 Adiabatic curve, calculated by Rankine's formula, p <*v . 
 
 Barometer corrected 30.099 = 14.78 Ibs. per square inch. 
 
 The difference between the two adiabatic curves shows 
 the amount of water present in the steam at cut-off 
 evaporated by heat derived from the 
 jacket; and the effect of the surface of 
 piston-rods, acting as a surface con- 
 denser, in condensing steam during ad- 
 mission, to be subsequently ree vapor - 
 ated. / 
 
 I axis of Isothermal curve. 
 
 A axis of Adiabatic curve. 
 
 ! ,f\y 
 
 _j j 1 j i ;_ . _ _ _ _ __4__-_.^_-_-_-4-'A|s^ 
 
 5 
 
 g 
 
 The indicated work done in smaller cylinder may be taken as including the portion 
 only fine shaded, in which case the work indicated in large cylinder includes from the 
 point A, the clearance being curved on the indicated work in small cylinder is to be 
 taken as including the portion with coarse shade, the clearance loss being vertical and 
 the work indicated in long cylinder is that fine shaded only. 
 
 NOTE. The Rankine formula for an adiabatic curve is p v = const., and not 
 
 p v = const., as given above. 
 
116 
 
 STEAM USING; OH, STEAM ENGINE PRACTICE. 
 
 DIAGRAMS FROM ENGINES OF S. S. "ARIZONA." 
 
 HIGH PRESSURE. 
 
 90-T 
 80- 
 
 10- 
 
 60- 
 
 oO- 
 40 - 
 
 ao - 
 20- 
 
 10- 
 
 LOW PRESSURE. 
 
 COMBINED DIAGRAM. 
 
 Boiler pressure 86 Ibs. 
 
 Cylinders, high pressure, 60" X 66" X 55 revolutions. 
 2 low " 90" X 66" X 55 
 
 INDICATED HORSE-POWER. 
 
 High pressure cylinder. 
 .FLow 
 A " " " 
 
 .2433 
 .1925 
 .1948 
 
 6306 
 
ON THE INDICATOR, ETC. 
 
 117 
 
 DIAGRAMS FROM ENGINES OF THE S. S. "ABERDEEN.' 
 
 70"X 54" 
 
 SINGLE DIAGRAMS. 
 
 140 e. /. 
 
 COMBINED DIAGBAM. 
 
118 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 gram is also given on page 91, from the St. Louis Water Works engine, 
 and there is drawn thereon an expansion curve for the same quantity of 
 steam in a non-conducting cylinder. 
 
 As illustrating the action of a non- receiver compound engine we give, 
 page 115, a combined diagram from the trial of the Lawrence Pumping 
 Engine, and have noted the matter of indicated work thereon, the point 
 being that is more than the. sum of the finely shaded areas. 
 
 Single and combined diagrams are given for the engines of the S. S. 
 "Arizona, "the two low pressure diagrams combining as one under the high 
 pressure diagram; and we also give single and combined diagrams from 
 the triple expansion engines of the S. S. "Aberdeen," see pages 116, 117. 
 
 As examples of the exteriors of modern marine engines we have re- 
 produced illustrations of the engines of the steamships "Grecian," the 
 "Parisian" and the "Aberdeen," see pages 125 to 129, which with the above 
 diagrams from the "Arizona" and "Aberdeen" are taken from "Engineer- 
 ing," of London. 
 
 We also give illustrations of a compound engine of the non-receiver 
 type, for the U. S. Lighthouse Steamer "Manzanita," pages 122 to 124, de- 
 signed and built by Mr. H. A. Kamsay, of the Vulcan Iron Works, Baltimore, 
 Md. The "throws" of the two cranks are placed immediately opposite each 
 other or 180 apart, and the steam after doing its work in the high pres- 
 sure cylinder, is released, and passes at once into the large cylinder when 
 the piston is found at the commencement of its stroke ready for the steam 
 to exert its force upon it. Hence the interposition of a receiver, which is 
 required when the cranks are placed at right angles to each other, is not 
 necessary. The only objection to the arrangement is the supposed greater 
 difficulty in handling the engine as both cranks are on their dead centres 
 at one time, but this is easily obviated by care on the part of the engineer. 
 The cylinders are 22 inches and 36 inches diameter, and the stroke of pis- 
 tons is 34 inches. In order to work the valves with one pair of eccentrics 
 the valve faces are placed at right angles to each other. This arrangement 
 simplifies the number of parts and renders them easily accessible for ad- 
 justment. There are no expansion or cut-off valves on either cylinder but 
 both which are of the ordinary locomotive type, are provided with suffi- 
 cient lap to enable them to suppress the steam at three -fourths of the 
 stroke of the pistons. By the proper arrangement of passages and valves, 
 the engines are arranged to work as simple condensing engines, compound 
 and high pressure or non-condensing, as may be desired. The surface 
 condenser which is of the usual rectangular cast-iron box shape forms the 
 frame for supporting the cylinders on one side, while the other side is 
 supported on wrought -iron columns. The air pump has a trunk plunger 
 and bucket actuated by wrought-iron levers connected to the cross-head 
 of the low-pressure cylinder. 
 
 The condenser is arranged to pass the condensing water, furnished by 
 an independent steam pump, making three turns through the condenser. 
 The circulating pump has a cylinder of 12 inches diameter and 15 inches 
 stroke. 
 
VARIETIES OF ENGINES. 
 
 119 
 
 MISSISSIPPI RIVER STEAMER "MONTANA." 
 Cylinder and Valves. 
 
 A 
 
 ill- 
 
 m 
 
 -- 
 
 
 ^sss 
 
 \ 
 
 - 
 
 w 
 
 
 
 
 ! 
 
 > 
 < 
 
 
 
 
 ; 
 
 
 
 
 & 
 
 
 
 
 tt 
 
 
 
 
 s 
 
 
 
 
 c 
 
 
 
 
 a 
 
 
 
 
 a 
 
 
 
 
 1 
 
 
 
 
 1 
 
 v^nm^" 
 
 4 
 1 
 
 
 
 ^T~ 
 
 k 
 1 
 
 
 
 i 
 
 
 TT 
 
 "a 1 tf 
 
120 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 DETAILS. 
 Scale, M inch = 1 foot. 
 
 Cutoff Cam. 
 
 MISSISSIPPI RIVER STEAMER "MONTANA.' 
 
VARIETIES OF ENGINES. 
 
 121 
 
ENGINES OF THE U. S. LIGHTHOUSE 
 STEAMER "MANZANITA." 
 
 ELEVATION. 
 
VARIETIES OF ENGINES. 
 
 123 
 
 ENGINES OF U. S. LIGHTHOUSE STEAMER " MANZANITA." 
 
 PLAN. 
 
ENGINES OF U. S. LIGHTHOUSE STEAMER " MANZANITA.' 
 
 ELEVATION. 
 
VARIETIES OF ENGINES. 
 
 125 
 
 TRIPLE-EXPANSION ENGINES OF S. S. "ABERDEEN. 1 
 
 PERSPECTIVE VIEW. 
 Cylinders 30", 45" and 70" X 54" stroke. 
 
COMPOUND ENGINES OF S. S. "GRECIAN." 
 
 SIDE ELEVATION. 
 Cylinders 48" and 84" X 54" stroke. 2000 horse power. 
 
VA EIETIES OF ENGINES. 
 
 127 
 
 COMPOUND ENGINES OF S. S. "GRECIAN. 
 
 END ELEVATION. 
 
128 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 THREE-CYLINDER COMPOUND ENGINE OF S. S. "PARISIAN.' 
 
 Cylinders 60", 85" and 85" X 60" stroke. 
 
VARIETIES OF ENGINES. 
 
 129 
 
 THREE-CYLINDER COMPOUND ENGINES OF S. S. "PARISIAN." 
 
13O STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 On pages 138 and 139 will be found illustrations of the engines of the 
 steam yacht "Leila," built by the Herreshoff Manufacturing Company, 
 Providence, B. I. 
 
 The Buckeye engine we illustrate in detail, see pages 141 to 152. It 
 appears at first sight to have only an expansion valve on the back of the 
 main slide, with the change in the cut-off given by the rotating weights 
 on the shaft turning the expansion eccentric round the shaft. But the 
 connection between the expansion eccentric and its slide is through a rock- 
 ing lever with equal arms, carried at its centre at the middle of a rocking 
 lever pivotted at the lower end. The effect of this, while throwing the 
 expansion slide to the other side of the main slide, is also to render the 
 movement of the expansion slide on the back of the main slide entirely 
 independent of the position of the latter; for the expansion slide is in 
 its central position on the main slide when the two working levers are 
 in line, which only requires the expansion eccentric to be in a given 
 place. The arrangement is therefore as if the expansion slide worked on 
 a fixed seat, dividing the steam chest, with the advantage of a very small 
 clearance. 
 
 In drawing the valve diagram we find from the distance circle for the 
 main valve, the points of admission, release and cushion; but we draw 
 either on the same, or a separate diagram, the distance circle for any posi- 
 tion of the expansion eccentric, and remembering the lap is negative, we 
 draw the lines for the opening and closing of the ports in the slide by the 
 expansion slide. 
 
 By turning these round from the place where the cut-off is made by 
 the main slide to any early limit desired, the range and throw, or range and 
 lap, may be determined. This presents no difficulty, and is simpler than 
 the case of the common engine in which the expansion eccentric is also 
 turned round the shaft. 
 
 On page 136 we give an illustration and description of the Steam 
 Engine Governor used by the Cummer Engine Company, of Cleve- 
 land, O. 
 
 Four port engines are the oldest type of vertical engines with poppet 
 valves, and the lifting is usually performed by cams. The first attempt at 
 expansion was made by Stevens. The rock shaft holding the cams was 
 moved by a rod from the eccentric; the end of the rock shaft carried an 
 arm with a pin thereon and the rod from the eccentrics, passing on this pin, 
 was slotted. By using a large eccentric, set much in advance, this rod ac- 
 quired considerable motion in either direction before the rock shaft was 
 put in motion, a.nd the opening of the valves was made much more rap- 
 idly. As the cams could move down without any effect after the valve had 
 seated, the closure was made earlier and also more promptly. A drop cut 
 off was afterwards devised by Sickles, and the steam valve stems were 
 lifted by clutches and released at any point desired. This gear could be 
 altered by hand while running. Pumping engine, No. 1, High Service, 
 St. Louis Water Works, has been given as a good example of this kind of 
 machine. 
 
VARIETIES OF ENGINES. 131 
 
 The credit of the Automatic cut-off, or expansion varied by the gov- 
 ernor, must be accorded to Geo. H. Corliss, of Providence, Khode Island. 
 He introduced a four port engine with twisting valves, adding various de- 
 vices in the way of connections and adjustments, until he secured a very 
 rapid opening of the steam valve, and rapid opening and closing of the 
 exhaust valves. The latter were permanently attached to, and moved by, 
 the eccentric, and only by moving the eccentric of the shaft and by chang- 
 ing the length of the rods can any change be produced in their movement. 
 The steam valves were lifted by clutches, which met with a releasing piece 
 at some point of the opening determined by the governor, the earlier the 
 faster; the mean effective pressure in the cylinder is thereby reduced, the 
 speed lowers and is again increased. By a delicate governor these vari- 
 ations may be made almost imperceptible. 
 
 The undoubted gain made by these engines in the use of steam, is due 
 to the fact that the initial pressure up to the cut off is kept nearly that of 
 the boiler, and that with the large ports and quick opening valves the 
 back pressure is reduced to the least possible amount. The piston 
 speed, number of revolutions, and steam pressure are all increased while 
 the clearance is very greatly reduced. And, in fact the gain is due to a 
 general fitness and skillful combination of things, all of which were 
 good in themselves, and which in no way disturbed each other, when 
 combined. 
 
 The closing of the steam valves after being tripped has been caused 
 by weights, by springs, by compressed air, by air of reduced pressure, and 
 by steam, as well as by combinations of these forces. The best results 
 have been given in this country by weight and vacuum tending to close 
 the valve, with an air dash pot to take up the shock. 
 
 At the expiration of the patents covering Mr. Corliss' improvements, 
 a number of builders commenced making engines of almost exactly the 
 same pattern. Of these imitations which contain very many improve- 
 ments, perhaps those built by Mr. Wm. A. Harris, of Providence, are best 
 known in the East, while those of Messrs. E. P. Allis and Co., of Milwaukee, 
 working from the designs and under the supervision of Mr. Edwin A. 
 Reynolds, have suited the requirements of the Western practice. 
 
 To take an example of a four- port engine with poppet valves we select 
 the western river Steamer "Montana," given on pages 119 to 121. The four 
 valves are moved by four cams from two rock shafts, driven by rods from 
 triangular cams on the shaft. The engines can be worked by hand or not, 
 at full stroke, backward or forward, and at a fixed cut-off never earlier 
 than one-half stroke while running forward. 
 
 A single cam on the shaft connected to the steam rock shaft is used for 
 the cut-off, while the exhaust side rockshaft is driven by the other cam. 
 To use full stroke this is disconnected and the two rockshafts hooked to 
 work together from the exhaust; while to reverse, the full stroke cam is 
 hooked to the rock arm driving the exhaust on the other side. The rapid 
 movements of the valves in opening and closing are seen from the dia- 
 grams taken from the steamer "Phil. E. Chappel." Two cards are given 
 
Ill 
 
 2 
 
 . 
 
 
 
 > 
 
 
 M 
 
 It] 
 
 > 
 
 * 
 
 W) 
 
 o 
 
 'i 
 
 3 
 
 s 
 
 (0 
 
 LJ 
 Ul 
 
 I 
 
 ! 
 
 
 
 TJ 
 
 _c 
 
 UJ 
 
 I 
 
 o" 
 
 h 
 
 "S 
 
 
 U) 
 
 
 c 
 o 
 
 
 '! 
 
 CD 
 
VARIETIES OF ENGINES. 133 
 
 from the "James Howard" shown with and without the use of the con- 
 denser. The condenser was opened to the cylinder after the bulk of the 
 steam had been exhausted into the air, say when the piston had moved 
 one- tenth of the stroke backwards, the increase of pressure found was 
 about ten pounds or 10 per cent., and the great size of condenser and air 
 pump were reduced because little of the steam was left. It was found 
 practically that the increase of the range of temperature between the in- 
 coming and outgoing steam was enough to so increase the internal waste 
 that no economy but a positive loss was attendant upon its use. 
 
 Following the Corliss type of engine closely is that of Mr. Jerome 
 Wheelock, of Worcester, Mass., which is however, only two-ported. The 
 valves are of the Corliss pattern and two are used for opening to steam and 
 for opening and closing the exhaust. The closing of the steam is governed 
 by two other valves of the same kind placed upon the steam passage 
 leading to the main valves, and moved from the main valves by clutch 
 links which are tripped by the governor. This engine has a future, but as 
 built, it leaves a separate eccentric for the steam and exhaust valves, and 
 a general increase in strength, to be desired. The use of two eccentrics, 
 one for the steam, and one for the exhaust, valves, has been followed in 
 this country by the Atlas Engine Works, of Indianapolis, and is generally 
 desirable. One of the best frames made with the Corliss engine has been 
 a modification of that used by Wm. Wright, in his four-port slide valve 
 engines and has been adopted by Messrs. Smith, Beggs & Kankin, of St. 
 Louis. 
 
 The Porter- Allen engine is a four-port slide valve engine with steam 
 slides balanced and worked from a link, the position in the link of the 
 radius rod attached to the valve stem being controlled by the governor; 
 the balanced exhaust valves are worked from some fixed part of the link. 
 The engine is intended to run at high speed both in revolutions and pis- 
 ton travel, and on this engine was first studied and developed the actual 
 effect of the weight of the reciprocating parts of the engine, the piston, 
 rod, cross -head and connecting rod. These pieces evidently have to be 
 started on the stroke before any pressure from the steam can reach the 
 crank pin and main journal, and if they are heavy and the speed high, al- 
 most all the pressure of the steam upon the piston may be absorbed in 
 starting this dead weight. In any case the full force of the steam acting 
 upon the piston is reduced before it reaches the crank pin. After the en- 
 gine has passed mid stroke where this weight is now moving at its great- 
 est speed we find that the steam pressure has very much fallen if the cut- 
 off be early, and that the inertia of the moving parts is now pressing with 
 the steam pressure against the crank. Thus the weight of the moving 
 pieces tends in connection with the varying cylinder pressures to equal- 
 ize the pressure, wear, and rotation. 
 
 In some cases we have examined we found that the load on the engine 
 was so light and the boiler pressure so low, the cut-off taking place early. 
 Under these conditions the crank pin was started entirely by the fly wheel 
 and the crank pin actually dragged from one end the weight which 
 
134 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 the steam was pushing at the other. At mid stroke the steam pressure 
 had fallen considerably and the inertia of the weight now moving fast did 
 all the work from there to the end of the stroke. This of course is not de- 
 sirable, but it may be considered that to have half the initial pressure 
 transmitted to the pin, while half is taken up in moving the weight to be 
 given out again at the end of the stroke, is the best adjustment possible. 
 
 THE RIDER AUTOMATIC EXPANSION GEAR. 
 
VARIETIES OF ENGINES. 135 
 
 This, of course, can only be done for one speed and steam pressure by 
 properly designing the weight, and average results only can be reached. 
 This action combined with the high rotation speed on the fly wheel 
 causes these engines to run with what appears to be a simply marvelous 
 smoothness, and the quick acting governor maintains the speed, chang- 
 ing in height under variations of revolution which are too small to be 
 otherwise noticed. Illustrations of this engine are given on pages 167 
 to 187. 
 
 Among stationary engines in England, by far the most common type is 
 that of a slide valve with an expansion slide on the back, and it is the most 
 common type in Europe. The arrangement is very often used with small 
 engines, as small as those with 8 inch cylinders, without means of vary- 
 ing the expansion, and with the speed regulated by a throttle and govenor 
 as usual. In the United States, on the other hand, this form (non-adjus- 
 table) is rarely to be met with. While the use of an expansion slide 
 varied by hand is by no means common, a very neat type is made at 
 Erie, Pa., of the class in which the negative lap is varied by a right 
 and left hand screw and hand wheel projecting from the end of the steam 
 chest. By this wheel the expansion can be changed while the engine is 
 running. 
 
 One of the neatest automatic, or governor, expansion gears, is 
 that of Eider, of the Delamater Iron Works, New York City. The back of 
 the main slide is hollowed into a part of a cylinder whose axis is the centre 
 of the expansion slide. The lap edges of the expansion slide and steam 
 edges of main slide are tapered in such a manner that by rotation of the 
 expansion slide the lap is changed as in the ordinary manner. This rota- 
 tion is accomplished by a spindle attached to the govenor, gearing into a 
 sector on the valve stem. 
 
 When an expansion slide is used in connection with a link motion for 
 the main valve, in many cases the expansion slide is driven by its eccen- 
 tric set opposite the crank, and change of cut-off is given by change of lap 
 and a right and left hand thread and handwheel. In such cases the link 
 is usually worked full-gear only, and the cut-off applied after the engine 
 is in motion. The lap is negative. 
 
 A more common form with large engines, where the weight becomes 
 harder to handle, is to cut into two parts the expansion valve rod and to 
 connect one of them to a link, making the other a radius rod pinned to the 
 valve stem, and working with slides in the link. The position of the 
 slider in the link is determined by a screw and handwheel. Thus a 
 change of travel is effected in the expansion valve. The diagram is 
 readily constructed, and the results are good, but the steam chest has to 
 be made longer than would be otherwise required. 
 
 A more complex arrangement is found in some cases with the link 
 motion, a third eccentric and second link being added. The second 
 link is attached to the expansion eccentric at one end and the main 
 valve stem at the other. The expansion slide is moved by its valve 
 stem, and either by a radius rod and link curved to fit the radius 
 
VMUET1KS Of-' K \irIXES. 137 
 
 rod, or by a double curved link. This combination of links can be 
 sketched as follows : 
 
 A is the crank, and C the shaft centers. B, D, and E are the virtual 
 centers of the three given eccentrics, while F is the virtual eccentric for a 
 
 given position of 
 the main link, and 
 G. on E F, dividing 
 J? it as the expansion 
 
 _._ _ _ slider divides the 
 
 -1 - 
 
 second link, is the 
 virtual eccentric 
 driving the expan- 
 sion slide. 
 
 The diagram 
 for the cut-off is 
 then readily constructed from these two virtual eccentrics by the ordinary 
 methods. 
 
 The use of this class has been confined as far as we are aware to the 
 beautiful marine engine of Messrs. Herreshoff, of Bristol, K. I. 
 
 An automatic engine built under patents of Messrs. Armington & 
 Simms is a good example of neat and careful design. The engine is worked 
 by a piston valve, and the governor is like that of the Buckeye and many 
 other engines, a pair of weights attached to the main shaft. But in the 
 Buckeye the governor acts only on the expansion valve eccentric, pulling 
 it ahead on the shaft if the speed is increased. In the Armington & Simms 
 engine the pair of weights flying out pull on what may be called a compound 
 eccentric. This compound eccentric is the peculiar feature of the engine; 
 the ordinary eccentric, loose on the shaft, carries instead of the usual 
 strap, a second eccentric with about one third the arm of the first. The 
 weight is a heavy casting, pivoted at one end to the governor disc; at the 
 free end it is connected to the lugs on the inner eccentrics, and at its mid- 
 dle to the outer eccentric, but on the other side of the shaft, the result 
 being that the eccentrics are pulled around the shaft in opposite direc- 
 tions. The effect is that the outer eccentric centre changes both in angle 
 and in arm, very much as it is varied in a more simple manner in the 
 Westinghouse automatic engine, already illustrated. The rest of the en- 
 gine requires no comment. 
 
 With large cylinders the valves become large, but while the steam 
 port edge which regulates admission increases with the diameter, the 
 volume to be filled varies with the square of the diameter. And in order 
 to give length enough, the only way it can be increased, the edges are 
 often doubled, or the body of the value cored with passages to the exhaust, 
 while steam is admitted from the sides of the value to the additional ports. 
 When the cylinders are large these valves become very complicated, and 
 are matters of special study. 
 
 Locomotive and portable engines assume special forms but the only 
 real difference is that boiler and engines have to be so attached as to move 
 
138 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 ENGINES OF THE STEAM YACHT "LEILA.' 
 
 SIDE VIEW. 
 Cylinders 9" and 16" X 18". 
 
VARIETIES OF E\<;IXK*. 
 
 139 
 
 ENGINES OF THE STEAM YACHT "LEILA." 
 
 
 END VIEW. 
 
140 STEAM USING; OR, STEAM ENGINE PEACTICE. 
 
 together. The locomotive has always two cylinders,* operating upon 
 cranks at right angles, and carried on from four to twelve wheels, of 
 which from two to twelve have the same diameter and are coupled to- 
 gether thus acting as driving-wheels. The remaining wheels are smaller 
 and act only as bearing wheels. The wheels are set with bearings in a 
 frame to which the cylinders are firmly connected. Upon this frame is set 
 the boiler and its appurtenances. The water and fuel are carried in some 
 forms upon this frame, but more usually upon a separate frame and wheels 
 called the "tender." 
 
 The engine proper is the pair of cylinders with their connections to 
 the wheels and the valve gear back from the wheels to the steam. The 
 valve is almost always the three ported slides driven by a link motion, and 
 there is no difference whatever in the use of steam in the engine itself from 
 any other engine of the same type. The number of revolutions is at times 
 high, and the piston speed great, but the engine is rarely kept at hard 
 work for more than an hour at a time. 
 
 Cylinders are used up to 24 inches in diameter while the stroke is rarely 
 over 26 inches. A few engines for the Central Pacific Railway have a 30- 
 inch stroke, and have an expansion slide on the back of the main slide, 
 but such engines have been tried before and the only gain is likely to be 
 that due clearance, for at slow speeds and late cut-off the expansion valve 
 adds nothing; while at high speed and early cut-off the cushion produced 
 by the link motion and single slide is found to be essential in reducing 
 shock and increasing durability. And it may be safe to say that no other 
 than the single slide in some of its forms, driven by the link or kindred 
 motion, is likely to be used; for in spite of many attempts to improve it, 
 George Stephenson left the locomotive as it still remains in all essential 
 points, and that the only improvement worth noting is the addition of the 
 automatic air brake which may be called the most important adjunct to 
 railway train service after the locomotive. 
 
 The varieties of engines and cut-off gear are almost endless, but we 
 have endeavored to restrict ourselves to the ordinary and well-defined 
 successful constructions. New patterns are being used every day, and we 
 have omitted a number of beautiful and apparently successful forms, 
 manufactured by well-known builders, wholly because they have only 
 been in operation for a comparatively short time. 
 
 *The London and Northwestern Bailway, England, have built a number with three 
 cylinders which have been very successful, and the Boston and Albany is building one 
 with four cylinders, both of these cases being compound engines. Mr. Fairlie and others 
 have built locomotives with four cylinders, which were, however, properly speaking, 
 duplex locomotives. 
 
H- a 
 
 o g 
 
 p 
 
 < ^ 
 
 o S 
 
 

VARIETIES Of 1 ENGINES. 
 
 143 
 
 THE BUCKEYE AUTOMATIC CUT-OFF ENGINE. 
 
144 STEAM USING; OK, STEAM ENGINE PRACTICE. 
 
 THE BUCKEYE AUTOMATIC CUT-OFF ENGINE. 
 
 THE BUCKEYE ENGINE COMPANY'S ISOCHRONAL REGULATOR. 
 
 EXPLANATION. 
 
 Two levers, a a, are pivoted to the arms of the containing case at one of their ends 
 as at 6, while the movable ends are connected by links B B, to ears, or flanges, on the 
 sleeve of the loose eccentric, C, so that their outward movement, in obedience to cen- 
 trifugal force, as indicated by dotted lines, advances the eccentric forward in the direc- 
 tion of revolution. Springs, F F, of tempered steel wire, furnish the centripetal force. 
 The tension of the springs is adjusted by a screw at c. The proper speed is obtained by 
 adding more or less weight at A, or within certain limits by shifting the weights along 
 the levers. The case, which is made in halves, is clamped to the shaft by pinch bolts, a- 
 The spring clips, d, are adjustable along the levers. The parts are shown in their 
 proper position for motion in the direction of the arrow; for motion in the opposite 
 direction, the levers are pivoted to the other arms of the case and the springs are 
 reversed. 
 
THE BUCKEYE AUTOMATIC CUT-OFF ENGINE. 
 
 GENERAL DETAILS. 
 
 a, Bock Shaft. & &, Adjustable 
 Bearings. A, Main Bock Arm. 
 6 6, Clamping Bolts. C D, Arms 
 of Cut-off Bock Shaft, B, actu- 
 ated by Cut-off Eccentric Bod at 
 E. H, Hole for Starting Bar. 
 
 CROSS SECTION OF BED AND COMPOUND BOCK AKM. 
 
 &, Gun-metal Spool in halves, a. Pin. c, 
 Bolt, d, Handle continued in a threaded 
 spindle, which bites into spool, 6, and takes 
 up lost motion . 
 
 BALL AND SOCKET 
 JOINT OF GOVER- 
 NOR LINK. 
 
 PINCH WRIST. 
 
 c, Wrist clamped to stem, 6, by bolt, a. 
 
 a, Ball-headed Stud, d, Head of 
 Link. &, Hardened steel button. 
 c, Cap, to screw down solid. 
 
THE BUCKEYE AUTOMATIC CUT-OFF ENGINE. 
 GENERAL DETAILS. 
 
 BALL, AND SOCKET JOINT IN CUT-OFF 
 VALVE STEM. 
 
 &, Sleeve Nut screwing over head of stem 
 d, jammed by Sleeve Nut c. e, Clamp Han- 
 dle set at about 45. 
 
 CROSS HEAD FOE GIBDEB BED TYPE. 
 
 Cross-head is in halves, pinched on the 
 thread of piston rod, a, by bolts, //. &, Pin. 
 c c, Tongue Gibs, d, Wedges, e e, Screws. 
 
 DEVICE FOR OILING CRANK PIN. 
 
 c, Hole in Crank Pin. b, Tube communicating 
 with c. a, Ball to receive oil ; a, being stationary, 
 the oil is fed into tube by the centrifugal force. 
 
 AUTOMATIC WASTE COCK. 
 
VARIETIES 0^ ENGINES. 
 
 147 
 
 THE BUCKEYE AUTOMATIC CUT-OFF ENGINE. 
 SOME DETAILS OF A 20" x 40" ENGINE. 
 
 Edge of foot -j | 
 
 CYLINDER. 
 
 CYLINDE3 HEADS. 
 
148 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 THE BUCKEYE AUTOMATIC CUT-OFF ENGINE. 
 SOME DETAILS or A 20" X 40" ENGINE. 
 
 VALVES. 
 
 VALVE CHEST COVEK, ETC. 
 
VARIETIES OF ENGINES. 
 
 
 149 
 
 
 THE BUCKEYE AUTOMATIC CUT-OFF 
 
 SOME DETAILS OF A 20" X 40" ENGINE. 
 
 ^iJFO; 
 
THE BUCKEYE AUTOMATIC CUT-OFF ENGINE. 
 SOME DETAILS OF A 20" x 40" ENGINE. 
 
 Of= ~*E1 
 
 TZ-^rrU-^--- =tm\- PT~ 
 X, 3 if^JWTS f JT 
 
VARIETIES OF ENGINES. 
 
 151 
 
 THE BUCKEYE AUTOMATIC CUT-OFF ENGINE. 
 
 SOME DETAILS or A 20" x 40" ENGINE. 
 
 . -. K^ J -4S 
 
 ^_p 
 
 SPLIT ECCENTRICS. 
 
152 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 THE BUCKEYE AUTOMATIC CUT-OFF ENGINE. 
 
 SOME DETAILS OF A 20" x 40" ENGINE. 
 
 BED PLATE. 
 
 CONNECTING KOD. 
 
-^ 
 
156 
 
 STEAM USING; OR, STEAM ENGINE PH. ACTIVE. 
 
VARIETIES OF ENGINES. 
 
 167 
 
 THE REYNOLDS' CORLISS ENGINE. 
 
 **% *j 
 
 END VIEW OF CYLINDER, ETC. 
 Showing arrangement of Valve Gear for 20" X 48" Engine. 
 
158 STEAM USING; OH, STEAM ENGINE PRACTICE. 
 
 THE REYNOLDS' CORLISS ENGINE. 
 
 -SSF 
 
 Valve - 1 1 $ t _i _-A'--~- - : 
 
 4ftggt Valve Stem tf%"Long 
 
 *~ - Exhaust, V_alyeSie?rLl5*/ 8 '%Q?ia_^ 
 Exhaust Bonnet 2 of this \ 
 
 4 of this 
 
 VALVES, ETC., FOE A 20" X 48" ENGINE. 
 
VARIETIES OF ENGINES. 
 
 THE REYNOLDS' CORLISS ENGINE. 
 
 Ki: 
 
 DETAILS OF VALVE GEAR, ETC. 
 FOE A 20" X 48" ENGINE. 
 
160 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 THE REYNOLDS' CORLISS ENGINE. 
 
 DETAILS OF GOVERNOR, ETC. 
 
 For a 20" X 48" Engine. 
 
VARIETIES OF ENGINES. 
 
 THE REYNOLDS' CORLISS ENGINE. 
 
 >" . ^ i , ' ' ''V _ ' *. i n 1 \ i '.'.".' < * 
 
 
 - i-;.^ r | * f 
 
 u-^-t-- J J l^L ! 1 
 
162 
 
 UTEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 THE REYNOLDS' CORLISS ENGINE. 
 
 
II 
 -f as! 
 
 "c 
 
 2 3 "8 a o - 
 
 iiiSig* 
 
 a , 
 
 ff 1 
 
 ! 9 
 i 
 
 oS 
 
 ,{iBJl 
 
 ||||i|s 
 
 lliSs5S|I 
 
 :-|lllEiU 
 
 z i 
 
 3*4 
 
 'cS5ftTass 
 a > -^ 
 
 03 5 -C C8 g 
 
 "- a 2 S 
 
 IfllHJiii: 
 
 tiMiiiiffi 
 
 sf^fis- s 
 
 liujji!!! 
 
 *?.5s a 2fSg 
 S3 a .||3Sg| 
 
 ;a^|5|58?? 
 
THE LAMBERTVILLE, N. J., IRON WORKS PATENT AUTOMATIC 
 CUT-OFF ENGINE. 
 
 VIEW OF GOVERNOR AND SECTION THROUGH BED. 
 
LAMBERTVILLE, N. J., IRON WORKS 
 
 PATENT AUTOMATIC CUT-OFF 
 
 ENGINE. 
 
 GOVERNOR SIDE VIEW. 
 
\ 
 
 
168 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 THE PORTER-ALLEN ENGINE. 
 
 VERTICAL CROSS SECTION THROUGH CYLINDER AND VALVES 
 
 SECTIONAL PLAN OF CYLINDER THROUGH STEAM AND EXHAUST VALVES, 
 
VARIETIES OF ENGINES. 
 
 169 
 
 THE PORTER-ALLEN ENGINE. 
 
 SIDE AND FRONT ELEVATION OF ECCENTRIC AND LINK. 
 
 ELEVATION AND PLAN OF VALVE CONNECTIONS. 
 
THE PORTER-ALLEN ENGINE. 
 
 PLAN AND ELEVATIpN OF MAIN SHAFT BEARING. 
 
 i 
 
 ijy 
 
 CEOSS-HEADS. 
 
VARIETIES OF ENGINES. 
 
 171 
 
 THE PORTER-ALLEN ENGINE. 
 CRANK PIN OILER. 
 
 
 CENTRE LINE OF SHAFT. 
 
IMM Ml 
 
 it it i I I in i i in 
 
174 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
VARIETIES OF ENGINES. 
 
 q 
 
 UUJ 
 
 175 
 
176 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 THE PORTER-ALLEN ENGINE. 
 
 f 
 
 PLAN OF COMPOUND 
 CONDENSING ENGINE. 
 
VARIETIES 
 
 > OF ENGINES. * ' 
 
 L "" ~" """" 
 
 177 
 
VARIETIES OF ENGINES. 
 
 i 
 
180 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE, 
 
THE 
 
 PORTER-ALLEN 
 ENGINE. 
 
 DETAILS OF 
 COMPOUND CON- 
 DENSING ENGINE. 
 
VARIETIES OF ENGINES. 
 
 183 
 
 C Line of Cylinder 
 
 Section on Line B B 
 (See previous page.) 
 
 Section through A A 
 (See previous page.) 
 
 THE PORTER-ALLEN ENGINE. 
 DETAILS FOR A 11^" X 20" ENGINE 
 
THE PORTER-ALLEN ENGINE. 
 DETAILS OF CYLINDER AND VALVES FOR llh>" X 20" ENGINE. 
 
 ' 
 
 
 - /,* 
 
 
 3 
 
 v ,o. 
 
 '7f"~O~ 
 
 O--T- 
 
 "fr 
 
 
 
 1 1 
 
 
 
 1, 
 
 
 1 
 
 ! 
 
 J 
 
 T 
 
 
 s+\ 
 
 
 1 
 
 
 
 \ ""* T* 
 
 i 
 
 
 
 
 \ I? 
 
 -|- 
 
 
 
 -Vj 
 
 
 !<p 
 
 
 
 O.'^o 
 
 --C) J 
 
 
 u i 
 
 
 ; 
 
 
 
 r* 
 
 
 
 
 
 
 
 
 IL4, , 
 
 00 
 
 5 
 
 
 , . 
 
 
 -i- [* iL-j -o:_ . 
 
 S 
 
 . J. .. 
 
 [ (p - 
 
 _ . -0 
 
 
 I 
 
 J "l~l 
 
 , ! 
 
 / i 
 
 
 
 
 .r- 
 
 ' ! '"'< 
 
 
 
 
 4 J 
 
 i 1 " 
 
 r::; 
 
 \ 
 
 I 
 
 ii 
 
 ii 
 
 
 
 
 
 ; 
 
 V . 
 
 
 
 _ !f. L. 
 
 
 
 _! _ _ - _ 
 
 
 
 
 . . ,, / 
 
 
 s^ 
 
 ----- 
 
 :ll 
 
 ; i r 
 
 
 
 
 __ fJnr ._ H 
 
 2 
 
 
 .1 
 
 x x 
 
 | 
 
 1 i'~ ; 
 
 i 
 
 ! 
 
 // 
 
 ! ,JQ, 
 -] 
 
 \ f| j 
 
 ; | 
 
 1 ~^ 
 
:TT 
 
 THE PORTER-ALLEN ENGINE. 
 VALVE GEAR FOB 11^" X 20" ENGINE. 
 
186 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 THE PORTER-ALLEN ENGINE. 
 
 DETAILS OF GOVERNOR FOR 11 V X 20' 
 ENGINE. 
 
VARIETIES OF ENGINES. 
 
 187 
 
 THE PORTER-ALLEN ENGINE. 
 
 \ *&**' 
 
 t J^>JSf' 
 
 In 
 
 t~ 
 
 DETAILS OF MAIN BEARINGS FOR A 11? 2 " X 20" ENGINE. 
 
CHAPTER V.* 
 
 THE ALSATIAN EXPERIMENTS THE WORK OF MESSRS. HIRN AND 
 
 HALLAUER. 
 
 KEPORT ON A MEMOIR PRESENTED BY M. O. HALLAUER, UPON "STEAM 
 ENGINES." BY M. KELLER. 
 
 The work of M. Hallauer is, as its title suggests, a study of the economic 
 influence of the degree of expansion (point of cut-off) in various types of 
 steam engines. It tends also to support by analyses, more and more 
 numerous, the conclusions of the last paper of the author the equality in 
 the matter of industrial consumption of simple and compound engines, 
 the advantage being rather on the side of,the former. 
 
 It is divided into three parts: the first comprises researches relative to 
 double cylinder engines; the second concerns single cylinder engines, 
 and the third sums up and compares the results obtained. We will follow 
 M. Hallauer successively in the order adopted by him in the study which 
 occupies us, viz., first, the compound, then the simple engines, and, sum- 
 ming up for each type, we will conclude upon the whole work. 
 
 FIRST PART. DOUBLE CYLINDER ENGINES, KNOWN AS " WOOLF," OR 
 
 COMPOUND. 
 
 M. Hallauer states, first, all that has been said concerning the ad- 
 vantages of this system, and after discussion arrives at the conclusion 
 that the only really serious consideration, "a priori," which cannot be 
 gainsaid, and which stands in its favor, is this: 
 
 That the difference of force at the commencement and at the end of 
 the stroke is smaller than for other types of engines, and that, in conse- 
 quence of this better distribution the running of the machine is much 
 smoother. In regard to the useful effect of the compound engine, the 
 brake experiments, executed under the auspices of your mechanical com- 
 mittee, have shown that the "Woolf " engines absorb more force in them- 
 selves than single cylinder engines, a result easily foreseen. 
 
 Passing then to the study of the influence of the cut-off in the small 
 cylinder, M. Hallauer commences by verifying the results obtained: 
 
 First. In a series of experiments executed by himself in 1877 with a 
 "Woolf" beam engine, at Miinster, which was made at the shops of the 
 Alsatian Society for the Construction of Machines. 
 
 Second. In the experiments executed in 1876, under the auspices of 
 
 *This chapter comprises translations of some valuable papers contained in the 
 "Bulletin de la Societede Mulhouse," a journal not commonly seen in the United States. 
 The author has endeavored to preserve the form of original thought as nearly as 
 possible. 
 
THE ALSA 77.1 .V A'.V/'A/,'/ l//..\ 7X ETC. 189 
 
 your Mechanical Committee upon a horizontal "Woolf " engine of the same 
 make, which appeared in the bulletin for July-August, 1877. 
 
 Third. In those executed in February, 1877, by the Alsatian Associa- 
 tion of Steam Owners upon an engine at Malmerspach built by the same 
 company, but provided with a variable expansion in the small cylinder, 
 controlled by the governor, and provided by the machine company of 
 Bitsch wilier -Thann. 
 
 Fourth. In the experiments executed at St. Remy by M. Que'm, and at 
 Rouen by the Normandy Association of Steam Owners, upon two "Woolf" 
 beam engines, from the shops of Messrs. Thomas & Powell, of Rouen. 
 
 Having discarded the experiments of which the verification is not 
 exact enough, M. Hallauer determined the consumption of dry saturated 
 steam per hour, and for each absolute horse-power, per indicated horse- 
 power per hour," and per effective horse -power per hour, based upon 
 the sum of the calories* brought into the cylinder by the steam and 
 water leaving the boiler; in other words, substituting for the entrained 
 water the quantity of steam which could furnish to the cylinder the same 
 number of calories which had been brought by that water, and taking 
 account of the calories left in the jacket by the steam which was con- 
 densed therein. He passes then to the analysis of each experiment, and 
 determines the quantity of steam and water contained at the end of admis- 
 sion and at the end of stroke; the variations of the internal heat, and fin- 
 ally of the cooling in the condenser Re.} This cooling is verified by two 
 different methods, already explained many times by Mr. Hallauer in your 
 bulletins. 
 
 I should stop a moment to say some words upon the verification of the 
 experiments. 
 
 M. Hallauer's very elegant manner of operating has already been given 
 in his last work. However, I believe it is not useless to again speak of it 
 here in order to clearly comprehend its importance. 
 
 When it is desired to render an account of the quantity of steam ex- 
 pended during a given time by an engine, the water fed to the boiler is 
 measured. This quantity of water, augmented or diminished by a weight 
 easily calculated, according as the level in the boiler is higher or lower at 
 the end of the experiment than it was at the beginning, gives the quantity 
 of steam used. A method given by M. G. A. Hirn, and many times 
 described in your bulletins, permits the determination of the proportion 
 of water entrained with this vapor. This constitutes the direct gauging, 
 Avhich is verified by the following method: 
 
 Knowing the weight of water and of steam leaving the boiler and the 
 pressure of steam therein, it is easy to determine the number of calories 
 brought to the engine per stroke. This number of calories, diminished 
 by the external cooling and the heat absorbed by the work done, should 
 be equal to the number of calories absorbed by the water of condensation, 
 
 *A calorie is X 2.204 = 3 9672 British heat units. 
 
 tl understand by lie the cooling effect of the condenser upon the steam in 
 cylinders. C. A. 8. 
 
19O STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 a quantity obtained by gauging the water leaving the condenser, and 
 measuring the initial and final temperature. The manner of operation has 
 been many times described. The difference between the number of 
 calories brought to the engine diminished by the work done and by ex- 
 ternal radiation, and that of the calories found in the condenser divided 
 by the total number of calories, gives the per cent, of error which has 
 been committed. 
 
 A word also concerning the term "absolute horse-power" which has 
 not always been easily understood. 
 
 The fine work of M. Hirn has shown the enormous iufluence of the 
 sides of the cylinder upon the action of the steam therein contained. 
 When two engines of different types and dimensions are compared, or the 
 work of the same engine under varying conditions, it is the influence of 
 the internal surfaces which should be determined to render an account of 
 the manner in which the steam is utilized in each experiment. But what- 
 ever system of condenser is used, the influence of the internal surfaces 
 can only vary little; but on the other hand, the vacuum may vary within 
 wide limits, and consequently the indicated work. The variable vacuum, 
 being a point to be considered if one compares engines by their indicated 
 horse-power, can falsify the comparison. We are then forced to suppose 
 that engines, presented for comparison, are furnished with an ideal con- 
 denser keeping a perfect vacuum behind the piston, and to compare be- 
 tween the engines an account of the work furnished with this perfect 
 vacuum. That is the work which constitutes the absolute work of the 
 engine, and it is the consumption for this absolute work which permits 
 the comparison between different engines or of different conditions of 
 working.* 
 
 It is well understood, for the rest, that from the practical and 
 industrial point of view the better the condenser the better the re- 
 sults; this is the affair of the builders, but can have little influence 
 upon the manner in which the steam comports itself in the interior of the 
 cylinder. 
 
 The absolute work is then the term to be employed for comparison of 
 two engines of different types, or working under different conditions, but 
 the effective work will be always the term of comparison to be employed 
 from the industrial point of view. 
 
 This stated, I sum up in the following table the results of the verifica- 
 tions and analyses of M. Hallauer, and with him I arrive at the deductions 
 shown. 
 
 The experiments executed upon the engine at Minister, in which the 
 variation of work is obtained by throttling, give as the difference of con- 
 sumption per absolute horse -power per hour for the extremes of work 
 3 per cent., and this represents the effect of throttling. 
 
 In passing to the effective horse-power per hour, the economy is much 
 more considerable, and reaches 20 per cent., which shows the influence of 
 
 *In English it is usually called the total, or forward work.- C. A. S. 
 
THE ALSATIAN EXPERIMENTS, ETC. 
 
 191 
 
 
 J 
 
 
 OrHg 
 
 
 t- CC 00 
 
 
 
 B 
 
 i 
 
 Is 
 
 - s ^- : j | j ^ j j 
 
 
 O 1C ^ rH 
 CO nrHCN 
 
 
 
 S 
 
 WO" 
 
 
 
 
 
 
 3 
 
 -1 
 
 
 
 
 
 
 - 
 
 f| 
 
 : : : : : . . 
 
 rH <S CO 00 C5 00 O CO rH IS 
 
 
 
 
 
 I 
 
 's 
 
 
 rH =t-0005 = ^ -* t^r-l O 
 
 
 sss- 
 
 
 
 TICAL 
 
 1 
 
 coSJiec ^ -^wioci 
 
 
 SSSo 
 
 
 
 i 
 
 M 
 
 1 
 
 2 rH rH rH CS 
 
 
 sss<- 
 
 . 
 
 
 
 t 
 
 
 
 ^ gs 
 
 1 
 
 H 
 
 w 
 
 H 
 
 
 
 2 rHrHC, 
 
 1 
 
 
 o 
 
 
 
 73 C-JSS rH ct 00 
 
 
 2?3Sc, 
 
 a 
 
 Z 
 | 
 
 i 
 
 H 
 
 rt cooooo ac = o t^ ooo t- 
 
 
 3 rn25 
 
 t 
 
 
 
 s 
 
 
 co co s t- 
 
 
 
 fl 
 
 I 
 
 S 
 
 
 T*oi cs ??iS 01 . 
 
 n 
 
 :- 
 
 sss 
 
 9 
 
 00 
 
 ~ 
 
 x 
 
 1-1 oj c* <s oo" o' * 
 
 
 S2S 
 
 2 
 
 
 H 
 
 H 
 
 <N ...... 
 
 
 rH T ,rHCS 
 
 
 
 i 
 
 H 
 
 J 
 
 2 *5o ^. 
 
 
 
 0<0 
 
 S 
 
 a 
 
 1 
 
 O 
 
 x^ 
 
 T ~ < IO <* X O) t- CO* CO* t-' a rH 00 
 
 1 
 
 S Sr^Sg 
 
 -u 
 
 5 
 
 
 H 
 
 Q, 
 
 ?. OolSI rH X ^.5 rH 
 
 
 00 Sg<0 
 
 1 
 
 2 
 
 
 S 
 
 rn CM t- 00 O IO S^I rHOOOC^J 
 
 
 | SSSSco- 
 
 S 
 
 1 
 
 M 
 
 ^ 
 
 t- oo 55 t- cs oo 
 
 
 , ssi 
 
 
 
 1 
 
 I 
 
 g rHt-OCM = rH t- ^ CO CO 
 
 
 CJ <O rH t- -* 
 
 3 -eso 
 
 5 
 
 
 i 
 
 H 
 
 i 
 
 2 s "" . 2 p S 
 
 
 S S2S8 
 
 
 
 2 
 
 4 
 
 C4 ^i ^4| 
 
 
 S2 
 
 
 
 1 
 
 S" 
 
 t- rH t- CO Ci OC O O O t- rH CO 
 CO 
 
 
 1 - SSg 
 
 
 
 
 
 '. ITS -a 3 :5 -2 :"S ' 
 
 : : .A : S :j :, : J '$ ' 
 
 
 jo j^^ 
 
 
 
 
 
 *ajST3''C aj S 'tcc"S>a <*> T3 
 
 nn2Sx>c83 I-e-aSBSb^L -a 
 
 ;:*% fllll! s li 
 
 *H e . ja o eS eg ac-t O aj 
 
 |lla- ' |l!|l|!fl| f| 
 
 i 
 
 Englishl. H. P 
 Pounds water per hour 
 wardH.P 
 Lba. water per hour I. ] 
 E. ] 
 ac in heat units 
 
 
192 STEAM USING; OK, STEAM ENGINE PRACTICE, 
 
 the back pressure measured with the work done, and the coefficient of 
 friction which augments in the same circumstances. 
 
 M. Hallauer has then reached the conclusion in his preceding work 
 that the most simple process of regulation is an expansion valve regulated 
 by hand and the governor throttle, the control by hand being used for the 
 larger variations and the throttle for maintaining the speed uniform in 
 spite of the minor variations of work which may occur every instant. 
 
 The author also explains by simple considerations, based always upon 
 the action of the internal surfaces, some apparent anomalies which seem 
 to occur in the experiments. I will not enter into these details, which 
 would compel us to present the entire work, they all prove that the practi- 
 cal theory pointed out by M. Him and applied here by M. Hallauer, per- 
 mits the explanation of all the phenomena which take place in the interior 
 of cylinders. After having studied the condensation and evaporation in 
 the interior of the small and large cylinders, and noting the marked differ- 
 ences which occur from an expansion more or less prolonged in the small 
 cylinder, M. Hallauer reaches the important conclusions which follow, 
 and which I sum up under a slightly different form from that which he 
 has adopted. 
 
 1. Given a boiler working at 5.5 k.* of pressure, for instance, and 
 a "Woolf " engine which can furnish a maximum work of A horse-power, 
 there is a possibility of obtaining, industrially at least, 10 per cent, econ- 
 omy by cut-off in the small cylinder, instead of throttling down the steam 
 at the times when, because of circumstances, the engine has to supply 
 only the half of A horse-power; this economy will be diminished by the 
 measure in which the work approaches to A horse-power. 
 
 2. The engine working nearly to its maximum capacity, there can 
 be produced by throttling a variation of force of 10 per cent, without 
 any marked change in the economic regime. 
 
 We see at once the importance of these conclusions, and in reality, 
 there are few "Woolf" engines working at full power; furthermore, some 
 builders of our district, when they furnish an engine, declare it at less 
 than one-half the power that it really is. 
 
 Thus, an engine sold at 100- horse power can ordinarily furnish 200- 
 horse power without reaching its maximum. This mode of operation has 
 practical advantages, but it is none the less true that when the engine 
 only gives 100-horse power, thanks to the throttling, it consumes 10 per 
 cent, more than it would have consumed if the work of 100-horse power 
 had been obtained by admitting full pressure and cutting off in the small 
 cylinder. 
 
 I do not agree with M. Hallauer when he recommends a cut-off vari- 
 able by hand. I believe that from the practical point of view a governor 
 cut-off works better, for it disposes of the neglect of the engineer, who 
 may not be able very well to give the desired cut-off the moment that it 
 should be applied. We have nearly always observed that an automatic 
 expansion procures a more regular speed than a governor throttle. 
 
 *14.223 pounds per square inch = 1 k. per square centimetre. 
 
THE ALSATIAN EXPERIMENTS, ETC. 193 
 
 SECOND PAET. INFLUENCE OF CUT-OFF IN SINGLE CYLINDER ENGINES. 
 
 M. Hallauer proceeds for these engines as he did for the compound 
 engine. He first verifies the experiments upon which he rests; then he 
 analyzes them. 
 
 The documents which have served for this study resulted from the 
 following experiments: 
 
 First. Those executed under the auspices of your mechanical com- 
 mittee in April and May, 1878, upon a Corliss engine, constructed by 
 Messrs. Berger, Andre & Co., of Thann. 
 
 Second. Those undertaken in 1873 and 1875 by Messrs. Hallauer, Gros 
 seteste and Dwelshauvers-Dery, under the inspiration of M. G. A. Hirn, 
 and executed upon an engine deprived of its jacket, and working with 
 superheated and with saturated steam. We group the results of the veri- 
 fications and analyses in the table which follows. 
 
 The examinations of these results show: 
 
 First. That for the Corliss engine, with steam jacket, there is a theo- 
 retical economy per absolute horse-power of 1^ per cent, when the cut- 
 off is changed from to Jj, but that industrially, this economy disappears 
 and changes sign, and there is a practical gain per effective horse -power 
 of 4^ per cent, by working at J cut-off in place of -f f . 
 
 Second. That for the engine, without jacket and without superheat- 
 ing, the economy furnished by the cut-off is much more considerable than 
 the Hirn engine working with saturated steam, and that there is a theo- 
 retic gain of 7.4 per cent, by changes to cut-off \ from , and that, indus- 
 trially, this economy remains at 4 cent. 
 
 Third. In taking account of the difference of superheating, the ex- 
 periments I. and II. (106 C. and 231 C.), establish the fact that the in- 
 fluence of the cut-off in the unjacketed engine, with superheated steam, 
 remains as it did with saturated steam. 
 
 Experiment III., with superheated steam, shows still more the great 
 economy of the cut-off in these circumstances, where one passes certain 
 limits, for between admissions of and | there is 15 per cent, economy, 
 which would have been more considerable if, in experiment I., we had 
 worked with the same superheating as in experiment III. 
 
 Fourth. In the Hirn engine the experiments with saturated steanj give 
 at the end of the stroke the same weight of water, 0.0940 k.* and 0.0927 k., 
 and the refrigeration is also the same, 37.53 C. and 37.02 C., while with- 
 out any jacket the same weights of water gave the same values of Re. For 
 the Corliss engine, on the contrary, the weights of water differ at the end 
 of the stroke, 0.0298 k. and 0.0398 k., and the same refrigeration, 11.21 C. 
 and 11.15 C., results showing the steam jacket. For experiments II. and 
 III., with superheating, the weights of water at the end of the stroke are 
 0.0367 k. and 0.0373 k., and the refrigerations 16.61 C. and 20.34 C., a dif- 
 ference which should be attributed to superheating, for in the same con- 
 
 *i k. = 2.204 pounds. 
 
194 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 O ** 
 
 X X 
 
 
 
 I 
 
 CM 
 
 TH ocboS J3CM 
 
 * 
 
 CO O 
 
 
 | 
 
 H 
 
 o o t-^ oo os oj co 
 
 l_ 
 
 s 
 
 
 * 
 
 s 
 
 I . SI . 
 
 s 
 
 I 
 
 I 
 
 K 
 
 * O t- t- 00 O CM 
 
 S" 
 
 
 1 
 
 02 
 
 1 
 
 2 .iiS ., 
 
 8 
 
 H 
 
 
 M 
 
 CO O O t- 00 < *H 
 
 CO CN 
 
 a 
 
 
 a 
 
 TH t- ^1 CO CO 
 
 CO 
 10 -* 
 
 
 1 
 
 
 O O 00 OS O TH O 
 
 C^S 
 
 
 1 
 
 
 T ScbS 
 
 TH TH 00 00 OS O CM 
 
 00 
 
 
 
 
 got-ccos ^ 
 
 1- TH 
 CO CM 
 
 
 
 
 o t-mo 
 
 iH * CO O O CO CO O 
 
 So 
 
 i 
 1 
 
 S 
 
 H- 1 
 
 oocNt-t-ooiomoo 
 
 rH 00 
 
 
 1 
 
 S* 
 
 oo o os -* 
 
 CO CO CM 
 TH CO CM OS t- t- CM 
 
 TH 00 
 
 * 
 
 2 
 
 
 t-CNt~L-OOOrHOS 
 CO CO i-H 
 
 THOS 
 
 1 
 
 i 
 
 M 
 
 i-l iHrHOSO>OCOt- 
 
 o 
 
 ^ 
 
 
 
 
 expansion 
 
 error committed 
 actual horse-power per hour 
 indicated horse-power per hour, 
 effective horse-power per hour. . 
 condensed in jacket 
 water at end of admission 
 expansion 
 radiated to the condenser by t 
 
 cent, of heat received 
 
 r""^ ^ h o 
 
 Q)= = +i S *-l 
 
 s;|^ft ^^ g'SS 
 
 g|S| ls 88 8 -|5 
 
 Illl || 4^ 
 
 <M -* O 00 CO 
 
 10 C-OS * 
 CO TH rH rH ^ 
 
 * O5 O5 
 00 W OS 
 
 -* Ift t~ OS -rl 
 rH TH rH * 
 
 total H. P 
 I. H. P 
 E. H. P 
 
 pe 
 
 British I. 
 Pounds w 
 
THE ALSA TIAN EXPERIMENTS, ETC. 195 
 
 ditions saturated steam rejects heat in the condenser according to the 
 weight of water evaporated at the end of the stroke. This shows that 
 superheating acts in a different manner from jacketing. 
 
 Analyzing the variations of internal heat in the various experiments 
 before us, and the values of the refrigeration in the condenser which re- 
 sult from differences of cut-off, according as they work with or without a 
 jacket, and with or without superheating, M. Hallauer comes to the same 
 conclusion as to the different modes of action of the jacket and super- 
 heater. The jacket acts more energetically than the superheater, and sim- 
 ilarly during the period of expansion, but becomes disadvantageous dur- 
 ing that of condensation, for then it furnishes heat to the water which 
 lines the internal surface of the cylinder, and augments that rejected 
 into the condenser, which is not the case with superheating; this renders 
 the latter more economical in most cases. Examining, then, the theoretic 
 economy, and comparing the consumption per absolute horse-power per 
 hour, the industrial economy and the amounts per effective horse-power 
 per hour, the author agrees with M. Zeuner that large expansions are 
 economical from a theoretical point of view, and with M. Hirn that the 
 reverse is the case from an industrial standpoint. 
 
 Thus the conclusions of these two savants, which had the air of con- 
 tradiction, are found to be in accord, taking account of the different con- 
 siderations which guided them; the one, M. Zeuner, having made a gen- 
 eral study of steam engines based upon conditions non- realizable (non- 
 conducting internal surfaces of the cylinder); the other, M. Hirn, on the 
 contrary, having studied them from the industrial basis, and taking 
 account of the action of the internal surfaces and the work done in 
 friction. 
 
 THIRD PAKT. COMPARISON OF THE DIFFERENT TYPES OF ENGINES 
 
 STUDIED. 
 
 The close of M. Hallauer's work includes the comparison of compound 
 and simple engines; still taking the analysis of the experiments which 
 served for the two first parts of the work, he restates what he understands 
 by absolute, indicated, and effective horse-power. 
 
 The absolute work is the work which would have been given by the 
 engine with an ideal condenser, making a perfect vacuum behind the pis- 
 ton, the theoretical work.* 
 
 The indicated work is that furnished by the steam upon the pistons; it 
 includes the influence of back pressure; it is the work which the dia- 
 grams traced by the indicator enables us to calculate. 
 
 The effective work is then the work industrially disposable; it takes 
 account of the back pressure and the friction of the various parts of the 
 engine itself. 
 
 That stated, we have two of the verified experiments: one executed 
 upon the Corliss engine, cutting off at , the other at the Malmerspach 
 
 *The forward work. 
 
196 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 engine, cutting off at ^ in which the figures approach each other as 
 closely as possible. 
 
 
 
 Corliss. Woolf. 
 
 
 
 per cent, per cent. 
 
 Water contained at the end of 
 
 admission 
 
 OK 3_ OQ 7_ 
 
 
 expansion 
 
 18 5 II- 9 - 
 
 Difference of initial and final 
 
 heat 
 
 
 Rejected in condenser. . . 
 
 
 
 The comparison of these two experiments brings us to very interest- 
 ing conclusions. 
 
 A phenomenon to be first noted for the "Woolf" engine destroys a false 
 idea held hitherto: that for the double -cylinder engine a portion of the 
 force is withdrawn from the cooling influence of the condenser. 
 
 In reality, in the small cylinder the expansion from half stroke gave 
 place to an evaporation of 10.6 per cent, of the weight of water contained 
 at the end of admission, and the internal heat increased much more rap- 
 idly than in the Corliss engine. Then the mixture of steam and water 
 passed to the large cylinder with 13.1 per cent, of water, and in place of 
 the evaporation continuing in this cylinder there was, on the contrary, 
 condensation, in spite of the jacket, till at the end of the stroke 4.8 per 
 cent, more water was deposited than at the end of the stroke in the small 
 cylinder. This shows that the great condensation at the moment of 
 entrance into the large cylinder is strong enough, in spite of the jacket; 
 that all the steam condensed in the large cylinder, and that which has 
 come from the small cylinder, cannot be evaporated. 
 
 In the Corliss engine, on the other hand, there is a continuous evapo- 
 ration till the end of the expansion. We see that the influence of the 
 large cylinder in the case of expansion, commenced in the small cylinder, 
 does not always tend to lessen the rejected heat, but, on the contrary, may 
 sometimes augment it, which is the reverse of that which has been hith- 
 erto admitted. 
 
 The consumption of dry saturated steam per absolute horse -power per 
 hour gives the theoretic economy realized by the one or the other motor. 
 If we compare the consumption of the Corliss with that which it would 
 have had with an expansion of 13, which is that of the "Woolf" engine, 
 for the experiment with which we compare, we find that the theoretic 
 economy is 4 per cent, in favor of the "Woolf" engine; but the influence 
 of the back pressure, and still more that of the friction, reduces this 
 economy when we pass into the industrial domain. It then changes sign, 
 and we find that the practical economy becomes 8.7 per cent, in favor 
 of the Corliss. 
 
 Comparing, then, the horizontal "Woolf" engine with the Corliss, and 
 taking account of the lack of compression in the former, M. Hallauer finds 
 
THE ALSATIAN EXPERIMENTS. ETC. 197 
 
 again a theoretic economy of 4 per cent, for the "Woolf" engine, but that 
 economy also disappears in practice, and the Corliss becomes industrially 
 superior by 9.05 per cent, to the horizontal compound. Meanwhile, con- 
 sidering the construction of this latter, the difference falls to 2 per cent., 
 and then the horizontal "Woolf" engine consumes only 8.8 k. of dry 
 saturated steam per effective horse-power per hour, which is the smallest 
 consumption at which we can arrive by a careful construction of condenser, 
 and a very strong compression. In closing, M. Hallauer sums up in a table 
 the consumption of steam per hour per horse-power, absolute, indicated 
 and effective, adding the consumption per effective horse-power per hour 
 of coal, on the basis of an evaporation of 8. 
 
 The table shows that the various types of double cylinder engines con- 
 sume from 9.1 to 9.5 k. of steam per effective horse-power per hour; the 
 Corliss, with cut-off , only uses 8.6 k., while the Hirn engine, with super- 
 heated steam, and f cut-off, only uses 8k., or 17.6 pounds. 
 
 To sum up: The work of M. Hallauer, which is one of those laborious 
 and conscientious studies to which he has so long accustomed our society, 
 is above all very remarkable in its conclusions an/1 tends once more to 
 prove the impossibility of stating anything with precision concerning 
 steam-engines, if one does not rest on verified experiments of existing 
 engines. 
 
 We have been compelled in this summary to leave untouched many 
 interesting things, aiming mainly to unite the divers conclusions of M. 
 Hallauer; but all parties concerned in steam-engines will certainly find in 
 his work exceedingly useful material to serve them in many circumstances. 
 We are cognizant of owing many thanks to M. Hallauer for all these inves- 
 tigations which demand so much time, patience, and reflection, all of 
 which he has not been sparing in this his last memoir; you know, for the 
 rest, what he is accustomed to do, by the numerous works which he has 
 already presented to our society, and of which this last gives the most 
 interesting practical conclusions. 
 
 EXPERIMENTS WITH A STEAM ENGINE. A MEMOIR, PRESENTED BY M. O. 
 HALLAUER, KELATING TO THE EXPERIMENTS DIRECTED BY M. G. A. 
 HIRN, EXECUTED BY M. M. DWELSHAUVERS-DERY, W. GROSSETESTE 
 AND O. HALLAUER. 
 
 In his recent work on Thermodynamics M. Hirn describes the remark- 
 able progress that the judicious employment of the principles of this new 
 science has brought about in the study of heat engines, at the same 
 time that it shows that various edifying theories were far from the reality; 
 finally it proves with evidence the impossibility of general theories. 
 
 If we take up this question so fully treated by him; if we analyze one 
 after the other the experiments made under his direction; it is to the sin- 
 gle end, as he himself has well said, "of shunning the numerous blunders" 
 of practical men who have wished to pursue such studies; and it will also 
 be seen how we have obtained an experimental solution of many important 
 
198 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 problems not yet studied, so that we can encourage engineers on a track 
 so fruitful of results, for we are certain that researches of this nature lead 
 to one useful end, the reduction of fuel. But in order that this memoir 
 may serve as a guide in new work, it is essential to give here the method of 
 experiment so happily inaugurated by M. Him and so often described by 
 him. 
 
 METHODS OF EXPERIMENTS. BRIEF DESCRIPTION OF THE ENGINE. 
 
 The engine with which we worked was a beam engine, with vertical un- 
 jacketed cylinder and four valves, two for admission and two for exhaust. 
 A differential movement of the two cams which worked the admission 
 permitted us to vary the angle at will, and, consequently, at various de- 
 grees following the needs of the experiments. The steam brought to the 
 cylinder immediately on leaving the boiler passed through many tubes, 
 of cast-iron, placed behind the boiler in a special chamber where the sim- 
 ple movement of two registers brought in the hot gas. The steam carried 
 with it a degree of superheating which ranged from 195 C. to 231 C., while 
 it could be brought to the engine in the state it left the boiler by chang- 
 ing the registers to give direct passage to the pipe. The consequence of 
 this analysis will lead us to see that M. Him in constructing the most 
 practical engine actually known, has also made a veritable instrument of 
 precision for researches in experimental physics. The dispositions, as sim- 
 ple, as ingenious, which he adopted to assure the exactness of the opera- 
 tions have served in most cases. 
 
 Observations Taken. These are grouped into three distinct series: the 
 measurement of the water consumed and rejected from the condenser, 
 those of temperatures and the work. 
 
 1. The quantity of steam that the engine consumed was obtained by 
 direct measurement; the easiest made the water for the feed pump 
 flow through a reservoir of constant known volume, alternately filled and 
 emptied. If the water level in the boilers is noted evening and morning, 
 and account taken of this difference of level, the only precaution neces- 
 sary is to ascertain by hydraulic pressure that no leaks exist in the boiler, 
 superheater and pipes. The total weight of feed water passing the gauge 
 tank divided by the number of strokes, B, during the day, gives us the 
 water consumed per stroke, M. 
 
 That rejected by the air pump is also gauged, but differently. At the 
 outlet a pit of masonry is arranged; its capacity at different heights was 
 determined by putting in known weights of water, and at the bottom of 
 this plate a copper plate was placed, pierced with a circular orifice 0.06 m. 
 diameter, out of which continually flows, under a head of from 0.6 m. to 
 0.8 m., the water brought from the condenser. The discharge of this 
 orifice is found experimentally by closing it tightly by a stopper, filling 
 to a known height and noting the time it takes to lower each 0.1 m. 
 
 Let S be the horizontal section of the pit, s the area of discharge, and 
 
THE ALSATIAN EXPERIMENTS, ETC. 
 
 199 
 
 n an unknown coefficient of contraction, h the head above the centre of 
 the orifice, and t the time. We then have: 
 
 Sdh = nadt Vfyh 
 2 & (VHt VHJ = naTV2g 
 
 This gives at once na V2g when the time T is noted, while the level 
 falls from H to H^ 
 
 The volume of water, or the weight rather, running out under head 
 H , is II = (na Vfy) VH. 
 
 During the days of experiments we noted every fifteen ininutes the 
 heads h , h lt h. . . these varied little and 
 
 - - ' ; or, more simply even 
 
 
 whence II = (ns VW v/7im 8 (hl ~ 
 
 &! and h* are initial and final heads of an experiment whose duration was 
 T. If in this time the engine made B strokes the weight of water rejected 
 
 II T 
 
 per stroke was J/ = - . 
 B 
 
 The injection per stroke was II M. 
 
 We give in tabular form the values obtained in eight experiments, 
 each consuming an entire day: 
 
 TABLE I. 
 
 DATE. 
 
 Superheat'ng. 
 
 Expansion. 
 
 Weight of 
 Steam 
 Consumed. 
 
 Injection 
 Water. 
 
 Nov. 18, 1873 
 
 231 C 
 
 4 
 
 3065 
 
 9.3500 
 
 Nov 28, 1873 
 
 none 
 
 4 
 
 3732 
 
 9 29175 
 
 Aug. 26, 1875 
 
 215 
 
 5 
 
 2651 
 
 8 7291 
 
 Aug. 27, " 
 
 223 
 
 2 throttled. 
 
 0. 2822 
 
 8 5983 
 
 Sept. 7, " 
 
 195 
 
 7 
 
 2240 
 
 8 7384 
 
 Sept. 8, " 
 
 none 
 
 7 
 
 0.2634 
 
 8 . 9132 
 
 Sept.29, " 
 Oct. 28, " 
 
 220 
 220 
 
 2 throttled, 
 non-condens - 
 
 0.2265 
 2715 
 
 5.9810 
 
 
 
 ing. 
 
 
 
 2. Temperatures. Part of these are noted directly, but with different 
 instruments. To arrive at the superheating a good thermometer is simply 
 placed in mercury in a copper pocket open at one end and closed at 
 the other and let into the steam pipe ; this measurement is easy to 
 obtain. 
 
 The rise of temperature of the injection water in the condenser neces- 
 sitates a more careful measurement a special thermometer known 
 already to the readers of our Bulletins, the differential air thermometer. 
 
200 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 The practical arrangements that M. Him brought about this instrument 
 permitted the obtaining to the fiftieth of a degree the increase in tempera- 
 ture. We will only indicate the method of procedure with its use without 
 entering upon any detailed account of the construction of the instrument. 
 
 Two thermal reservoirs filled with perfectly dry air, connected to indi- 
 cating tubes by capillary copper tubes, are both plunged into the cold 
 injection water and put, by opening three-way valves with which they 
 are fitted, into communication with the external air. The temperature of 
 the cold water is also noted with a good mercurial thermometer reading 
 to the tenth of a degree, also the divisions where the indicating liquid in 
 the pressure tubes stops; finally, the height of the barometer. A turn of 
 the three-way valve closes the connection to the external air, while it 
 leaves open that between the reservoirs and indicating tubes of the appa- 
 ratus. Placing then one reservoir in the cold water brought to the injec- 
 tion, the other in the warm water rejected, we note every quarter of an 
 hour the heights of the liquid. The averages for an entire experimental 
 day are transformed to degree, centigrade, and give by their difference the 
 rise in temperature of the water during that day. The formula for the 
 air thermometer is as follows : 
 
 Let .B=height of barometer at start. 
 
 i= temperature of cold water at start. 
 
 5 / =mean height of barometer during the day. 
 
 = temperature sought. 
 
 y Coefficient of dilatation of the reservoir, usually copper. 
 
 a=that of the air. 
 
 p_ tt tt indicating liquid. 
 
 a = temperature " " 
 
 V= volume of reservoir. 
 
 s/i= " " liquid moved in tube. 
 
 Then: 
 
 / V \ /I + /A /I + a 
 
 13.596 B f 
 
 ,o r 
 
 = 13 - 
 
 + 
 
 By taking large reservoirs, - = 1 nearly. 
 
 Practically, when we only have in view industrial results, the employ- 
 ment of a good mercurial thermometer reading to tenths of a degree con- 
 
 TABLE II. 
 
 
 Nov. 18. 
 
 Nov. 28. 
 
 Aug. 26. 
 
 Aug. 27. 
 
 Sept. 7. 
 
 Sept. 8. \ Sept. 29. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Temperature of in- 
 jection 
 
 31.3 
 
 33.65 
 
 33.09 
 
 35.26 
 
 30.42 
 
 32.25 37.81 
 
 Temperature of re- 
 jection . . . 
 
 12.6 
 
 11.83 
 
 16.50 
 
 16.50 
 
 16.37 
 
 16.50 15.85 
 
 
 
 
 
 
 
 j 
 
 Difference 
 
 18.7 
 
 21.82 
 
 16.59 
 
 18.76 
 
 14.05 
 
 15.75 i 21.96 
 
THE ALSATIAN EXPERIMENTS, ETC. 
 
 2O1 
 
 duces to results sufficiently close, but there are times when their use is far 
 from being as convenient as that of the air thermometer. 
 
 The other temperatures corresponding to the steam pressures are 
 taken from the tables of Eegnault. They are thus obtained indirectly, and 
 are experimental; for these, tables of pressure and temperature are 
 deduced from observation. 
 
 The values which serve us were obtained by a generator and free -air 
 manometer, and in the interior of the cylinder by measurement upon the 
 diagrams of work, which gives, as all know, the curve of pressure for each 
 point in the stroke of the piston. 
 
 TABLE III 
 
 
 00 
 
 1 
 
 1 
 
 
 
 150.00 
 124.00 
 94.40 
 58.25 
 
 I 
 
 06 8' 
 
 43 
 
 a 
 
 O 
 
 Temperature in boiler. 
 
 150.15 148.20151.00 
 144.96 140.78148.74 
 97.24 98.24 92.05 
 73.49J 73.42 58.86 
 
 | 
 150.77150.77 151.20146.20 
 142.00 141.30 115.80 137.49 
 85.00 84.33) 87.36101.89 
 58.44 61.15! 57.11 
 
 Temperature at cut-off 
 
 Temperature at end stroke 
 
 Temperature during exhaust 
 
 3. Measurement of the Work. This requires the most minute care and 
 was effected by the aid of two sets of apparatus, the one checking the 
 other the indicator of Watt and the pandynamometer of flexion of Hirn 
 this latter, above all, installed in very favorable conditions, has fur- 
 nished us values most remarkable for their exactness. We have then, 
 based upon the results which it gave, analyses in which the de- 
 termination of the various works, with full pressure, with expansion, 
 the total forward work, plays an important part ; also the measurement of 
 pressures at the points of the stroke where we wish to study the thermic 
 conditions of the steam and the transformations to which it is sub- 
 mitted. 
 
 It is owing to the happy idea of utilizing as a spring the beam of the 
 engine that M. Hirn arrived at the construction of the instrument which 
 he has already described in our Bulletins. This apparatus draws at each 
 instant the deflection of the beam, and gives a curve, of which the 
 abscissas are proportioned to the stroke and the ordinates to the 
 pressures. 
 
 The study of the diagrams drawn by these instruments is nearly the 
 same. The indicator, then, is less exact for engines with a single cylinder, 
 but it has been long used, not as an instrument of analysis, but as a means 
 of measuring work, while we only insist on the curves traced by the 
 pan dynamometer. 
 
 With this latter the scale of abscissas is obtained by direct measure- 
 ment upon the movable plate and for equal parts of the stroke; that of the 
 ordinates by a direct comparison; the engine is brought to one of its dead 
 points, the steam turned into the cylinder at a known pressure, the force 
 
2O2 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE, 
 
 exerted on the piston is transmitted to the beam which deflects an amount 
 measured by the pencil of the apparatus. Then to check in the reverse 
 direction, the crank is put on the other dead point and the operation re- 
 peated. M. Hirn has established, every twentieth of the stroke, the load, 
 in kilogrammes, needed to bend the beam such an amount that the pencil 
 of the dynamometer shall describe an arc of a circle of one metre in de- 
 velopment; it is the unit of measure at each corresponding twentieth of 
 the stroke. 
 
 All the curves drawn have been carefully measured; the mean ordinates 
 transformed into kilogrammes upon the piston are set off on the corres- 
 ponding lines, 10 millimetres for 1,000 kilogrammes. At the same time we 
 trace below the axis, to the same scale, the values of the vacuum given by 
 the indicator alone, joining the extremities by a curve, drawn full for the 
 pressure, and dotted for the vacuum. Between these curves we have the 
 pressure carried at each moment by the piston. If the absolute vacuum 
 could have been reached on one side, while the steam acted on the other, 
 nothing is more easy than to establish the pressure per square metre; the 
 load divided by the piston area gives it. 
 
 This known, the tables of Regnault, or the formula of Roche, fix the 
 corresponding temperature, and we can calculate then : 
 
 y the densities, 
 
 A the total heat of evaporation, 
 
 ZJthe internal heat, etc., 
 
 of the steam at points of the stroke which interest us, without which 
 values any analysis is impossible. 
 
 Finally, the planimeter, by a measure of the area, gives us the different 
 kinds of work, the forward work at full pressure, the total forward work, 
 the total indicated work, and the effective work, or the difference between 
 the indicated work and that absorbed by the friction of the parts of the 
 engine. 
 
 TABLE IV. 
 
 
 Nov. 18. 
 
 Nov. 28. 
 
 Aug. 26. 
 
 Aug. 27. 
 
 Sept. 7. 
 
 Sept. 8. 
 
 Sept. 29. 
 
 Oct. 28. 
 
 Revolutions per 
 minute 
 
 30.1736 
 
 30.5494 
 
 29 .969 
 
 30 306 
 
 29.98 
 
 30 41 
 
 30 13 
 
 30.00 
 
 T. H. P. in. ch.d. 
 v. ot 75 kgms. in 
 forward work.. 
 
 144.36 
 
 136.46 
 
 135.77 
 
 125.17 
 
 113.08 
 
 107.81 
 
 99.53 
 
 78.30 
 
 Checking the Consumption. Thanks to the data taken, we can estab- 
 lish an equality in the heat brought to the engine and that used in work 
 and rejected from the condenser, so that we can make a first verification 
 showing with what approximation we have noted the consumption of 
 steam. 
 
 The first fundamental proposition of thermodynamics experimentally 
 established is that, heat acting upon a body gives place to mechanical 
 work and a quantity of heat proportioned to the work disappears; the 
 
THE ALSATIAN EXPERIMENTS, ETC. 
 
 2O3 
 
 relation between work produced and heat disappearing is constant and 
 depends in no way upon the body by which it acts. We should then have 
 between the heat brought to the engine and that found in the injection 
 water a difference, a decrease proportioned to the external work done, 
 or the forward work. The value of this decrease is no other than the quo- 
 tient of the number of kilogrammetres furnished by the engine per stroke, 
 divided by 425 kilogrammetres, the equivalent of a calorie transformed 
 into work. 
 
 Take for example the experiment of August 26, 1875, with steam at 
 21 5 C. The mean consumption per stroke for this day is 0.2651 k. Taking 
 the boiler pressure, 49,938 kgs. per sq. metre, a temperature of 151 C., it 
 takes from the boiler 0.2651 k. x h = 0.2651 k. (606.5 + 0.305 x 151); but 
 before reaching the cylinder it traverses the superheater, which raises its 
 temperature to 215 C., and furnishes to the fluid 0.5 x 0.2651 (215151) 
 more. The specific heat of steam being 0.5, finally, when this fluid leaves 
 the condenser, it takes with it 0.2651 / = 0.2651 x 33.09, which it is necessary 
 to take away from the amount brought, and we have for the heat available 
 in the cylinder. 
 
 Q a = 0.2651 k. (606.5 + .305 x 15*1 + 0.5 (215 151) 33.09.) = 
 
 172.79 c. 
 
 The cold water of injection, measured as above is, 8.7291 k.; it receives 
 from the steam which it condenses Q l = 8.7291 k. (fi) = 8.7291 k. + 16.59 
 = 144.82 c. 
 
 There has disappeared in the interval: 
 
 172.79 144.82 = 27.97 c. 
 
 But we have used externally 135.77 ch., say 10,193 kgms. per stroke, which 
 has absorbed 23.99c. The external radiation of the cylinder has been 2.5c., 
 total of 26.49c., a difference of 1.48c., say of 0.8 per cent, of 172.79c. brought 
 to the cylinder. This is the error and check upon the consumption, as the 
 distinction between the observations is clearly defined. 
 
 TABLE V. 
 
 
 Nov. 18. 
 
 NOV. 28. 
 
 Aug. 26. 
 
 Aug. 27. 
 
 Sept. 7. 
 
 Sept. 8. 
 
 Sept. 29. 
 
 Oct. 28. 
 
 Heat brought. 
 
 202.72 
 
 228.80 
 
 172.79 
 
 184.41 
 
 144.34 
 
 161.51 
 
 147.05 
 
 g 
 
 Heat in work 
 
 
 
 
 
 
 
 
 p 
 
 and radiated 
 
 27.82 
 
 26 15 
 
 26.49 
 
 24.36 
 
 22.47 
 
 21.27 
 
 19.99 
 
 Is 
 
 Difference.. 
 
 174.90 
 
 202.65 
 
 146.30 
 
 160.05 
 
 121.87 
 
 140.24 
 
 127.06 
 
 2L P 
 
 Heat rejected 
 
 174.84 
 
 202.75 
 
 144.82 
 
 161.30 
 
 122.77 
 
 140.38 
 
 131.34 
 
 ' 
 
 Error 
 
 +0.06 
 
 0.10 
 
 + 1.48 
 
 1.25 
 
 0.90 
 
 0.14 
 
 4.28 
 
 
 We see that all the errors but that of the 29th September are less than 
 1 per cent. The latter has an error, of which we cannot discover the cause, 
 which does not exceed 3 per cent. 
 
 Verification of the Work. To be more correct, we should describe the 
 verification of the indicator, for the simple inspection of the curves drawn 
 
2O4 
 
 STEAM USING; OK, STEAM ENGINE PRACTICE. 
 
 by the aid of the pandynamometer renders evident the superiority of this 
 method of valuing the force. And we may add that many successive ex- 
 periments always led M. Hirn to the same results and the same coefficients 
 for the scales of pressures. On the other hand the sufficiently large curves 
 permit us to obtain, with much exactness, not only the pressures but also 
 the fractions of the stroke and consequently the volumes affected by the 
 cut-off; also, by taking these curves and comparing them with the indi- 
 cator diagrams taken under the same conditions, we form the following 
 table: 
 
 TABLE VI. 
 
 
 Aug. 26. 
 
 Aug. 27. 
 
 Sept. 7. 
 
 Sept. 8. 
 
 Sept. 29 
 
 Oct.28. 
 
 Pressure in metres of mercury : 
 Pandynamometer 
 
 3 013 
 
 2 790 
 
 2.91 
 
 2 91 
 
 2 93 
 
 2.475 
 
 Indicator 
 
 3 007 
 
 2 806 
 
 2 80 
 
 2 91 
 
 2 94 
 
 2 475 
 
 Full pressure forward work, kgm. : 
 Pandynamometer 
 
 4 175 
 
 6 285 
 
 3 365 
 
 3 235 
 
 4 880 
 
 4 800 
 
 Indicator 
 
 3.990 
 
 6 148 
 
 3 120 
 
 3 135 
 
 4 930 
 
 4 702 
 
 Difference per cent 
 
 +4 43 
 
 + 2 17 
 
 +7 25 
 
 + 3.09 
 
 1.00 
 
 + 2 08 
 
 Forward work of expansion : 
 Pandynamometer 
 
 6 700 
 
 3 930 
 
 6 185 
 
 5 965 
 
 3 120 
 
 6 575 
 
 Indicator 
 
 6 621 
 
 3 875 
 
 5 885 
 
 5 985 
 
 3 192 
 
 6 697 
 
 Difference per cent 
 
 + 1.19 
 
 + 1.39 
 
 +4 85 
 
 0.33 
 
 2 30 
 
 1.85 
 
 Total forward work : 
 Pandynamometer 
 
 10 875 
 
 10 215 
 
 9 550 
 
 9 200 
 
 8 000 
 
 11 375 
 
 Indicator .... 
 
 10 611 
 
 10 023 
 
 9 006 
 
 9 120 
 
 8 122 
 
 11 399 
 
 DilTerence per cent . . . 
 
 + 2.42 
 
 +1 91 
 
 + 5 69 
 
 + 86 
 
 1.52 
 
 0.21 
 
 Indicated work : 
 Pandynamometer 
 
 9 870 
 
 9 293 
 
 8 675 
 
 8 220 
 
 7 065 
 
 6 163 
 
 
 9 606 
 
 9 101 
 
 8.131 
 
 8 140 
 
 7 187 
 
 6 187 
 
 Difference per cent 
 
 + 2.67 
 
 +2.07 
 
 +6.27 
 
 + 0.'98 
 
 1'.72 
 
 0.38 
 
 The Watt indicator should give us the pressure in the interior of the 
 cylinder and should give greater values than the pandynamometer, as the 
 friction of the packings should diminish the results obtained by the latter. 
 This difference only exists on September 29 and October 28, clearly denned 
 on September 29, but for all the other experiments is reversed. To what 
 cause of error is this anomaly to be attributed? We can only see such as 
 may be caused by the construction of the apparatus in the small dimen- 
 sions, a little piston, strong springs and too high a speed of the engine giving 
 results which have not the exactness wished.* M. Hirn has avoided the 
 oscillations due to the motion of springs, and notwithstanding all our care 
 with the indicator the pressures are too low. I insert the diagrams of 
 boiler pressure taken on September 7, in which they are notably lower 
 than given by the dynamometer. But the greatest error with these two 
 pieces of apparatus does not exceed 2.4 per cent, of the work, and the 
 indicator is valuable in practice to measure, rapidly and closely, the power 
 of an engine in motion; in certain circumstances it can serve for an exact 
 
 *Probably leaky piston of indicator more than all others, C. A. S. 
 
THE ALSA TIAN A'.V/'A' /,'/.)/ A AT.S K T< '. 2O5 
 
 analysis, as in the case of compound engines making 20 to 30 revolutions 
 per minute. In this case the uniformity of the pressure is greater than in 
 simple engines, which is favorable to the indicator, and the beam dynamo- 
 meter is useless, as it would only give the resultant force on the beam 
 without separating the effect of each cylinder. The indicator giving the 
 action of each cylinder permits as a consequence the obtaining of the 
 temperatures, density, internal heat, etc., of the working steam and to 
 follow the transfers of heat. 
 
 THE INFLUENCE OF THE INTERNAL SURFACES THE ERRORS WHICH MAY BE 
 COMMITTED BY NEGLECTING THEIR ACTION. 
 
 We believe that it will be useful before commencing this study to recall 
 briefly some preliminary notions, to enumerate some of the facts incon- 
 testable in themselves but of which the consequences have been violently 
 discussed. 
 
 Each of us knows that the cause of the movement of engines, the force 
 which acts through the medium of water reduced to steam, which sets in 
 action the different pieces of the engine, is the force of heat. Heat is the 
 direct cause of the movement. The first study naturally imposed upon us 
 is, then, the phenomena which give birth to the action of this force upon an 
 intermediate body. Actually completed this study has been made in the 
 cabinet of the physicist beyond all practical applications, and holding no 
 account of the circumstances in which the fluid is called to work. 
 
 Assuming that there is no interchange of heat between the surface of 
 the cylinders and the fluid which they contain, considering them as sim- 
 ple geometrical receptacles impenetrable to heat, is evidently contrary to 
 the truth; but for a long time it was considered that the errors of this 
 theory in practice could be neglected, and to destroy this error, to prove 
 how far it was from the reality, a series of precise experiments, so well 
 verified, of which some are given in the work of M. Him, was needed, to 
 which we shall add others which we will develop. 
 
 Without doubt it is possible in that which concerns the engine, that is 
 to say, the weight of steam used per stroke, to directly measure it, for we 
 have checked it by comparing with it the heat used and rejected by the 
 condenser, directly measured also. If the consumption had been incor- 
 rect this error would have exceeded 1 per cent. 
 
 If we desire to calculate at cut-off and the end of the stroke the weight 
 of dry saturated steam inclosed in the cylinder, it is a simple matter. 
 
 The volume swept by the piston and the densities of the steam are all 
 that is needed. We recall the diagrams taken with the indicator, or the 
 pandynamometer, where the abscissas are proportioned to the stroke, and 
 seeing how clearly the intersection of the curves with full pressure and 
 with expansion are defined when produced, this point fixes the cut-off 
 and pressure thereat. To the volume generated by the piston we add the 
 clearance volume which is occupied by steam. In the same way at the 
 end of the stroke the last ordinate gives the pressure, and the- clearance 
 
2O6 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 added to the volume generated gives the volume occupied; the same could, 
 of course, be done for any other point of the expansion. 
 
 The densities are obtained by two relations, one established by Zeuner, 
 a direct function of the pressure, 
 
 7 = 0.6061 P 0.939 3j 
 
 P being the pressure in atmospheres, the other a function of temperatures 
 and therefore indirectly of the pressures, the corresponding temperatures 
 being taken from the tables of Regnault or the formula of Roche, deduced 
 from the same tables: 
 
 1 1 
 
 and u = 
 
 (31.1 c + 0.096 t 0.00002 * 2 . 0.000000 
 
 J 
 u + w A P 
 
 where w is the volume of the unit of water and u 
 
 = i', that of the 
 We will take as examples the calculations rela- 
 
 steam 7 = - = density. 
 
 tive to the experiment of August 26, 1875, cut-off at one-fifth. 
 
 The volume swept during admission augmented by the clearance is, 
 i> = 0.1048 me., and the pressure is 41,415 kgms. per sq. metre, whence the 
 density is y = 2.5175 k. The weight of dry steam at the commencement 
 of expansion is m = 0.1048 me. x 2.5175 k. = 0.26383 k., while the feed water 
 
 per stroke was 0.2651 k., an error of - 65 i^ 0.26383 _ legs than one .half 
 
 0.2651 
 
 of one per cent. Passing to the end of the stroke, the final volume in- 
 cluding clearance is 0.490 me., the pressure 7,722 kgms. persq. m., whence 
 we have a density -}\ = 0.46096 and a weight of steam of m 1 = 0.490 me. x 
 0.46096 k. = 0.21859 k. ; the difference is 17.5 per cent. What has become of 
 this steam which has disappeared during the expansion? We hope that 
 the examination of all our experiments will make this clear to all. We 
 place them in the following tables with the differences: 
 
 TABLE VII. 
 
 
 s 
 
 00 
 
 
 
 |j 
 
 ^j 
 
 06 
 
 gj 
 
 5 
 
 
 > 
 
 > 
 
 M 
 
 si 
 
 
 
 
 
 ^ 
 
 
 
 
 , 
 
 < 
 
 3 
 
 
 
 i 
 
 O 
 
 Weight of feed 
 
 0.3065 
 
 0.3732 
 
 0.2651 
 
 0.2822 
 
 0.2240 
 
 0.2634 
 
 0. 2265 
 
 0.2982 
 
 Weight at cut-off mo 
 
 0.28656 J0.2571 
 
 0.26383 
 
 0.2866 
 
 0.1688 
 
 0.1656 
 
 0.22080.2625 
 
 Difference 
 
 9.01994:0.1124 
 
 0.00217 
 
 0.0044 
 
 0.0552 
 
 0.0948 
 
 0.0057 0.0357 
 
 Per cent 
 
 +6.5 
 
 + 30. 4i +0.83 
 
 1.5 
 
 +24.64 
 
 +36 
 
 +2.52 
 
 + 12 
 
 Weight at end of stroke, mi 
 
 0.269810.2792 0.21859 
 
 0.24496 
 
 0.1761 
 
 0.1707 
 
 0.19060.2982 
 
 
 0367 J 094.0 
 
 04651 
 
 0.03724 
 
 0.0479 
 
 0927 
 
 0.0359 
 
 Percent 
 
 +12 
 
 +25.2 
 
 + 17.5 
 
 + 13.2 
 
 + 21.38 
 
 + 35.18 
 
 + 15. 85 1 
 
 We find their marked differences nearly always less between the calcu- 
 lated and measured weights. What are the causes? Let us review 
 them. 
 
 For those who know how difficult it is to keep a metal- packed piston 
 tight, nothing is more simple than to suppose leakage. To those who 
 have not learned all the arrangements adopted, we explain that it is easy 
 
THE ALSA TIAN EXPERIMENTS, ETC. 2O7 
 
 to make a tight piston. Suspending the piston and with two cast-iron rings 
 sprung in make a tight piston. But in many cases a vertical cylinder 
 cannot be used. We believe that the piston rod should be carried at 
 both ends of the cylinder, to avoid leaks, and to use softer cast-iron for 
 the rings than for the cylinder. 
 
 What is important for us is to see that the vertical engine on which we 
 experimented possessed a tight piston, and that the natural hypothesis of 
 leakage is inadmissible. 
 
 How could we believe that these losses could give an increase over 
 that directly gained, for example on August 27, and how could the piston 
 of the same engine, working with almost the same initial pressure, vary 
 from 0.83 to 36 per cent, in a few days, August 26 to September 7 and 8.; 
 or from 25 to 36 per cent, from September 7 to September 8? There is no 
 doubt that all conditions were the same, except the superheating, and we 
 should suppose that the hotter steam would leak the easier; on the con- 
 trary the leakage is least then. 
 
 Finally, we will note some results which will support the views we 
 shall advance, knowing that the piston was tight. 
 
 It will be remarked that in the table given there is for many of the 
 experiments less steam accounted for at the end of the stroke than at the 
 cut-off. On November 28 it decreases from 30.4 to 25.2 per cent., and on 
 September 7 from 24.64 to 21.38 per cent., but in the non- condensing run 
 of October 28 it changed from 12 to per cent, at the end of the stroke. 
 
 Let us recall the progress of the steam in the engine, and see if this is 
 not an impossibility with a leaky piston. 
 
 The consumption per stroke was measured with all the precautions 
 stated above; it is then the weight of fluid which passes through the 
 cylinder, leakage or no leakage. The computation based on the volume 
 and density gives the actual quantity of steam present. Is the difference 
 lost? If so, how explain the irregularity in amount, or the excess in some 
 cases? Every one must see that this is absurd, and the hypothesis of leak- 
 age cannot be maintained. One of the most important propositions of 
 applied physics gives us the key. 
 
 When steam is introduced into a reservoir of invariable dimensions, of 
 which the surface has not everywhere the same temperature, the final pressure 
 of the steam is that which corresponds to the lowest temperature. 
 
 It was upon the facts from which this proposition is deduced, that 
 Watt based one of his best discoveries, the condenser; it will serve us to 
 explain the apparent disappearance of the steam which we have stated. 
 
 The study of the phenomena to which the action of heat upon water 
 gives birth has been made in the cabinet of the physicist, ignoring at the 
 start perturbing influences. The results thus obtained are as exact as the 
 laws from which they are derived, and if we have an error to note, at least 
 it is not from applying erroneous principles. We will not repeat too much, 
 transporting to the domain of practice the physical data relative to steam, 
 without considering the^circumstances in which the steam is called to 
 
208 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 work. Kegarding engine cylinders as simple geometric receptacles im- 
 penetrable to heat, of stating that there is no exchange of heat between 
 the steam and its surrounding metal, is from all evidence contrary to the 
 truth and has never been sustained. But for a long time it has been im- 
 plicitly considered that the various errors arising from this supposition 
 were insignificant. We shall see if it is possible to neglect them. 
 
 Kemarking, first, that when the steam is taken directly from the boiler, 
 experiments, November 28, 1873, and September 8, 1875, as is usually the 
 case, we have to do with a vapor in contact with its liquid, a so-called 
 saturated vapor, that is to say, in such a state of equilibrium that it is 
 impossible to take away the smallest quantity of heat without condensing 
 a portion, that almost always the gas itself has entrained and mixed with 
 itself a portion more or less great of the fluid from which it came. Neg- 
 lected for the two experiments which occupy us it is sometimes 5 or 6 per 
 cent, of the weight of steam introduced. In this condition it is impossi- 
 ble to add heat without evaporating a part of the fluid in suspension. 
 
 This state of saturation or unstable equilibrium of a vapor mixed with 
 its liquid in more or less quantities is such that any addition or subtrac- 
 tion of heat, how small soever it may be, brings immediately and neces- 
 sarily the evaporation or condensation of the liquid or vapor. In such a 
 condition does the mixed fluid pass from the boiler to the cylinder. 
 
 At the end of the steam pipe it finds in the steam chest the valve open 
 to the cylinder, the piston at or near the end of the stroke, and the steam 
 then fills the clearance spaces between the valve and the piston. In the 
 engine we are studying the clearance is 5 litres, very small relatively to the 
 inclosing surfaces, the cylinder head and piston have 0.5699 sq. in., and are 
 instantly filled; the steam is thus brought against an extended surface 
 which has been cooled during the preceding strokes by the expansion 
 and the exhaust to the condenser. The incoming fluid tends to impart its 
 temperature to the surfaces, and a large portion condenses, yielding its 
 heat of evaporation. By virtue of the proposition enunciated above, the 
 pressure of the fluid would fall if the communication from the boiler was 
 not open and did not permit a constant influx of steam coming to replace 
 that liquified till the moment that the interior of the cylinder has acquired 
 the temperature due the pressure. 
 
 All these phenomena are produced in the almost inappreciable inter- 
 val of time the piston is at the end of the stroke, and afterward the piston 
 uncovers fresh cooled surface which also condenses, but much less rapidly 
 than than at the first instant, for whatever be the speed of the piston at 
 midstroke the surface is much less in proportion to the inclosed volume 
 than at the end of the stroke. Finally the valve, after opening, closes the 
 steam port and the expansion commences without interrupting the action 
 of the surfaces, but the supply of heat from the boiler being ended the 
 reverse action begins, while at the same time steam is condensing on the 
 cool surface uncovered by the piston it is forming from the heads which 
 had been previously warmed. 
 
 Take the experiments of November 28 and September 8 made with 
 
THE ALSA TIAN EXPERIMENTS, ETC. 2O9 
 
 saturated steam. When the expansion commences, we have shut up 
 in the cylinder a mixture of 69.6 per cent, steam and 30| per cent, 
 water in one case; 64 per cent, steam, and 33 per cent, water in the 
 other. 
 
 This water, as we have shown before, came nearly all from direct con- 
 densation on the metal; it is deposited on the surface lining it and pos- 
 sessing its temperature and that of the steam in the cylinder. The piston 
 advances, the work of expansion demands a certain quantity of heat, the 
 fresh cool surface, uncovered, condenses more steam, the pressure falls, 
 and the water lining the surfaces instead of increasing, grows less or re- 
 mains the same in quantity, 25.2 per cent, instead of 30.4 per cent, and 
 35.19 per cent, instead of 36 per cent. This shows clearly that evaporation 
 has been produced from the surfaces which first condensed and were 
 then warmed by the steam . 
 
 During the expansion the pressure and temperature fall at each instant. 
 The cylinder surface and water covering it, keep at each instant a temper- 
 ature little higher than that of the steam, but at each instant this temper- 
 ature tends to equality with that of the mass of steam, which can only 
 occur through the medium of the deposited water evaporating at the 
 expense of its own heat, or yielding its excess to the metal, or drawing 
 from it. The steam and condensed liquid, evaporating or condensing, 
 serve as the vehicle of heat so well that at the end of the stroke we have 
 different proportions from those at the end of admission. 
 
 All this concerning the vapor of saturated steam is only the natural 
 consequence of the laws of the transmission of heat, and it is matter of 
 astonishment that it has been contested; not that the principle has been 
 denied, but it has been called insignificant in its influence from the fact 
 that gases are bad conductors of heat, and assuming the time of a stroke 
 to be too short to permit any considerable exchange of heat by radiation, 
 we have come to see that it is by direct contact and not by radiation that 
 this action can condense up to 36 per cent. In this case the error would 
 be in not taking account of the action of the surfaces: it is far from being 
 one that may be neglected, as it has been. 
 
 Let us see what happens when the steam from the boiler by a special 
 apparatus has its temperature raised about 100 C. above that of saturated 
 steam when it is superheated. Brought in that state to the cylinder, one 
 can believe that it will act as a gas, losing, without doubt, its heat, its 
 superheat, but never falling below that of saturated steam. The experi- 
 ment of September 7 shows us the contrary, that steam at 195 C. can con- 
 dense on the surface, giving 24.64 per cent, of water, and the heat of 
 evaporation of this water is given to the metal, besides the superheat, 
 which it gives first. When the expansion commences there is then only 
 saturated steam containing one-fourth water in the cylinder, most of the 
 water being on the surfaces. We are in identical conditions with the 
 experiment of November 28 and September 8. All the phenomena we 
 have already described take place. Condensation and evaporation going 
 on simultaneously in different parts of the cylinder, we find at the end 
 
210 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 of the stroke 21.38 per cent, water, showing that 3 per cent, has been re- 
 evaporated. 
 
 Between this experiment and that of November 28, one with super- 
 heated steam and the other with saturated steam, the analogy is striking. 
 It is far from being so with the others, which appear almost in part as 
 exceptions to the laws which we state. 
 
 On September 29, and August 26 and 27, the condensation at the begin- 
 ning of expansions, 2.52 per cent, and 0.83 per cent., are very small, and 
 on August 27 the steam remained superheated, since the weight of steam 
 calculated is greater than that directly measured. To give an account of 
 what passes let us recall what we said in our two experiments with satu- 
 rated steam. 
 
 At the commencement of the stroke only the clearance space is open 
 to the incoming steam. The condensation is then very energetic; we can 
 perhaps say that nearly all the steam which comes first is liquified against 
 the metallic surface, whether it be superheated to 223 C. or not, for what 
 is the heat of superheating compared with the weight of surrounding metal 
 to be warmed, and the steam first introduced into the clearance space is 
 small in amount of heat compared with the quantity to be given up. The 
 steam then introduced instantly loses its superheat and the piston begins 
 to move. We have then saturated steam in contact with a large portion of 
 water, which we unfortunately cannot directly determine, but of which we 
 can affirm the existence, for our experiments proved that the proportion 
 of water liquified was very considerable, though the time was very short, 
 that is to say, with the surface increased inversely as the weight of steam 
 introduced, we could bring it all to water. The piston then moves more 
 and more swiftly, and steam flows in from the boiler through the super- 
 heater at a temperature of 223 C. ; it meets the saturated steam, with which 
 it mixes, then the water covering the surfaces and yielding heat to it evapo- 
 rates it so well that on August 27 the whole was superheated, since the cal- 
 culated weight of steam at the end of admission was one per cent, greater 
 than the gauged weight per stroke. At other times, since the weight com- 
 puted is little greater than measured, the difference being with the errors 
 of observation, we will therefore suppose the vapor saturated but dry. 
 During the expansion we see that the liquif action, in spite of an introduction 
 of nearly half stroke, reaches 13 per cent, at the end of the stroke. A con- 
 densation takes place because the surface originally warmed did not receive 
 heat enough to prevent it. This is shown by the experiments of August 
 26 and September 29, and for November 18. 
 
 We then arrive at the last of our experiments, remarkable as we shall 
 see, in analyzing that of October 28 at high pressure non- condensing. The 
 steam heated to 220 C. is admitted for one- quarter of the stroke, expands 
 to one atmosphere and is exhausted into the air. 
 
 The clearances of the engine we are studying are 5 litres, say 1 per 
 cent, of the cylinder volume. We have in our preceding experiments 
 neglected the weight of steam shut up at the closing of the exhaust. With 
 low pressure and density the weight may be neglected, but such is not 
 
THE ALSATIAN EXPERIMENTS, ETC. 211 
 
 the case when the exhaust is at the atmospheric pressure. The compressed 
 steam rises even above the pressure of admission, and its weight is 0268 
 k., say 10 per cent, of that consumed per stroke. 
 
 It is easy to obtain, as we know the closing of the exhaust valve, the 
 pressure and volume. We then find the density a function of the pressure. 
 As this steam at this point is dry and saturated, as we shall see, its weight 
 represents the whole steam in the cylinder, a quantity put into the cylin- 
 der at the first revolution and in a manner remaining there till the engine 
 is stopped; but if this weight is constant its temperature is not, for it par- 
 ticipates along with the new steam in all the exchanges of heat of which 
 the engine is the seat. 
 
 The first action is daring the cushion and before the steam valve opens. 
 The weight calculated at the closing of the exhaust was 0.0268 k. of steam 
 that we know to be dry. If we value it again when it only fills five litres 
 of the clearance volume we find only 0.0093 k.; the balance has been con- 
 densed upon the surfaces, having a lower temperature than the compressed 
 vapor, of which the pressure is constantly increasing. Then the steam 
 valve opens, and steam rushes in and mixes with that compressed, aban- 
 doning its superheat and at the end of admission containing 12 per cent, 
 of water. The expansion commences, and at the end of the stroke the 
 deposited water has evaporated, and we have dry steam. 
 
 Summing up, the examination of each of our experiments brings us 
 to the conclusion that we can by no means neglect the action of the sur- 
 faces. To a certain depth the metal of the cylinder is penetrable by heat; 
 it plays the role of a reservoir, which receives heat during admission and 
 gives it out during expansion, or continues to receive but gives it out 
 again during exhaust. This action is shown clearly by the figures in Table 
 VII. There we find the proportion of water, which sometimes could be 
 neglected, following the conditions under which the experiments have 
 been made. Thus with steam superheated, cut-off, *, and throttle, the 
 condensations are 1 per cent., 2 per cent., and even superheated; while 
 with saturated steam, cut-off \, we find water 25 per cent., 30 per cent, and 
 36 per cent. 
 
 Above all, this series of eight experiments removes all doubts, and it 
 no longer can be denied that these exchanges of heat actually take place, 
 variable in intensity and intimately connected with the conditions of tem- 
 perature and expansion in which the engine is worked. 
 
 Cooling due to the Condenser. We can enumerate the various changes 
 to which the steam submits during its passage through the cylinder, fol- 
 lowing the exchanges of heat upon the surfaces or in the fluid inclosed by 
 them; this is not the only question, but only the first step in the road to 
 which our analysis has led us, for we can find not only the manner of 
 the distribution of heat, but the exact values by obtaining that which is con- 
 sumed on the one side by work done, and on the other the various losses 
 produced during expansion and exhaust. This study requires us to refresh 
 our memory with some of the facts established by thermodynamics, among 
 others what is meant by the internal heat U of a mixture of steam m and 
 
212 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 water M m, this value 7" is the total heat of the mixture less that of the 
 external A Pu; it is expressed 
 
 U = l-APu = M 
 
 cdt + mp 
 
 U= M (t + 0.00002 * 2 -f 0.0000003 J 3 ) + m (575 0.791 *) 
 From the elementary principles of thermodynamics the value of U 
 can only vary: 1. If the total mass M does external work, positive or 
 negative, augmenting or diminishing in volume under an external pressure, 
 and then the variation of U is proportioned to the work done. 2. lithe 
 mass, without doing work, receives or loses heat by contact with any other 
 body. 3. It' these two phenomena take place together; in this case the 
 change in U may be zero. We know the heat consumed by the work of 
 expansion, since this is given by the diagrams, and it suffices to divide the 
 number of kilogrammetres by 425 to have this quantity in heat units. 
 Putting then the internal heat [7 and the work of expansion AF" we have: 
 
 TABLE VIII. 
 
 
 CC 00 
 
 so 
 
 jj 
 
 . 
 
 00 
 
 3 
 
 . rf 
 
 
 > > 
 
 bt 
 
 A 
 
 t 
 
 i 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 fc fc 
 
 * 
 
 3 
 
 02 
 
 02 
 
 02 
 
 O 
 
 Internal heat at cut-off 
 
 
 
 
 
 
 
 
 /o G 
 
 ' 176.81 173.92 
 
 160.42 
 
 169.94 
 
 110.22 
 
 114.19 
 
 133.10 
 
 163.73 
 
 Internal heat at end of 
 
 
 
 
 
 
 
 
 stroke, U 
 
 164.44 175 79 
 
 134.37 
 
 149.42 
 
 108.26 
 
 109 .07 
 
 116.28 
 
 117.91 
 
 Difference 
 
 + 12.37 1.87 
 
 + 26.05 
 
 +20.52 
 
 +1.566 
 
 +5.12 
 
 + 16.82 
 
 14.24 
 
 Heat given to iron 
 
 23;26; 57.73 
 
 9.84 
 
 11.56 
 
 33.95 
 
 48.10 
 
 14.80 
 
 29.40 
 
 Work of Exp. A F A 
 
 16.521 15.835 
 
 15.79 
 
 9.24 
 
 14.31 
 
 13.70 
 
 7.22 
 
 14.50 
 
 In this table we also give the heat which the fraction M m gives to 
 the metal during admission, by condensing upon the surfaces increased by 
 the superheat lost by the mass M, we shall call it the heat stored by the 
 surfaces. 
 
 The internal heat U sometimes increases, or diminishes, or remains 
 nearly stationary. Let us examine each case. 
 
 1. On October 28 it is 14.24 c. more at the end than the beginning of 
 the expansion, and there has also been done 14.5 c. of external work which 
 should have been at the expense of the internal heat U and this should 
 have decreased instead of increased. This must have had heat from out- 
 side, and as the boiler is cut-off it must have come from the cylinder 
 metal, and as there is no jacket the metal must have taken it from the 
 incoming steam as previously explained. The metallic surface has con- 
 densed a portion of the steam brought from the boiler, the temperature is 
 raised and the heat penetrates the metal to a depth more or less, but which 
 matters little; it in a manner stores up heat which we can value directly 
 from the superheating, M x 0.5 (6 t) = 0.2717 k. x 0.5 (220 137.49) = 
 11.20 c.; where M = the mass, 0.5 specific heat of steam, 6 = temper- 
 ature of steam, t = temperature of saturated steam at same pressure. To 
 
THE A L > - 1 77. 1 A" K XPE /,'/.!/ K XTS, E TC. 213 
 
 this value is added the heat obtained from condensing 0.0357 k. liquified 
 during admission. 
 
 0.0357 k. x r = 0.0357 X 509.78 c. = 18.29 c. 
 11.20 x 18.20 = 29.40 c. 
 
 The work of expansion was 14.60 c 
 
 The external radiation was 2.50 c 
 
 The increase in internal heat was 14.24 c 
 
 Total 31 . 24 c 
 
 being with 1.84 c. of the other, an error of less than 1 per cent, of the 
 186.72 c. brought to the cylinder. 
 
 2. When the variations of internal heat are very small, November 28 
 and September 7, for example, we may consider it as remaining nearly 
 stationary during the expansion, and the external work done during ex- 
 pansion must have been furnished by the metal and not by the internal 
 heat; the metal must have received it during admission, but this is only a 
 portion of their action, for if we compare the amount the surface has re- 
 ceived 33.95 c. on September 7, when the work of expansion is 14.31, the 
 external radiations 2.50, leaving 18.80 c. to be accounted for per stroke 
 what has become of them? Given to the surface per stroke it is impossible 
 to have them remain there, for the temperature would rise to such a point 
 as to melt the iron under an increase of heat of 18.80 c. per stroke; they 
 must have gone to the condenser during exhaust unless the piston leaked, 
 and we showed above that it was tight. The difference we call Re, refrig- 
 eration by the condenser. It is the form first known of the action of the 
 internal surfaces, an influence long doubted and far from being admitted 
 in our day. We refer to table VII., and find there, except for the non- 
 condensing experiment, that there is from 12 to 35 per cent, of water at 
 the end of the stroke whether it entered saturated, wet, or superheated, the 
 result of all the exchanges of heat being the condensation of a greater or 
 less portion of the steam which works. This action is due to the sides cov- 
 ered with a layer of water very thin and at the temperature of the metal. 
 
 When the exhaust valve opens the steam rushes out, its pressure falls 
 rapidly, and its temperature still faster till it reaches that of the water in 
 the condenser, this, in the cylinder in spite of the smallness of the connect- 
 ing passages, and the temperature is lower than that at the end of the 
 stroke; but the surfaces of the cylinder and the water which covers them 
 are higher in temperature and the water upon them evaporates, and the 
 heat is taken from the metal in which it had accumulated during admis- 
 sion. 
 
 We see here the reverse of the phenomena during admission, the sur- 
 faces cooled during exhaust are warmed by the incoming steam which 
 they condense and evaporate during exhaust, and are again cooled. These 
 two opposing actions are the result of the same physical cause, the per- 
 meability to heat of the surfaces and that which they inclose. 
 
 If even to-day, resting upon the poor heat-conducting power of gase- 
 ous fluids and the shortness of time, it is believed that the effect of the 
 
214 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 surfaces can be neglected, the following table will again prove that it is 
 not small, that the cylinder loses during exhaust a certain number of 
 calories which are far from being small, and which do no work whatever, 
 and which even exceed those expended in the work done. 
 
 As we have seen, the values of Re are deduced from the very simple 
 relations of the internal heat, the work done and the heat given up to the 
 metal during admission. 
 
 r (Mm) = C/i U + A F. + 2.5 c. + .Re. 
 
 For September 7, 33.95 c.= 1.66 + H.31 + 2.5 c.+ .Re. 
 Re = 35.61 16.81 = 18.80 c. 
 
 But these values can be checked by a different computation, which we 
 shall follow out, knowing that Re is the heat taken from the iron during 
 exhaust. 
 
 If this heat is retained in the metal up to the opening of exhaust, it 
 will not be in the final internal heat at the end of the stroke U v , but we 
 shall find it increased by the work of expulsion, in the water rejected 
 from the condenser, and the difference will give it to us. For September 
 7, U l = 108.56 c., less that remaining after condensing 6.81, added to the 
 
 back pressure work 2.14 c. gives 103.89 c. 
 
 The heat found in condenser is 122.77 c. 
 
 Difference, Re '. 18.88 c. 
 
 The other method gave 18.80 c. 
 
 Error 
 
 0.08 c. 
 
 TABLE IX. 
 
 Rc 1st method , 15.61! 37.53 
 
 Reid. i 15.79! 35.33 
 
 Error i 0.82 2.20 
 
 Total heat per stroke |212.31 241 .36 
 
 Rc per cent, of this ! 7.80 15.60 
 
 * i 
 
 bi 
 
 
 
 <J 
 
 17.60 20.34; 18.80| 37.02 21.90 1.84 
 17.03 19.66 18.88 37.37 20.09 p, 
 
 0.57 0.68 0.08 0.35 1.81 : bl 
 181. 56 194. 36 151. 15 170. 00 154. lO^q .2 
 
 9.70 1 10.50 12.43 ! 21.76 14. 21 I 8 " 
 
 Here again it is not easy to construct any gratuitous hypothesis The 
 explanation we have given of the observed phenomena is based upon pre- 
 cise figures. It is the expression of the truth. We state from direct ob- 
 servation that the surfaces inside the cylinder are covered with a water 
 film, condensed from dry or even superheated steam. This water upon 
 the surface at a temperature greater than the exhaust is evaporated and 
 carries away from the iron a certain quantity of heat which is taken again 
 from the boiler during the admission at the next stroke. The consequence 
 is that during admission it condenses steam enough to do the work of 
 expansion, and more yet the external radiation, and the internal radiation, 
 to the condenser Re, and we can also value this Re in two ways, and the 
 
THE A L SA T1A N EXPEB1MENTS, E TC. 215 
 
 greatest error has been 1.84 or 1.17 per cent, of the heat used by the engine 
 per stroke. 
 
 v lf I insist upon this point, it is because this value Re has been not only 
 discussed but badly understood by some engineers who have found for 
 Be large negative values, and who have tried to give a reason for this 
 anomaly which was simply due to an error of observation as the only 
 negative value under exceptional circumstances, that we have in our table. 
 In a word, any negative value of Re is absurd. 
 
 3. The internal heat diminishing to the end of the stroke in this case, 
 the loss of internal heat added to that stored by the surfaces goes into the 
 work of expansion and into the external radiation, and any excess is lost 
 by the cooling by the condenser, Re, which is found as before. 
 
 The water present in the steam at the end of the stroke is the cause of 
 the phenomena we are studying. Nothing is easier than to see that the 
 figures of our table are closely connected with the final proportion of 
 water varying with it. Even the Re = 1.84 c., that is within 1 per cent, of 
 nothing for the non- condensing experiment, for which the steam is dry at 
 the end of the stroke; but this case is uncommon. If we admitted that 
 Re could be negative, it would only be saying that the condenser was 
 sending heat to the cylinder in the place of cooling the steam contained 
 in it. The second method, serving as a check, could only give Re nega- 
 tive, by having too large a value of U^ after accounting for the work of 
 expulsion, but as this internal heat is all that the steam contains after 
 having worked in the engine, we should at least find that in the water of 
 condensation. If the observation gives us a quantity too small, some heat 
 must have been lost between the cylinder and the point at which the 
 measurement is made; this can only be produced by leakage, but one could 
 easily detect that, for it could only be in the exhaust pipe, or by the loss 
 of heat by conduction from a very moderately heated pipe 70 or 80 C. 
 only. 
 
 Kesuming, this loss Re unexplained and neglected in the study which 
 denies the action of the surfaces, which may reach 22 per cent, of the 
 steam used, constitutes a useless expense which merits attention and for 
 which we should seek a remedy. 
 
 At the point to which we have attained in our analysis there should no 
 longer remain any doubt in the minds of our readers of the action of the 
 internal surfaces of the cylinder; but it may be useful before quitting the 
 subject to show how errors committed in the account of steam can extend 
 in the account of work; to see if the steam condensed during admission 
 really corresponds 1st, to the external work of expansion; 2d, to the exter- 
 nal radiation, and 3d, to the internal radiation during exhaust. 
 
 Thermodynamics establishes the general equation 
 
 dQ = Mcdt + dmr dt 
 
 which connects the quantity of heat d Q at each instant with the mass 
 m of steam and M m of water inclosed in the cylinder, the heat of 
 
216 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 evaporation r, the absolute temperature T and the specific heat c of the 
 liquid. 
 
 Since in whatever manner the heat Q has been added to or subtracted 
 from the work of expansion, A Ft Q + ( Z7 t/i) the internal heat at the 
 beginning and end of expansion is known from the temperatures and 
 pressures. Proceeding thus, the expansion curve is verified from its two 
 ends. 
 
 But for Q we find ourselves in the presence of two theories. The 
 generic theory of the engine with non-conducting cylinder, and the other 
 called the practical theory, which admits the action of the surfaces as a 
 reservoir of heat. We shall see under another form which answers to the 
 facts. Denying the action of the surfaces in stating that 
 
 dQ and Q = 0, or, AF A = U U v \ 
 Tabulating these values for all our experiments we have 
 
 
 GO 
 
 g 
 
 K 
 
 s 
 
 t-^ 
 
 00 
 
 
 
 s 
 
 
 > 
 
 > 
 
 bi 
 
 M 
 
 ft 
 
 Q 
 
 a 
 
 4J 
 
 
 g 
 
 1 
 
 1 
 
 f 
 
 4 
 
 I 
 
 6 
 
 CD 
 
 0> 
 GO 
 
 1 
 
 T _ JJ c _ 
 
 +12 37 
 
 1 870 
 
 + 26 05 
 
 + 20 52 
 
 +1 66 
 
 +5.12 
 
 +16 82 
 
 14.24 
 
 AFdC 
 
 16 52 
 
 15 835 
 
 15 79 
 
 9 24 
 
 14 31 
 
 13 17 
 
 7 22 
 
 14.50 
 
 Differences c 
 
 + 4.15 
 
 + 17.705 
 
 10. 26 
 
 11. '28 
 
 +12.65 
 
 + 8.58 
 
 9.60 
 
 +28.74 
 
 The figures of this table are, as we see, very eloquent; there is even 
 the absurdity of negative work on November 28 and October 28, for in- 
 stance, and the generic theory is untenable. 
 
 We have to return to the equation 
 
 dQ = Mcdt + dmr 
 
 mr 
 
 and sum these quantities during the expansion to find a function which 
 may be integrated. M. Hirn has arrived at it by a very natural idea which 
 he developes as follows: 
 
 "After many fruitless researches I decided to return here in the track 
 traced by the experimental method itself. As the action of the surface 
 consisted not only in taking heat from or yielding it to a gaseous mass, 
 but in partially condensing a mass of saturated vapor, or of evaporating 
 partially a mass of water in contact with it, I thought that the hypothesis 
 nearest truth would consider the active part of the surface as a portion of, 
 and at the temperature of, the water covering it. Whatever, in reality, 
 may be the temperature of the surface, the water covering and evaporat- 
 ing or condensing must be at the temperature of the saturated steam. 
 The exactness of this view has been fully sanctioned by experience. " 
 
 We can always represent, by a proper weight of water at temperature 
 T, varying by dT, the position of the mass of the surface which is warmed 
 
THE A L S-.-l 77. 1 .V EX PE111MEXTX, ETC. 217 
 
 during admission and cooled during expansion and exhaust. Let // be this 
 weight, changing by a quantity of heat, n c d T. 
 
 n ^ r d T. 
 mr 
 
 T T T* ' T. 
 
 Integrating both sides from T to T, 
 
 - CM + /to J" c ^ = ~ + constant. 
 
 - M + 
 
 Taking one of our experiments for example that of September 7, for 
 instance: 
 
 M = 0.2240 k., m = 0.1688 k., m l = 0.1761 k. 
 C m = 1.006096, T = 414.85, 2\ = 357.85. 
 r = 506.55 c., r, = 547.10 c. 
 
 0.2240 + ^ = 0.42445, u = 0.20045, and the heat yielded by this equivalent 
 weight of water is // (g qj 0.20045 (143.26 85.32) = 11.61 c. 
 
 Finally, the work of expansion deduced from this heat yielded, and 
 the difference between the internal heats at the beginning and end of 
 expansion is A F., = Q + 7 C/i = 11.61 + 1.66 = 13.27 c., while the value 
 from the diagrams is 14.31, an error of only 1.04 c., while the generic 
 hypothesis was an error of 12.65 c. 
 
 The following table gives the value of the work of expansion, calcu- 
 lated in this way, and also from the diagrams, with their difference. The 
 greatest error is 1.51 c., while the others are mostly less than 1 an error 
 of less than 1 per cent, of the total heat brought. 
 
 TABLE XI. 
 
 
 , a ! 
 > 
 
 o 
 
 
 
 > 
 
 r 
 
 
 * 
 
 ; 
 
 j 
 
 \ 
 
 3 
 
 i 
 
 5 
 
 4 
 
 i 
 
 3 
 
 t-^ 
 
 i 
 
 X 
 
 "ft 
 
 8 
 
 ! 
 
 % 
 
 t 
 
 
 K 
 
 z 
 
 
 < 
 
 ( 
 
 
 (J 
 
 OQ 
 
 
 
 OQ 
 
 
 
 Calculated A FA 
 
 16.003 
 
 U 
 
 73 
 
 Uf 
 
 01 
 
 7 
 
 73 
 
 13.27 
 
 12.75 
 
 6.78 
 
 14.52 
 
 Direct .... 
 
 .... 16 520 
 
 u 
 
 ga 
 
 15 
 
 79 
 
 i 
 
 "4 
 
 14.31 
 
 13.70 
 
 7.22 
 
 14.50 
 
 Difference 
 
 0.523 
 
 
 
 u 
 
 1 
 
 .78 
 
 1 
 
 .51 
 
 1.04 
 
 0.95 
 
 0.44 
 
 0.02 
 
 We have shown how the water deposited upon the surface at the end 
 of the stroke is partly evaporated during exhaust, how it carried with it a 
 certain quantity of heat Re, which we have called the cooling by the con- 
 denser. The cylinder in these conditions works as a boiler, producing 
 steam which goes to the condenser with a certain quantity of water en- 
 trained with it the amount of which is easily determined. The condenser 
 
218 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 here being a large edition of that used in determining priming, the 
 method of calculation is identical; that for September 7, for example, is as 
 follows: 
 
 Mean back pressure 1881 kgs. per square metre: 
 
 Temperature t corresponding 58.44. 
 
 q = 58.57 c. 7. = 624.32 c. r = 565.75 c. 
 
 The heat found in the condenser, as we have seen above, iy 122.77 c., the 
 weight of fluid per stroke is 0.2240 k., and the final temperature of the 
 water is 30.42, whence the weight of entrained water: 
 
 0.2240 (A 30.42) 122.77 
 m c ---------- 
 
 r 
 
 _ 0.2250 (624.32 30.42) 122.77 
 565.75 
 
 = 0.01815 k. - = 8.1 per cent. 
 
 The following table shows the weight of water and its proportion of 
 the steam. It is easy to see that it depends solely on the proportion of 
 water at the end of the stroke. 
 
 TABLE XII. 
 
 
 i 
 
 06 ! oo 
 
 i-l O* 
 
 8 ! S 
 
 t-' 
 
 . 1 oi 
 
 00 : <M 
 
 oo 
 
 
 o 
 
 fc 
 
 o 
 
 1 
 
 I 
 
 CO 
 
 "ft "ft 
 
 CO | OQ 
 
 
 O 
 
 Carried over wt. of water k. 
 Per cent.. 
 
 0.0150 
 4.6 
 
 0.0349 
 9.4 
 
 0.02110.0087 
 7-9 3.11 
 
 0.01815 
 
 8.1 
 
 0.0280 
 10.63 
 
 0.0025 
 1.1 
 
 Non- 
 condensing. 
 
 
 
 General Conclusions. The statement of the method used in experi- 
 menting, the checks upon the results reached, which accumulate, shows 
 us with what exactness the data have been obtained; they are indisputable, 
 as well as the results derived from them, such as the temperatures, heats 
 of evaporation, densities, etc., of the steam which proceeds from the 
 pressures following certain physical laws, which are mostly expressed by 
 empirical formulae from experiments, and are contested by no one in our 
 day. 
 
 We find ourselves with this group of exact data in the presence of two 
 theories of the engine, that which till to-day has been universally re- 
 ceived, the generic theory, which does not consider the properties of the 
 bodies and fluid which are under the influence of heat; in a word that 
 which considers the steam as working in a non-conducting cylinder. The 
 other, the theory that M. Him has so judiciously called "practical," hold- 
 ing account of the actions which take place between the steam and the 
 surrounding masses of metal. This latter theory could only be established 
 by experiments. It is the study of the working engine, and each condi- 
 tion imposed demands a new experiment and a new analysis. 
 
THE ALSA TIAN EXPERIMENTS, ETC. 219 
 
 One easily understands that it is convenient to construct from the laws 
 of physics a theory of the steam engine without the long and tedious pro- 
 cess of experimenting and computing which we have shown, and this is a 
 great part of the reason that for so long we have had the generic theory 
 only. We are far from contesting its utility, as it often points out the 
 road to be followed in seeking for the truth ; but we are obliged to discard 
 its equations when we wish to get even approximate results. 
 
 To harmonize the data and derived results, for they are only the 
 manifestations of a single cause, the action of 'heat on steam and the 
 metallic masses surrounding it while doing external work, has naturally led 
 us to condemn practically the generic theory of the steam engine. 
 
 Meanwhile it is very useful, serving as a guide in our researches, for we 
 have been led by it to make this study of the densities and volumes of the 
 steam. We have had to give up the hypothesis of leaks, the proportions 
 were so variable; but it became evident that the generic theory led to 
 errors of 36 per cent., for the weight of steam given by the product of the 
 volume and density led us to the discrepancy and to seek its cause. 
 
 In a non-conducting cylinder the difference of internal heat at begin- 
 ning and end of expansion, U U-^, should give as the work of expansion 
 AF A . Our researches showed us that heat furnished from outside from 
 the metal which had stored it up by condensing steam during admission, 
 but storing up more than was needed for the work of expansion, the ex- 
 ternal radiation and the change of internal heat, which led us directly to 
 the discovery of the internal radiation or cooling due the condenser Re, 
 which we verified by the heat found in the condenser, another form of the 
 action of the surfaces. Finally, when we wish, by the aid of the relation 
 
 dQ=McdT + dmr ^^ d t to verify the work of expansion AF A , by 
 
 supposing it in a non-conducting cylinder, we arrive at absurdities and 
 impossibilities, and on the other hand the natural hypothesis of consider- 
 ing a part of the mass of metal as water, has led M. Him to a relation 
 which verifies the work of expansion in a very remarkable manner. To 
 sum up all of our analyses shows that it is only by the practical theory 
 that we can render an exact account of the facts. 
 
 DIRECT PRACTICAL RESULTS OF THESE EXPERIMENTS. 
 
 In all that has preceded we have only had one aim, that of showing the 
 very energetic action of the metallic surfaces inclosing the steam, showing 
 that the errors found in the results of the ordinary theories of steam en- 
 gines were due to neglecting their influence. 
 
 Each of our readers will have already perceived the importance of this 
 group of experimants in designing engines. Without doubt practice has 
 led to more than one happy idea, for example, the separate condenser and 
 the steam jacket of Watt; but these improvements, to be valued in exact 
 figures, demanded a complete analysis of the engine, such as we have 
 given, for it is only then that the results cannot be gainsaid. For in ou 
 
220 STEAM USING; 07s", STEAM ENGINE PRACTICE. 
 
 day we find some engineers affirming that it is possible to work a high- 
 pressure engine as economically as with condensation; and the utility of 
 steam jackets is not yet put beyond doubt, as M. Him has well said in his 
 "Analytical and Experimental Exposition of the Mechanical Theory of 
 Heat:" "I have already said that the effect of the steam jacket has been al- 
 ternately affirmed and denied without there being any real knowledge of 
 the matter. Some say the jacket has no useful effect, others, that it gives 
 40 per cent, more work with the same steam. It is easy for us to perceive 
 the origin of so diverse statements, knowing that under certain circum- 
 stances each of them has a foundation. The essential action of the jacket 
 consists in diminishing the quantity Re of heat that the steam takes from 
 the surface to the condenser, -and in augmenting the work of expansion 
 AF d , but this action varies with the engine itself. In the study of the sin- 
 gle cylinder engine we found that in diminishing Re and increasing 
 AF A the jacket only gave little heat to steam during expansion, and that 
 the greater part of the useful heat given by the surface came from the 
 heat stored during admission. 
 
 "The study of the double cylinder 'Woolf engine shows us on the 
 contrary that it is the heat given by the jacket which increases A F {{ . 
 Before such striking differences, due to such apparently insignificant 
 difference of details, we are brought to recognize that a single cylinder 
 un jacketed engine may by reason of details better utilize the heat stored 
 in the surface during admission than any other. It is not impossible, but 
 even probable, that the relation between A F A and Re depends, for in- 
 stance, upon the proportion between diameter and stroke, or the total 
 volume to the volume at cut-off. It is evident that the jacket will give 
 results less marked upon the engine which works best without it, and 
 better when A F tl is small compared with Re. In a word, the results of a 
 steam jacket may vary within 10 to 25 per cent." 
 
 The same method of analysis gives the economy which should be 
 realized by compression. This economy stated by Zeuner can be experi- 
 mentally verified. I have found 10 per cent, for a " Woolf " engine, but it was 
 only by following all the transformations of steam in the cylinders, that it 
 was possible for me to solve this problem and to bring a first experimental 
 confirmation to the fine theorem of Zeuner. 
 
 We give, then, summing up in a last table, on opposite page, the re- 
 sults of our eight experiments, believing it easier to follow the problems 
 with their experimental conclusions, for a glance will establish the 
 relations. 
 
 The first anomaly that strikes us is that the consumption in experi- 
 ments made in 1873 is different from those of 1875 under the same condi- 
 tions. The explanation is that the engine had a new cylinder with larger 
 ports, and the exhaust was considerably earlier. Thus a modification 
 which in the limits made would seem unimportant, has produced an 
 improvement of 9 per cent. This proves that nothing should be neglected 
 in designing an engine. 
 
 We shall be forgiven if we speak'again upon the question of leakage, 
 
THE ALSATIAN EXPERIMENTS, ETC. 
 
 221 
 
 the practical importance of the subject is reason for returning to it, not 
 for this engine, which we showed to be sufficiently tight, but as general 
 conditions bearing upon the construction of engines. 
 
 TABLE XIIL 
 
 = 
 = 
 
 s ; 
 
 e*n 
 
 - 0; 
 
 5 
 
 OD 
 
 i, 
 
 Nov. 18, 1873, steam at 231 C 44.2449144.36 7.688 
 
 Nov. 28, 1878, saturated 43-7773 136.4610.026 
 
 52 
 I| I 
 
 s s s 
 
 12.00 
 25.20 
 
 Aug. 26, 1875, steam at 215 C 5 4.1415 135.77 7.002! 17.05 
 
 Aug. 27, 
 Sept. ,, 
 
 " 29, 
 Oct. 28, 
 
 223 C C. (throttled) ; 2 2.3070 125.17 8.199 13.20 
 
 195 C ! 73.9128113.08 7.126; 21.38 
 
 saturated 73.8339107-81 8.915' 35.19 
 
 steam at 220 C. (throttled) 21.7458 99.53 8.227 15.85 
 
 steam at 220 C. (non-condensing) 43.4333 78-3012.3151 ! 
 
 7.8 
 15.6 
 
 9.7 
 10.5 
 12.43 
 21.76 
 14.21 
 
 We have seen that it was impossible to attribute to leakage the steam 
 condensed during admission, and we had to conclude that the engine 
 piston was tight, the packing of the most simple kind two cast iron rings 
 turned larger than the cylinder, cut and sprung in, the two ends coming 
 together. Some builders think this lacks elasticity; for cylinders of large 
 diameter, over one metre, they prefer to have segments set out with 
 springs. Whichever are used we can always have tight packings when 
 well set up and working vertically. 
 
 It is easy to see that it is the vertical disposition which keeps the pack- 
 ing in order, for it places the segments in the most favorable conditions 
 possible. Besting on the follower, they are in a manner equilibrated dur- 
 ing motion in the same condition at all points of the stroke, with nothing 
 to counteract the lateral pressure they exert on the cylinder. They can be 
 set with little tension so as to keep them tight with the least possible 
 friction. One should not hesitate for large power with room enough, 
 almost always to be had with stationary engines, to give the preference to 
 vertical engines. 
 
 But if the problem of light pistpns has been solved for vertical en- 
 gines, the results with others is far from being satisfactory; they nearly 
 always leak, if not at first, after a short time; and it could not be otherwise 
 from the cylinder wearing oval from the weight, and the segments not 
 being carried by the piston. 
 
 They have tried to remedy this defect, more or less happily in many 
 cases, by a steam packing; but this is difficult to adjust and gives the 
 greatest wear at the ends of the cylinder. Sometimes the leaks may be 
 neglected, but we usually find the horizontal piston in default. 
 
222 STEAM USING; OS, STEAM ENGINE PRACTICE. 
 
 In the marine service they have obtained very good results with rings 
 of "anti-friction" metal, but these are often changed, and watched with 
 great care. In any case, it is indispensable to set the piston in such a 
 manner that its weight shall not be carried by the ring; to give it a rod 
 with rigidity enough to keep it from flexures and to carry it on guides at 
 each end. Unfortunately this gives a double extent of cooling surface for 
 the rod alternately exposed to steam and air. 
 
 We do not insist upon the economy of the use of superheated steam, 
 for the experiments made in 1865 by the Committee of the Society have 
 sufficiently proved that. And we only verify the results of that time. 
 There is 23 per cent, for the experiments of November 18 and 28, and 20 
 per cent, for those of September 7 and 8. 
 
 But it may be useful to rapidly enumerate the evils which are said to 
 be involved with its use; to examine its mode of action which Him has 
 so well described in his new work on thermodynamics. After having 
 analyzed the effect of the jacket, he shows that bringing into the interior 
 of the cylinder a greater quantity of heat than comes \v ith the saturated 
 steam is more energetic than surrounding the cylinder by a jacket. For 
 the heat brought by the jacket, steam condensing on the outside of the 
 cylinder has to cross the metal before it can modify the condensation 
 during admission, it can not do this rapidly enough, and we find conden- 
 sations even in the small cylinder of compound "Woolf" engines, which 
 are open to the boiler nearly full stroke; for the heat, and above all, the 
 superheat, the steam brings directly into contact with the surface, which 
 has been cooled, and we have seen in one of our experiments the expan- 
 sion commences with superheated steam. Another precious advantage, it 
 furnishes heat without condensing, giving dryer steam at the end of the 
 stroke, diminishing by that the internal radiation to the condenser Re. 
 It does not, as the jacket does, furnish its heat at the wrong time when 
 open to the exhaust. When this commences the surfaces of cylinders 
 using superheated steam are at temperatures little higher than that corres- 
 ponding to the terminal pressure, the heat lost when the steam escapes is 
 then small, while the jacket steam at the boiler temperature accelerates 
 the evaporation of the water which covers the surface, sending to the 
 condenser the most heat when it should send the least to make Re a 
 minimum. 
 
 As for the objection, raised against superheating, we shall say with 
 Him, that "Setting aside some excellent but purely theoretical works we 
 stop before the critical judgment of those who to-day call themselves 
 'practical.' We discard all opinions resting upon anything but facts and 
 the spirit of impartial investigation." We can only discuss those among 
 them which appear to have a real basis. They always say that superheat- 
 ing burns the oil which should lubricate the piston and rod packings. 
 Practice has shown that this is not true up to 230 C. (44G F.), and that it 
 is not necessary to renew the oil more frequently than with common 
 steam; that the surfaces of the cylinder are kept in as good condition 
 without cutting, but we know that the steam in the cylinder is still satu- 
 
THE ALSA TIAN EXPERT)!!: A 7 >, K T< . 223 
 
 rated, only dryer. Even if brought in at 230 C., it is impossible to sustain 
 that it will burn the oil and destroy the cylinder in which it is worked. 
 
 There are three questions of the highest interest long discussed in 
 many theoretical works which can only be answered by experiments, and 
 this solution is so natural that one is astonished to see it so long unem- 
 ployed: 1. The limits of economic expansion. 2. The effects of throt- 
 tling. 3. The use of the condenser. 
 
 Influence of Expansion. The universal opinion to-day is that the 
 greater the expansion the greater the economy of fuel, and one is brought 
 naturally to continue it to the pressure of exhaust, and the dimensions 
 given the cylinder are only limited thereby. But as M. Hirn has remarked 
 in his book, if we push expansion too far we have less upon the piston 
 than is required to move the engine and overcome friction we then do no 
 good. It is a lower limit which should never be passed, and if we wish to 
 know how far we can reduce the initial volume compared with the final 
 volume with a constant consumption our experiments give us the neces- 
 sary figures for this comparison. For a long time the question has been 
 treated differently in purely theoretical works, and we shall see to what 
 errors the generic theory has led, such as calculating from inexact experi- 
 mental data the law of expansion, for this law is only an empirical state- 
 ment of the exchanges of heat during expansion, changes which vary 
 with the conditions imposed upon the engine, and of which analyses such 
 as we have given can alone define the value and employment. 
 
 We have operated with an introduction from to \ the limits of valve 
 gear. The experiment with cut-off , it is true, was made with engine throt- 
 tled; but we will justify this later, treating of the question of throttling, and 
 we have the results of a lower pressure than with the other experiments. 
 
 Taking the figures from the preceding table, with superheated steam 
 we have 1 per cent, in favor of five expansions over seven expansions. 
 Exact values: Introduction 0.1628, consumption 7.126 k.; introduction 
 0.2570, consumption, 7.002. If we correct by 9 per cent, the experiments 
 of November 18 and 28, 1883, for the reasons given above, we see that the 
 consumption of November 18, 6.996 k., is very close to those cited. This 
 constancy of consumption holds also with saturated steam as well as with 
 superheated, a more remarkable circumstance. Thus correcting, the ex- 
 periment of November 28, 1873, 9.024 k., and for September 8, 1885, 8.915 
 k. per I. H. P., figures within 1.2 per cent. Such are the results of exper- 
 ience. Let us see, if possible without a full analysis, which is preferable. 
 
 Between the experiments of 7th September and 26th August, 1875, the 
 work varied from 113 to 136 H. P.; the water at the commencement of 
 expansion from 24.6 to 0.8 per cent. But it is objected, on the 7th Sep- 
 tember the superheating is 20 less. This is only 2.24 c. loss, or 2 ' 24 =1.4 
 
 151.5 
 
 per cent., almost exactly the difference in consumption. 
 7.126 7.002 
 
 7.126 
 
 = 1.7 per cent. 
 
224 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 Thus, while using superheated steam, we have in the one case the 
 steam dry, in the other with one-fourth water condensed on the surfaces. 
 All the functions of the engines are completely changed in these two ex- 
 periments. On August 26 the internal heat U of the steam diminishes 
 during the expansion, doing the work thereof, and as the surfaces have 
 only absorbed 9 c., during admission, a part also is radiated to the con- 
 denser in Re. While on September 7 the internal heat U of the steam is 
 nearly constant during expansion within 1.66 c., it is the surface which 
 does the work of expansion, Re having received 33.95 c. As we see a rad- 
 ical difference in the mode of transmitting heat, and in each experiment 
 the same weight of steam has produced the same work, the final result 
 being the same. In the presence of such facts, how can one say, from the 
 diagram only, for example, how much expansion should be given? To 
 establish this, the verified data are needed which we have already used. 
 
 We find that when we have carried the introduction to half stroke, but 
 throttling also to give 125 H. P. for one and 99 H. P. for the other experi- 
 ment, that for each case the consumption is not affected by the throttling, 
 which was different in the two cases. This permits us to compare these 
 experiments with those cutting off at | and f stroke, finding a saving by 
 
 O -J QQ 7 AAQ 
 
 the latter of _ -L^ = 14.59 per cent, by the greater expansion. 
 8.190 
 
 On the other hand, we see that as the work was diminished the pro- 
 portion of water at the end of the stroke was increased, as also the dead 
 loss Re. By cutting off less than % we should find a point where Re would 
 change the law and the consumption would increase, but, unfortunately, 
 the valve gear would not admit of such a trial. 
 
 We have referred to the indicated work, but this is lessened by friction 
 for the useful work, and the friction is not proportioned to the indicated 
 work. This limits the cut off to between | and 1 for the best results for 
 the single cylinder engine. 
 
 Effects of Throttling. The question of restricting the area of the 
 orifices of admission has been less often agitated than the cut off, and 
 like that can only be resolved by direct experiments, fully analyzed. 
 Before stating the results we have obtained, we will review the opinions 
 put forward. Some engineers, basing on the proposition that dry steam 
 falling in pressure without doing external work becomes superheated, 
 have asserted that throttling was beneficial by evaporating the water 
 entrained with the steam. But if they look at the second part of the 
 proposition "without doing external work," they must admit that the 
 amount of heat wrought per unit of weight is the same in either case. 
 
 Others, considering only the loss of work resulting, condemn entirely 
 all methods of regulation based on the throttle as essentially defective, pro- 
 scribing all governor throttles. They have generally attributed the 
 economy that they believed to exist in the Corliss engines, or others of the 
 class, to the rapid introduction of steam at nearly the boiler pressure. 
 
 We should attribute this radical difference of opinion to the conditions 
 in the engines observed. We have seen that in certain limits of expan- 
 
THE A LSA TIAN EXPERIMENTS, ETC, 225 
 
 sion the consumption is constant, or is increased. If we wish light work 
 we may compare cut off with cut off, and should not be astonished at 
 a difference of 14 per cent, between those of August 26 and September 7 
 with and \ cut-off, and those of August 27 and September 29 with 
 throttle. 
 
 To compare, we should not take the basis of work done, but the cut- 
 off, and the two experiments at half stroke, with the valve more or less 
 closed. 
 
 The pressure, on August 27, of steam at 223 C. was brought to 2. 307 k. at 
 cut-off, and on September 29, with a temperature of 220 C., to 1.7458 k., a 
 difference of 0.5612 k. more than half an atmosphere, and the consumption 
 
 was altered ':^ = o.3 per cent. The increase is then due to the 
 
 8.227 
 
 less expansion, as we have seen. 
 
 But here, contrary to what was found with different expansions, the 
 percentage of water is almost the same 1.5, or slight superheating, and 
 2.52 per cent, at the beginning and 13.02 and 15.85 per cent, at the end of 
 expansion, while the internal heat has decreased 20.52 c. and 16.82 c. In 
 wide enough limits, as we see, the throttle has no influence upon the con- 
 sumption. 
 
 Effect of the Condenser. As we have said, resting upon facts from 
 which the following proposition is derived, that a vapor introduced into 
 a reservoir with constant volume, of which the surfaces are not every- 
 where of the same temperature, its final pressure depends upon the 
 lowest temperature, that Watt deduced for his condenser. 
 
 The figures that we have show an economy of 43 per cent, by the con- 
 denser over exhaust to the air, a result needing little comment. 
 
 Let us see, however, as we have done, how this effects the changes 
 of heat. 
 
 At the cut-off we have 12 per cent, condensed and 29.4 c. stored heat, 
 the work of expansion requiring only 14.5 c., and Re is zero, as also the 
 proportion of final water. The excess from the surfaces has increased 
 the internal heat from 163.72 to 177.97 c. We see that the loss Re exists 
 in a different form, the exhaust carrying off an excess of 14.24 c. to the 
 air, as an increase in its internal heat, and we have lost work by the in- 
 crease of back pressure. 
 
 It is not enough that Re should be zero, as we have said, but the 
 internal heat should not increase to put the engine in the best condition, 
 and there should be no lost work, or the best vacuum should be obtained. 
 This imposes the following condition the surfaces should absorb only 
 the work of expansion. 
 
 Although all our study has shown us how little freedom we have in 
 imposing conditions upon the action of the surfaces, we believe, resting 
 upon the experiment of August 26, cut off , that by the use of super- 
 heated steam and a jacket, the loss of internal heat would have been 
 reduced from 26.05 c., and with dry steam at the end of the stroke dimin- 
 
226 STEAM USING; OK, STEAM ENGINE PRACTICE. 
 
 ished Re., but the jacket should be fed by a separate supply pipe, in order 
 not to cool too much the working steam. 
 
 There only remains for us to examine what proportion of heat given 
 has been utilized to finish these practical deductions. 
 
 When one measures the efficiency of an hydraulic motor, one divides 
 the power utilized by that furnished, but for heat engines this is not the 
 case. 
 
 Whatever be the body used, in a heat engine of maximum efficiency, 
 the efficiency depends upon the difference of temperatures between which 
 it works, divided by the absolute temperature of the source of heat. In 
 these unrealizable conditions we should get 249 H. P. for 100 calories, as 
 shown by Hirn. The best of our experiments give 135.77 H. P. for 
 
 172.79 c., or 75.8 H. P. for 100 calories ; 7 ^ = 31.5 per cent. 
 
 Av9 
 
 We see how far we are from the theoretic effect, and while stating that 
 it never can be reached, we should hope for an improvement over what 
 is practically 30 per cent. 
 
 To sum up, thanks to the numerous checks which the method employed 
 permits, we have established the considerable influence of the action of 
 the internal surfaces upon the action of steam in engine -cylinders, and 
 have shown how, by employing superheated steam without prejudice to 
 jackets, a considerable loss may be brought to a minimum, the cooling 
 due the condenser Re., and within what limits it is judicious to confine 
 expansion. 
 
 Such are the results of the series of experiments carried out under 
 the direction of M. Hirn, and our readers can judge of their great im- 
 portance.* 
 
 EXPERIMENTAL STUDY COMPARING THE INFLUENCE OF EXPANSION IN 
 SIMPLE AND COMPOUND ENGINES. A PAPER BEAD BEFORE THE IN- 
 DUSTRIAL SOCIETY OF MULHOUSE, DECEMBER 30, 1878, BY M. 0. HAL- 
 
 LAUER.f 
 
 The comparison of the many experiments made upon " Woolf " engines, 
 and the engine of M. Hirn, with superheated steam, led me to a principle 
 which has been confirmed by th,e analysis of the compound engines in 
 use in the French navy . I had stated the conclusion in a paper presented 
 to the Society on the 30th January, 1878: 
 
 One can always construct a single cylinder vertical- beam engine, steam 
 jacketed with four valves, which shall be at least as economical as the verti- 
 cal " Woolf" beam engine, for expansions from 4 to 7, if the clearance does not 
 exceed 1 per cent, of the cylinder volume. 
 
 This conclusion is based upon the total work of the engine, supposing 
 
 *These papers are given in the direct reverse of the order of their original publica- 
 tion, but perhaps not of their value. C. A. S. 
 
 tM. Keller's summary of the following experiments was given in the opening section 
 of this Chapter. C. A. S. 
 
THE ALSA TIAN EXPERIMENTS, ETC. 227 
 
 a perfect vacuum in a word, we consider the intrinsic work of the steam 
 itself. 
 
 In this memoir I have had occasion to examine the various considera- 
 tions which serve to establish the superiority of the "Woolf " system, outside 
 of the experimental domain. 
 
 These same considerations I have again found developed under a form 
 nearly identical but very marked in two works, concerning the "Woolf" 
 engines, with expansion in the small cylinder. The authors there sum up 
 what is generally admitted in favor of the "Woolf" system, which I will 
 cite literally to allow the reader to appreciate the utility of my previous 
 paper. 
 
 The first of these works was published at Kouen by MM. Thomas & 
 Powell, engineers. It contains the experiments made in June, 1876, by 
 M. H. Roland, Engineer of the Norman Association of Steam Users, and it 
 opens thus: 
 
 "Double cylinder engines, in which the steam acts successively, pro- 
 duce motive force most economically when well constructed and managed. 
 The advantage is because the small cylinder is only in communication 
 with the condenser for a moment, the large cylinder only being more con- 
 tinually so, and the first action is to withdraw a portion of the force 
 produced from the cooling action of the condenser and the internal con- 
 densation which is the immediate consequence. 
 
 "The steam arrives at the large cylinder partly expanded, and conse- 
 quently at a lower temperature than that in the jacket, and is easier 
 warmed and the condensation notably lessened. 
 
 "The employment of two cylinders permits us to carry the principle 
 of expansion to its extreme limit with the best economic conditions, the 
 force generated is divided, the efforts better carried and the differences of 
 power between beginning and end of stroke are less than in a single 
 cylinder engine; working with the same admissions there results a 
 smoother operation. Because of the vertical cylinders and perfect equi- 
 librium of the pieces attached to the beam the frictions are reduced and 
 the useful effect is very high. It is to these qualities that the long life of 
 these engines is to be attributed. We can cite some which have worked 
 thirty years and which, after modifications with comparatively little cost, 
 are in perfect order for work and consumption. 
 
 "The addition of a 'Correy Governor Expansion Gear,' assures to the 
 engines which are furnished with it a perfect uniformity of speed and 
 economic utilization under all loads." 
 
 The second work, published in 1878, in the Annual of the Society of 
 Graduates of the Schools of Arts and Trades, under the title of "Notes 
 Upon Double- Cylinder Engines," contains the results of experiments made 
 by M. Que'm, upon engines at St. Eemy, constructed by MM. Powell. 
 
 "Among the different types of engines actually in use," says M. Quern, 
 "the 'Woolf,' with two cylinders jacketed, in which the steam acts suc- 
 cessively, is that which gives the best economy in production of motive 
 force. 
 
228 STEAM USING; OS, STEAM ENGINE PRACTICE. 
 
 "In these engines the steam acts first with or without expansion in the 
 small cylinder, then with expansion in the large cylinder. The latter only 
 is in communication with the condenser. By this arrangement a portion 
 of the force produced escapes the cooling action of the condenser and the 
 internal cylinder condensation. 
 
 "Finally, because the jackets are connected with the boilers, the ex- 
 panding steam in the large cylinder is at a temperature below that of the 
 jacket, and is warmed thereby, and the cylinder condensation is notably 
 lessened. 
 
 "The employment of two cylinders permits the best realization of ex- 
 pansion, which in these 'Woolf engines can be carried to its limit. 
 
 "The difference of force between the beginning and end of the stroke 
 is less in double than in single cylinder engines; there results smoother 
 working. 
 
 "Because of the lesser difference of pressures there are less risks of 
 breaking. 
 
 "Finally, leakage of steam by the admission valve is less prejudicial 
 than in single cylinder engines. 
 
 "In 'Woolf beam engines the balancing of weights reduces the fric- 
 tion, and the useful effect is consequently high. 
 
 "We have said that their principle assures to the 'Woolf engines regu- 
 larity of speed. That is true, but the regulators which have been applied 
 for the purpose of rendering the speed uniform under variable loads have 
 been far from perfect or from giving the desired results. 
 
 "The apparatus, long employed upon single -cylinder engines, is the 
 conical governor and butterfly throttle. 
 
 "Not only is the governor throttle unsatisfactory in point of speed, but 
 its operation is bad from the standpoint of economy. 
 
 "In effect it operates upon the steampipe, opening or closing a pas- 
 sage. 
 
 "There results a throttling which produces an expansion not only 
 useless but prejudicial in the pipe and steam chest, consequently a low- 
 ering of initial pressure, which loss of force augments the consumption of 
 fuel. 
 
 "It had been desirable to put on 'Woolf beam engines a variable expan- 
 sion gear which should be easily put on, which should give these engines 
 great regularity of speed, avoid the evils of throttling, and obtain a 
 greater expansion. 
 
 "Valves with lap which had been applied for some years to these en- 
 gines were a great improvement, but the expansion was fixed and was not 
 sufficient in most cases, and moreover the throttle was retained. 
 
 "Correy's variable gear permits us to add to the advantages of the 
 'Woolf engines the removal of the throttle, retaining an economic use of 
 steam under all loads. " 
 
 Of all the foregoing considerations one only is not to be contested; it 
 is as MM. Powell say, that the efforts are better distributed and the differ- 
 ences of force between the beginning and end of the stroke are less than 
 
THE ALSATIAN EXPERIMENTS, ETC. 229 
 
 in single- cylinder engines, and the movement smoother. But it should 
 not be concluded from this long- known fact that the useful effect of 
 double-cylinder engines is high and economical. The brake experiments 
 made by the Mechanical Committee of the Industrial Society of Mulhouse 
 have proved that the friction of the engines absorbs more power in 'Woolf 
 engines than in single -cylinder engines. 
 
 I have already shown in my paper of 1878 what economy can be real- 
 ized by expansion in a separate cylinder. But the principle which I have 
 stated has raised so many contradictions that our mechanical committee 
 has deemed it prudent to hold itself in reserve when it states in these 
 terms at the close of my work: "Many times already the committee has 
 given its entire approbation to the fruitful experimental method followed 
 by our colleague, and recommends to the attention of all engineers the 
 results of the experiments contained in this work, results which appear to 
 him unattackable. On the contrary, the committee believes it should be 
 less positive in the conclusions of the author; it desires to see them con- 
 firmed by a great number of cases, and above all by varied experience in 
 the widest field." I believed it useful to renew this question with new 
 data, and more, I have added the study of an expansion, more or less, in 
 the small cylinder of the 'Woolf engine. 
 
 INFLUENCE OF EXPANSION IN " WOOLF" ENGINES. 
 
 Can there be a notable economy in cutting off in the small cylinder of 
 a "Woolf" engine and expanding, for example, 28 times? Such is the first 
 question which we shall attempt, for it is necessary to verify the con- 
 sumption reported in each of the experiments which we shall cite, and 
 this defines the degree of confidence which we shall give them. 
 
 It may be useful to recall to our readers, in the interest of the ques- 
 tion which occupies us, the passage in my memoir of 1878, bearing upon 
 this question of the influence of expansion. 
 
 The three engines where the expansion was effected in a separate 
 cylinder are ranged in order of their consumption per total horse-power 
 per hour.* 
 
 Vertical Woolf engine, 7.112 k. (15.4 Ibs.); horizontal "Woolf" engine, 
 7.290 k. (15.9 Ibs.); compound engine, 7,510 k. (16.4 Ibs.). But this is also 
 the order of expansion: Vertical "Woolf," 7 times; horizontal "Woolf," 6 
 times; compound, 5 times. 
 
 The fact that the consumption per total horse-power per hour was 
 increased by changing the cut-off from f to was also found with the 
 single cylinder engine using superheated steam. But we should observe 
 that the reduction of the volume at cut-off causes a reduction of useful 
 work by the engine, and at the same time a relative increase in the back 
 pressure work. In the engine with superheating, and above all in the 
 "Woolf" engines, this increase of back pressure work not only annuls the 
 
 *The French weights are for a Cheval de Vapeur, translated H. P. English equiva- 
 lents are in parentheses. C. A. 8. 
 
230 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 economy of a prolonged expansion, but even causes a greater expense. 
 Also the back pressure work passing from 17 to 20 per cent, destroys the 
 economy of the vertical "Woolf," when the regulating valve lowering the 
 pressure reduces the work from 347 to 267 horse-power. 
 
 The documents which will serve us in the study of expansion are: 
 
 1. Experiments made in 1877 at Munster upon a "Woolf" beam engine, 
 built by the firm of Andre' Koechlin (really the Alsatian Society of Mechan- 
 ical Constructions), and figuring in my memoir of 1878. 
 
 2. Brake Experiments by the Mechanical Committee of the Industrial 
 Society in 1876 upon a horizontal *"Woolf " by the same builder, and given 
 in the Bulletins, July, 1877. 
 
 3. Experiments made in 1877 by the Alsatian Association of Steam 
 Users upon a vertical "Woolf" engine at Malmerspach having a variable cut- 
 off in the small cylinder, by the same builder. 
 
 4. Experiments made upon "Woolf" beam engines with expansion in 
 the small cylinder, built by MM. Thomas and T. Powell, of Kouen, and 
 tried, one in 1877 at St. Kemy upon Arne by M. Quern, and the other, in 
 1876, by the Norman Association of Steam Users, this latter running the 
 shops of MM. Fauquet-Lemaitre at Bolbec. The direct results of these 
 experiments upon the Powell engines and that at Malmerspach have been 
 given me by M. H. Walther-Meunier, Engineer of the Alsatian Association 
 of Steam Users. I have checked and analyzed them. The analysis of the 
 other experiments is given in my preceding paper, as I have already 
 said: Andre' Koechlin, "Woolf" Beam Engine working at Munster, variable 
 power by throttle expansion, 7 times. 
 
 CHECK UPON CONSUMPTION, GAUGED DIRECTLY, TAKING AS A BASE THE 
 HEAT GAINED BY THE COLD WATER INJECTED TO THE CONDENSER. f 
 
 I. Forces des Chevauxs, translated horse-power, I. H.P., 347.16; revo- 
 lutions per minute, 25.25; net H. P. on brake, 303.16; mechanical effi- 
 ciency, 87.3 per cent.; proportion of back pressure work to total work, 
 17. 43 per cent.; back pressure on large piston, 0.293 k. (4.2 Ibs. per sq. in.) 
 (Boiler pressure, 67 Ibs. above atmosphere.) 
 
 Per Single Stroke. 
 
 Heat brought by dry saturated steam 0.9123 k. x 654.03 c. = 597.19 c. 
 
 " " " water entrained 0.0290 k. x 157.47 c. = 4.44 c. 
 
 " steam condensed in jackets 0.0884k. x 496.56 c. = 43.89 c. 
 
 Total heat brought to engine 645.52 c. 
 
 Heat kept by the steam leaving condenser 0.9413 x 34.25 = 32.24 c. 
 
 Qo 613.28 c. 
 
 *In the table given, on page 191 this engine is headed as "Vertical Woolf." The error is 
 in the original. C. A. S. 
 
 tAs these computations are checks, all the French units will be retained, and any- 
 thing added from other sources will be put in ( ). 
 
THE ALSA TIAN EXPERIMENT^ ETC. 231 
 
 Heat given to water of condensation, Q t 29.1072 x 18.05 = 525.38 c. 
 
 Q _Q t = 613.28 c. 525.38 c = 87.90 c. 
 
 The total work absorbed 72.79 c. 
 
 The external radiation 7. c. 
 
 79.79 c. 
 
 Instead of 87.90; or an error of 87.90^- 79.79^ l 25 cent 
 
 645.52 
 
 The heat found in the water of condensation should have been 613. 28c. 
 79.79 c. = 533.49 calories; it was only 525.38 c. consequently too little. 
 The total heat brought to the machine per stroke is 645.52 c. which to be 
 more intelligible we will transform into a weight of dry saturated steam. 
 In accounting for the work of the engine this weight will serve as a unit 
 of comparison for other engines, and will be better comprehended under 
 that form than the number of calories expended per horse -power which 
 it stands in place of. 
 
 Consumption of dry steam per stroke, 645 - 52 _j_ = 0.98698 k. ; weight 
 
 654.03 c. 
 
 of dry steam per total horse-power per hour, 7,112 k.; per indicated 
 horse power, 8.614 k.; per net horse-power, 9.864k.* 
 
 II.- -I. H. P., 267.85; revolutions per min., 25.2.; net on brake, 226; 
 mechanical efficiency, 84.3 per cent. 
 
 Proportion of back pressure to total work 20.52 per cent. 
 
 " back pressure, 0.277 k. (3.9 Ibs.). 
 (Boiler pressure, 60 Ibs. above atmosphere.) 
 
 Per Single Stroke. 
 
 Heat brought by dry saturated steam 0.7124 k. x 652.93 c. = 465.14 c. 
 
 " " entrained water 0.0238 k. x 153.74 c. = 3.66 c. 
 
 " steam to jackets 0.0794 k. x 499.19 c. = 36.63 c. 
 
 to engine 505.43 c. 
 
 " kept by steam leaving condenser 0.7362 k. x 29.10 c. = -21.42 c 
 
 Qo 484.01 c. 
 
 Heat given to, injection water 29.3406 k. x 14.3 c. = Q l = 419.57 c. 
 
 Difference ..................................................... 64.44 c. 
 
 The total work ........................................... 56.27 c. 
 
 " external radiation .................................. 7. c. 
 
 -- 63.27 c. 
 
 Error, =0 , i3 per cent. 
 
 The heat found in condenser is too small, it should have been 484.01 
 63.27 = 420.74 c. 
 
 *These weights are from feed water at Qo c. I should suggest as divisor the term 
 Qo. C. A. S. 
 
232 
 
 STEAM USING; OK, STEAM ENGINE PRACTICE. 
 
 The heat brought per stroke is 505.43 c.; it represents a consumption 
 of dry saturated steam of 
 
 505.43 
 
 = 0.77409 k. 
 
 652.93 
 
 Weight dry saturated steam per hour per total horse power 6.945 k. 
 
 " " " indicated horse power .. 8.739 k. 
 
 " " " ' " " " net " " ..10.357k. 
 
 III. I. H. P. 185.75; revolutions per minute, 25.4; net on brake, 
 145.52; mechanical efficiency, 78.3 per cent. 
 
 Proportion of back pressure work to total work, 24.1 per cent.; back 
 pressure on large piston, 0.234 k. (3.3 Ibs.). 
 
 (Boiler pressure 50 Ibs. above atmosphere.) 
 Per Single Stroke. 
 
 Heat brought by dry steam 0.5401 k. x 650.69 c. = 351.43 c. 
 
 " " " entrained water 0.0145 k. x 146.22 c. = 2.12 c. 
 
 " " steam condensed in jacket . 0.0640 k. x 504.47 c. = 32.28 c. 
 
 Total 385.83 c. 
 
 Heat kept by steam leaving condenser 0.5546 k. x 23.46 c. = 12.65 c. 
 
 Qo 373.18 c. 
 
 " found in injection water ?. 30.5688 k. x 10.56 c. = Q a = 322.80 c. 
 
 Qo Q! , 50.38 c. 
 
 The total work 38.71 c. 
 
 External radiation . . . 7. c. 
 
 45.71 c . 
 
 Error. 
 
 50.38 45.71 
 
 = 1.2 per cent. 
 
 385.83 
 
 The heat found in the condenser is too little, it should have been 
 373.1845.71 = 327.47 c. 
 
 The heat brought per stroke, b85.83c.; it represents a consumption 
 
 of 385 ' 83 = 0.59295 k. 
 650.69 
 
 Weight of dry saturated steam per hour per total H. P 7.384 k. 
 
 " " " " " " " ind. H. P 9.730k. 
 
 " " " " " " " net H. P 12.411k. 
 
 Uniting in one table the results of these three experiments, we find 
 little difference per total horse-power, only 3.7 per cent, for a change 
 from 183 to 347 horse-power. 
 
 TABLE I. 
 
 Force indicated 
 
 347 
 
 267 
 
 Steam per hour per total H. P.. kilos 
 
 7.112 
 
 6.945 
 
 Back pressure work in per cent, of total work 
 
 17.43 
 
 20.52 
 
 Steam per hour per indicated H. P 
 
 8 614 
 
 8 739 
 
 Net work in per cent, indicated work 
 
 87 30 
 
 84 30 
 
 Steam per hour per net H. P 
 
 9 864 
 
 10 357 
 
 
 
 
 I. 
 
 II. 
 
 III. 
 
 185 
 7.384 
 24.10 
 9.730 
 78.30 
 12.411 
 
THE ALSA TIAN EXPERIMENTS. ETf '. 233 
 
 We note that the cost of a total H. P. is 6 per cent, less for 267 than 
 for 185 horse-power. This economy disappears for the indicated H. P., 
 which is best for 347 H. P. 
 
 If these two sorts of consumption follow a distinct law we owe it to 
 the back pressure work, which changes to 17 per cent, from 24 per cent. 
 An analagous cause produces greater differences in the cost of a net H. P. 
 The efficiency changes between 87 and 78 per cent, because of the friction, 
 and we are not astonished to see the cost of a net H. P. differ by 20.5 per cent. 
 It is the practical loss to which we put an engine working at 185 H. P. which 
 can give 347, and is due to the back pressure and friction. But we should 
 not conclude, as is often done, that this loss is due to throttling. 
 
 The difference of 3.7 per cent, that we find in the cost of a total H. P. 
 is that due this evil influence, and is very little; or adding the slight in- 
 crease over Experiment II. 
 
 I then legitimately concluded in my last work " that we are led to 
 adopt the most simple regulator, an expansion variable by hand and a 
 governor throttle." When the variations of work are large we can, by 
 hand, without stopping the engine, change the introduction for the 
 small intermediate differences the governor acts upon the valve. It is 
 well understood that we do not here speak of engines where the force 
 varies nearly instantly, for example, to double. This disposition per- 
 mits us, as we have seen, to obtain all the benefits of a prolonged expan- 
 sion, admitting that it gives a notable economy, of which the results of 
 the following experiments will permit us to judge. 
 
 "WOOLF" BEAM ENGINE BY ANDRE KOECHLIN. AT MAiMERSPACH. 
 
 Expansion in the small cylinder. Checks on the gauged consumption 
 from the heat gained by the injection water. 
 
 E. I. H. P., 143.11; revolutions per minute, 26.2; net horse-power, 
 118.38; efficiency, 82.7 per cent.; back pressure work in per cent, of total 
 work, 18.6; back pressure on large piston. 0.181 k. (2.5 Ibs.); expansion, 28 
 times. (Boiler pressure, 67 Ibs. above atmosphere.) 
 Heat brought by dry steam 0.3479 k. x 654.03 c. = 227.53 c. 
 
 " " entrained water 0.0200 k. x 157.47 c. = 3.15 c. 
 
 " steam to jackets. . . . 0.0324 k. x 496.56 c. = 16.09 c. 
 
 246.77 c. 
 Heat kept by steam leaving condenser 0.3679 k. x 19.01 c. = 6.93 c. 
 
 " expended, Q 239.78 c. 
 
 " rejected in condenser. . 21.8781 k. x 9.06 c. = 198.21 c. 
 
 41.57 c. 
 
 " in work done 28.97 c. 
 
 " external radiation 4.6 c. 
 
 33.57 c. 
 
 Error, 417 = 3.2 per cent. 
 
234 STEAM USING; OK, STEAM ENGINE PRACTICE. 
 
 The heat found in the injection water is too small, it should have been 
 239.7833.57 = 206.21 c. 
 
 The total heat brought to the engine per single stroke is 246.77 c., it 
 
 94.fi 77 
 
 represents a consumption of drv steam of - = 0.3773 k. 
 
 654.03 
 
 Weight of dry steam per hour per total H. P 6.731 k. 
 
 " <k " " " " ind. H. P 8.273k. 
 
 " " " " " " net H. P 10.019k. 
 
 C. I. H. P. ,215. 7; revolutions per minute, 25.47; net on brake, 185.69; 
 mechanical efficiency, 86.1 percent.; back pressure work in per cent, of 
 total work, 15.6; backpressure, 0.226 k. (3.2 fts.); expansion, 13; (boiler 
 pressure, 70 fibs, above atmosphere). 
 
 Heat brought by dry steam 0.5338 k. x 654.45 c. = 349.34 c. 
 
 " entrained water 0.0304 k. x 158.88 c. = 4.83 c. 
 
 " steam to jacket 0.0449 k. x 405.57 c. = 22.25 c 
 
 Total 376.42 c. 
 
 Heat kept by steam leaving condenser 0.5642 k. x 23.21 c. 13.09 c. 
 
 " expended, Q 363.33 c. 
 
 " gained by injection water 21.9136 k. x 13.48 c. = 295 39 c. 
 
 67.94 c. 
 
 " in work done 44.83 
 
 " external radiation 4.6 
 
 Error, 67 - 24 - 49 - 43 = 4.9 per cent. 49 ' c - 
 
 376.42 
 
 Heat found in condenser should have been 363.3349.43 = 313.90 c. 
 
 " brought per single stroke 376.42 c. 
 
 376 42 
 Represents a consumption of dry steam - = 0.5751 k. 
 
 Dry steam per hour per total H. P 6.878 k. 
 
 " " " " " ind. H. P 8.149k. 
 
 " " " " " net H. P 9.465k. 
 
 F. I. H. P., 149.53; revolutions per minute, 25.93; net on brake, 124.74; 
 mechanical efficiency, 83.4 per cent.; back pressure work in per cent, 
 total work, 17.5; back pressure on large piston, 0.175 k. (2.4 Ibs.); expan- 
 sion, 25. 
 
 Heat brought by dry steam .0.3652 k. x 654.03 c. = 238.85 c. 
 
 " entrained water 0.0210 k. x 157.47 c. = 3.30 c. 
 
 " to jackets 0.0350 k. x 496.56 c. = 17.38 c. 
 
 Total. : 259.53 c. 
 
 Heat kept by steam leaving condenser 0.3862 k. x 19.43 c. = 7.50 c. 
 
 " expended, Q 252.03 c. 
 
 " given to injection water 21.2674 k. x 9.89 c. = Q l = 210.33 c. 
 
 41.70 c. 
 
THE ALSA TIA N EXPERIMENTS, ETC. 235 
 
 Heat in work done ....................................... 30.51 
 
 " " externa Iradiation ................................ 4.6 
 
 - 35.11 c. 
 
 Error, == 2.5 per cent. 
 
 Heat found in condenser should have been ....... 252.03 - 35.11 = 216.92 c. 
 
 Heat expended per single stroke .................................. 259.53 c. 
 
 Represents dry steam ................................ 259 ' 53 = 0.3968 k. 
 
 654.03 
 
 Dry steam per hour per total H. P ................................. 6.821 k. 
 
 " " " ind. H. P ................................. 8.260k. 
 
 " " " net H. P ................................. 9.898k. 
 
 D. I. H. P., 212.92; revolutions per minute, 24.83; net on brake, 
 183.67; mechanical efficiency, 86.2 per cent,; back pressure work in per 
 cent, of total work, 14.9; back pressure on large piston, 0.218 k. (3 Ibs.); 
 expansion, 13; (boiler pressure, 70 fos. above atmosphere). 
 
 Heat brought by dry steam ................ 0.5443 k. x 654.45 c. = 356.22 c. 
 
 " entrained water .......... 0.0310 k. x 158.88 c. = 4.92 c. 
 
 to jackets .................. 0.0462 k. x 495.57 c. = 22.89 c. 
 
 Total ..................... ............................. 384.03 c. 
 
 Heat kept by steam leaving condenser ____ 0.5753 k. x 27.02 c. = 15.54 c. 
 
 " expended, Q ............................................ ... 368.49 c. 
 
 " found in injection water ........ 18.0071 k. x 17.07 c. = Q l = 307.38 c. 
 
 61.11 c. 
 " in work done ........................................ 45.39 
 
 " " external radiation ................................ 4.6 
 
 49.99 c. 
 
 Heat found in condenser should have been. ...368.49 49.99 = 318.50 c. 
 Heat brought to engine per stroke . ^ ......................... 384.03 c. 
 
 Represents dry steam per stroke ............. . 38<L03 = 0.5867 k. 
 
 654.45 
 
 Dry steam per hour per total H. P ................................. 6.983 k. 
 
 " " ind. H. P ................................. 8.210k. 
 
 " " net H. P ................................. 9.517k. 
 
 This Malmerspach engine is composed of two coupled on the same 
 shaft, and experiments E and C were made on the left, while experiments 
 F and D were made on the right-hand engine. We note that three of 
 these experiments, E, F, D, check to 3 per cent, nearly, but C only to 5 
 per cent.; however the direct measurement is correct, being sensibly that 
 of experiment D. 
 
236 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 TABLE II. 
 
 
 E. 
 
 F. 
 
 C. 
 
 D. 
 
 Expansion 
 
 28 
 
 ' 
 25 
 
 13 
 
 13 
 
 I H P , Cfa. de V 
 
 143 
 
 149 
 
 215 
 
 213 
 
 Dry steam per hour per total H P . ks. 
 
 6 731 
 
 6.821 
 
 6.878 
 
 6.983 
 
 Per cent back pressure work ... .... 
 
 18.6 
 
 17.5 
 
 15.6 
 
 14.9 
 
 Dry steam per hour per ind. H. P., ks. . . 
 Per cent, of mechanical efficiency 
 Dry steam per hour per net H. P., ks 
 
 8.273 
 82.7 
 10.019 
 
 8.260 
 83.4 
 9.898 
 
 8.149 
 86.1 
 9.465 
 
 8.210 
 86.2 
 9.517 
 
 Table II. sums our four experiments. The figures which are there, 
 compared with Table I., permit us to decide if it is more advantageous to 
 work with the throttle than with a cut-off in the small cylinder, or the re- 
 verse. If one is obliged to produce from an engine too small a power, either 
 way is used, but for a careful comparison it is necessary to take if possible 
 the work with the same difference in each. For example, Malmerspach 
 Engine C and E; I. H. P., 215 and 143, being 3.2; expansions, 13 and 28; 
 and the Miinster engine, II. and III., I. H. P., 267 and 185, being nearly 
 the same ratio, expansion 7. 
 
 In dry steam per hour per total H. P. there is a difference between 
 
 6.945 k., with 267 H. P., and 7.384 k., with 185 H. P. of Zi^J^?l 5 = 5.9 
 
 7.oo4 
 
 per cent, in favor of the larger power, both obtained by throttling. 
 
 If in the second engine we pass from expansion 13 to 28, the differ- 
 
 ence in dry steam per total H. P. is 
 
 G.878 6.731 
 6.878 
 
 = 2.1 per cent, in favor of 
 
 the lesser power, 143 H. P.,; in these limits there is little to choose, for the 
 difference is within the errors of observation. 
 
 It seems as if there was an economy of 5.9 + 2.1 = 8 per cent.; on the 
 other hand, the 267 H. P. corresponds to the least consumption, 6.945 k. 
 which has been obtained by throttling 1.342 k. at the cut-off (18.8 ft>s.). As 
 
 this only differs 6>94 AzL 6 j^I? =0.9 per cent, from that for 215 H. P., 
 
 we conclude, as before, that it is unimportant. 
 
 We find ourselves here in the face of a contradiction which we can 
 elucidate later after having given the complete analysis of these engines. 
 For the time we have only to note how the heat of the injection water checks 
 the consumption and fixes the degree of confidence which we should give 
 to each experiment. 
 
 These remarks, based upon the amounts used per total H. P. per hour, 
 only refer to the work of the steam itself in the cylinder. They are not 
 affected by poor vacuum, nor the friction of the engine. The influence 
 of these two elements is only felt when we consider the indicated work 
 and the net work. Thus, in passing from experiment E 143. H. P., ex- 
 pansion, 28, to experiment C, 215 H. P., expansion 13, the consumptions 
 differ per indicated H. P. 1.5 per cent, and per net H. P. 5.5 per cent. 
 
THE ALSA TIAN EXPERIMENTS, ETC. 237 
 
 This difference is in the reverse order of that for the total H. P.; it shows 
 that practically there is 5.5 per cent, loss in changing from expansion 13 
 to 28. These same causes, back pressure work and friction, have brought 
 for the Munster engine, with fixed expansion 7, stronger effects increased 
 to 16J per cent, when throttle causes the work to fall from 267 H. P. to 185 
 H. P. 
 
 Upon the same engine at Malmerspach, and before the application of 
 expansion gear, there had been made an experiment, with the object of 
 defining the economy realized, which we will check as before. 
 
 B. Indicated H. P., 201.64 H. P.; revolutions per minute, 24.18; net 
 on brake, 172.80 H. P.; mechanical efficiency, 85.6 per cent.: back pressure 
 work in per cent, of total work, 16.4; back pressure, 0.235 k. (3.3 Ibs. per 
 sq. in.); expansion, 6; (boiler pressure, 67 tt>s. above atmosphere). 
 
 Heat brought by dry steam 0.5823 k. x 654.03 c. = 380.84 c. 
 
 " entrained water 0.0312 k. X 157.47 c. = 4.91 c. 
 
 to jackets 0.0330 k. x 496.56 c. = 16.38 c. 
 
 Total ' 402.13 c. 
 
 Heat kept by steam leaving condenser 0.6135 k. x 22.73 c. = 13.94 c. 
 
 " expended, Q 388.19 c. 
 
 found in injection water 22.6234 x 14.50 = Q l = 328.04 c. 
 
 60.15 c. 
 
 " in work done 44.14 c. 
 
 " " external radiation 4.6 c. 
 
 48.74 c. 
 
 Error ^^8-^ = 2.8 per cent. 
 
 The heat found in injection is too small; it should have been 388.19 
 48.74 = 339.45 c. The heat per stroke is 402.13 c. 
 
 It represents 4?i4i = 0.6148 k. dry steam. 
 654. Oo 
 
 Dry steam per hour per total H. P 7.402 k. 
 
 a ind. H.P 8.847k. 
 
 " net H.P 10.301k. 
 
 The result of this analysis, if we consider the engine in good order 
 when the experiment was made, which we will suppose to be the case, 
 compared with experiment C. 
 
 B. Expansion 6, C. Expansion 13. 
 Per Total H. P. - 7 ' 402 ~ 6 ' 878 = 7.1 per cent. 
 
 Per Ind. H. P. 8 ' 847 ~ 8 ' 149 = 8 per cent. 
 
 Per Net. H. P. 10 ' 3 ^ 66 = 8 per cent. 
 
 by C over B. 
 
238 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 I have valued the friction by the brake experiments of our Mechanical 
 Committee. We shall see how far MM. Powell have obtained the same co- 
 efficients upon their "Woolf " engines. 
 
 HORIZONTAL "WOOLF" ENGINE, BY ANDRE KOECHLIN, TEIED WITH BRAKE 
 BY THE MECHANICAL COMMITTEE. 
 
 I. I. H. P., 130; revolutions per minute, 39.37; net on brake, 112.08; 
 mechanical efficiency, 86.1 per cent.; back pressure work in per cent, of 
 total work, 20.06; back pressure, 0.253 k. (3.5 Ibs.); expansion, 6; (boiler 
 pressure, 53 Ibs. above atmosphere). 
 
 Heat brought by dry steam 0.2286 k. x 652.46 c. = 149.15 c. 
 
 " entrained water 0.0079 k. x 152.17 c. = 1.20 c. 
 
 to jackets .0.0263 k. x 500.29 c. = 13.16 c. 
 
 Total 163.51 c. 
 
 Heat kept by steam leaving condenser . . .0.2365 k. x 26.20 c. = 6.19 c. 
 
 expended, Q 157.32 c. 
 
 " found in injection water 14.6202 k. x 9.20 c. = 134.50 c. 
 
 22.82 c. 
 
 " in work done 17.45 c. 
 
 14 in external radiation 3.5 c. 
 
 20.95 c. 
 
 Error ? 2 ^p = 1.13 per cent. 
 
 The heat gained by injection should have been 157.32 20.95 = 136.37 c. 
 Heat per single stroke 163.51 c. 
 
 Represents dry steam per stroke. . . .^-r 1 = 0.2506 k. 
 
 652.46 
 
 Dry steam per hour per total H. P 7.290 k. 
 
 " ind. H. P 9.120k. 
 
 " net H. P 10.563k. 
 
 There is only 1.3 per cent, difference between the consumption per 
 total H. P., and that in experiment B; while for the indicated H. P. there 
 is 3 per cent, in the other direction. 
 
 II. I. H. P., 181; revolutions per minute, 39.67; net on brake, 161 H. 
 P.; mechanical efficiency, 89 per cent.; back pressure work in per cent, of 
 total work, 17.3; back pressure, 0.295 k. (4.1 Ibs.); boiler pressure, 54 Ibs. 
 above atmosphere). 
 Heat brought by dry steam 0.3134 k. x 652.69 c. = 204.55 c. 
 
 " entrained water 0.0134 k. x 152.96 c. = 2.05 c. 
 
 to jackets 0.0271k. x 499.73 c. = 13.54 c. 
 
 Total 220.14 c. 
 
 Heat kept by steam leaving condenser 0.3268 k. x 23.90 c. = 7.81 c. 
 
 " expended, # 212.33 c. 
 
THE ALSA TIAN EXPERIMENTS, ETC. 239 
 
 Heat rejected in injection water 14.5795 k. x 12.58 c. = 183.41 c. 
 
 212.33 183.41 = 28.92 c. 
 
 " in work done 24.12 
 
 " in external radiation . 3.50 
 
 27.62 c. 
 
 Error ^-^ = 0.6 per cent. 
 
 Heat brought per stroke 220.14 c. 
 
 Represents dry steam per stroke 22(U4 = 0.3372 k. 
 
 Dry steam per hour per total H. P 7.328 k. 
 
 " ind. H. P 8.878k. 
 
 " net H. P 9.975 k. 
 
 The consumption per total H. P. is nearly the same as for the preced- 
 ing experiment, but for the indicated and net H. P. there is a marked 
 improvement by better efficiency, and less proportion of back pressure 
 work, although the vacuum is not so good. This is a fact which we have 
 many times noted. 
 
 Finally, that we may not lack generality in our conclusions of the 
 influence of expansion in the small cylinder, I will give the results of ex- 
 periments made upon the engines constructed by MM. Powell. Little 
 different from the preceding, they are designed with a view to an early 
 cut-off in the small cylinder, but distinguish themselves by the excellent 
 vacuum, 0.100 k. of back pressure (1.4 Ibs.) on the large piston. We shall 
 see that for 19 expansions the consumption per total H. P. is nearly the 
 same as that we have found for 13 and 28 times. 
 
 "WOOLF" BEAM ENGINE BY POWEIZL, WORKING AT ST. BEMY. EXPERIMENT 
 
 BY M. R. QUEM. 
 
 Expansion, 19 times; indicated horse-power, 137; revolutions per min- 
 ute, 24.503; net on brake, 107.88 H. P. ; mechanical efficiency, 78.7 per cent.; 
 back pressure work in per cent, of total work, 9.9; back pressure, 0.102 k. 
 (1.4 tbs.); (boiler pressure, 70 Ibs. above atmosphere). 
 
 Heat brought by dry steam 0.3222 k. x 654.42 c. = 210.85 c. 
 
 " entrained water 0.0110 k. x 158.80 c. = 1.75 c. 
 
 to jackets 0.0412 k. x 495.62 c. = 20.42 c. 
 
 Total 233.02 c. 
 
 Heat kept by steam leaving condenser 0.3332 k. x 27 c. = 8.99 c. 
 
 expended, Q 224.03c. 
 
 found gained by injection water. . . .10.8511 k. x 18.24 c. = 197.92 c. 
 
 26.11 c. 
 
 in work done 29.79 
 
 in radiation 4.50 
 
 34.29 c. 
 
24O STEAM USING; OE, STEAM ENGINE PRACTICE. 
 
 Heat found should have been 224.03 34.29 = 189.74 c. 
 
 " per single stroke 233.02 k. 
 
 Kepresents dry steam. . . 233 ' 02 = 0.356 k. 
 
 654.42 
 
 Heat per hour per total H. P 6.840 k. 
 
 " ' " " ind. H. P 7.591k. 
 
 " " " " net H. P 9.702k. 
 
 The consumption per total H. P. of this last experiment, made with 
 19 expansions on the Powell engine, is exactly between the two experi- 
 ments E and C, made upon the Koechlin engine with expansions of 13 
 and 28. These three consumptions are 6.731 k., 6.840 k., 6.878 k., differing 
 among themselves 1.4 per cent. Upon different engines they prove, be- 
 tween expansions of 13 and 28, how little the effect of expansion is upon 
 the good work of steam. The very good vacuum of the Powell engine, 
 0.102 k. (1.4 Ibs.), gives it practically a marked superiority over the 
 Koechlin engine, expanding 13 times, a circumstance remarkably excep- 
 tional for a "Woolf " engine, which demands justification by more numerous 
 experiments ; it loses only 9.9 per cent, in back pressure work, in the 
 place of 15 per cent., and we shall not be astonished to find there is 
 
 S ' M9 ~^A ? ' 591 = 6 ' 8 P er cent - in its favor ' 
 o.i^y 
 
 We had wished to join in these results those that were obtained by M. 
 H. Koland upon the same Powell engines, tried at Bolbec ; but, as we 
 shall see, their exactitude leaves much to be desired, and we only remark 
 the excellent vacuum and the error committed, 9.9 per cent., which causes 
 us to set aside the results.* 
 
 The results of the experiments C, D, E, F, of the Malmerspach engine, 
 built by Andre Koechlin, with that on the St. Remy engine, by 
 Powell, prove to us the small influence of an expansion from 13 to 28 ; in 
 these limits the total H. P. cost varies only 2 per cent, in one case from 
 the other. This fact acquired, we shall seek to render an account of the 
 following anomaly, which we have already noted, between cost per total 
 H. P. of experiments B and C of the Malmerspach engine. With expan- 
 sion 6 and 13 we found 7 per cent, difference, which should represent the 
 economy of expansion 13 ; but we should be too high, for it does not 
 agree with the figures of C, D, E and F, nor with the results of the 
 Powell engine. This difference is more when compared with II., 267 H. 
 P., of the Munster engine, expanding 7 times, which gives the least cost 
 as 6.945 k. per total H. P., when the expansion 13 gives 6.878 k. difference 
 of 1 per cent. We shall see if analysis will show us the cause of this irre- 
 gularity. 
 
 ANALYSIS OF EXPERIMENT III. 
 
 Account of heat and cooling by condenser, per stroke: 
 
 Weight of fluid in small cylinder 0.5776 k. 
 
 " dry steam at cut-off 0.4553 k. 
 
 "The computations are not transcribed. C. A. S. 
 
THE ALSATIAN EXPERIMENTS, ETC. 
 
 TJU, 31 T 7 
 
 Weight of water at cut-off (21.17 per cent.) 0.1223 k. 
 
 " " " entrained 0.0145 k. 
 
 " " " condensed up to cut-off 0.1078 k. 
 
 Heat given to iron 0.1078 k. x 516.77 c. = 55.70 c. 
 
 Weight of fluid in large cylinder 0.5830 k. 
 
 " " dry steam at end of stroke 0.5399 k. 
 
 " " water at end of stroke (7.39 per cent.) 0.0431 k. 
 
 Internal heat at the end of admission, U 290.12 c. 
 
 Internal heat at the end of stroke, U 1 322.01 c. 
 
 U Uj 31.89 c. 
 
 Heat given by jacket 32.28 c. 
 
 " " " condensation in small cylinder 55.70 c. 
 
 " furnished during expansion 56.09 c. 
 
 " absorbed by total work of expansion 35.61 c. 
 
 " radiated externally 7.00 c. 
 
 42.61 c. 
 
 56.09 42.61 = 13.48 c. = Re, cooling by condenser being 3. 5 per cent, 
 of the heat brought to the engine. 
 
 The final internal heat compared with the heat gained by the injection 
 water furnishes a check on Re. 
 
 Internal heat at end of stroke, U l 322.01 c. 
 
 Heat of back pressure work + 12.29 c. 
 
 " remaining in cushion 14.92 c. 
 
 after condensation 12.65 c. 
 
 306.73 c. 
 " gained by injection water 322.80 c. 
 
 Re 16.07 c. 
 
 The other method gave Re 13.48 c.; the error is only 
 
 *1^? ?*-* 
 
 ANALYSIS OF EXPERIMENT II. 
 
 Account of heat and cooling by the condenser: 
 
 Weight of fluid in small cylinder 0.7697 k. 
 
 " '' dry steam at cut-off 0.6603 k. 
 
 " " water at cut-off (14.21 per cent.) 0.1094 k. 
 
 " " " entrained 0.0238k. 
 
 " " " condensed 0.0856 k. 
 
 Heat given to iron up to cut-off. . . 43.40 c. 
 
242 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 Weight of fluid in large cylinder 0.7650 k. 
 
 " " drv steam at end of stroke. . . . 0.7237 k. 
 
 " water at end of stroke (5.39 per cent.) 0.0413 k. 
 
 Internal heat at cut-off, TJ 415.57 c. 
 
 Internal heat at end of stroke, U l 432.15 c. 
 
 U Uj 16.58 c. 
 
 Heat furnished by jacket 36.63 c. 
 
 " " condensation in small cylinder 43.40 c. 
 
 " during expansion 63.45 c. 
 
 " absorbed by work of expansion 49.80 c. 
 
 " external radiation 7. c. 
 
 56.78 c. 
 
 Re = 6.65 c.; jRc is a loss of 6 = 1.32 per cent, of total heat per 
 
 505.43 
 
 stroke furnished engine. 
 Check on Re: 
 
 Internal heat at end of stroke, U^ 432.15 c. 
 
 Back pressure work + 14.53 c. 
 
 Heat remaining in cushion 15.13 c. 
 
 Heat remaining in fluid after condensation 21.42 c. 
 
 410.13 c. 
 Heat gained by injection water 419.57 c. 
 
 Re 9.34 c. 
 
 The other method Re = 6.65; 9j4 t ~ 6 ' 65 = .55 per cent. 
 
 505.45 
 
 ANALYSIS OF EXPEBIMENT I. 
 
 Account of heat and cooling by condenser, Re: 
 
 Weight of fluid in small cylinder per stroke 0.9847 k. 
 
 dry steam at cut-off 0.8254 k. 
 
 water at cut-off (16.17 per cent.) 0.1593 k. 
 
 water entrained . . . .0.0290 k. 
 
 water condensed at cut-off 0.1303 k. 
 
 Heat given to iron at cut-off 65.25 c. 
 
 Weight of fluid in large cylinder 0.9691 k. 
 
 dry steam at end of stroke 0.9051 k. 
 
 water at end of stroke (6.6 per cent.) 0.0640 k. 
 
 Internal heat at cut-off, U 525.79 c. 
 
 " end of stroke, U, 543.19 c. 
 
 UU... 17.40 c. 
 
THE ALSA TIAN EXPERIMENTS, ETC. 243 
 
 Heat furnished by jacket 43.89 c. 
 
 by iron small cylinder 65.25 c. 
 
 during expansion 91.74 c. 
 
 Heat in total work done during expansion 62.85 c. 
 
 Heat of external radiation 7.00 c. 
 
 .Re = 21.89 c.; per cent, of heat furnished, 3.38. 
 
 Check on Re.: 
 
 Internal heat at end of stroke, U l 543.19 c. 
 
 Back pressure work + 15.37 c. 
 
 Heat remaining in cushion 14.60 c. 
 
 fluid after condensing 32.24 c. 
 
 511.72 c. 
 
 Heat gained by injection water 525.38 c. 
 
 Re 13.66 c. 
 
 The other method gave 21.89: 
 
 Error, ^'^^ M = 1-26 per cent. 
 
 HOBIZONAL "WOOLF," 130 H. P.; EXPANSION, 6. 
 
 Account of heat and cooling by condenser Re: 
 
 Weight of fluid per stroke in small cylinder 0.2500 k. 
 
 dry steam at cut-off 0.2220 k. 
 
 water at cut-off (11.2 per cent.) 0.0280 k. 
 
 water entrained . . . . . 0079 k. 
 
 water condensed at cut-off 0.0201 k. 
 
 Heat given to iron at cut-off 10.30 c. 
 
 Weight of fluid in large cylinder 0.2506 k. 
 
 dry steam at end of stroke . . .0.2372 k. 
 
 water at end of stroke (5.34 per cent.) 0.0134 k. 
 
 Internal heat at cut-off, U . .137.85 c. 
 
 * end of stroke, U t 141 .48 c. 
 
 Z7o #i .3.63 c. 
 
 Heat furnished by jacket 13 . 16 c. 
 
 " iron above.. . 10.30 c. 
 
 during expansion 19 . 83 c. 
 
 Heat absorbed during expansion, total work 14.38 c. 
 
 " external radiation 3 . 50 c. 
 
 Re = 1.95 c.; per cent. Re of total heat furnished, 1.19 c. 
 Check on Re. : 
 
 Internal heat at end of stroke 141 .48 c. 
 
 Back pressure work . . +4 . 38 c. 
 
244 STEAM USING; OK, STEAM ENGINE PRACTICE, 
 
 Heat retained in cushion ......................................... 7.43 c. 
 
 in condenser ....................................... 6 . 19 c. 
 
 132. 24 c. 
 gained by injection water ............... .............. ....... 134.53 c. 
 
 Re ............................................................ 2 . 26 c. 
 
 Error, 2 .-2_lj.?5 = 0>1 per cent. 
 
 HOKIZONTAL "WOOLF," 181 H. P.; EXPANSION, 6. 
 
 Account of heat and cooling by condenser Re: 
 Weight of fluid per stroke in small cylinder ..................... 0.3459 k. 
 
 " dry steam per stroke at cut-off ......................... 0.3082 k. 
 
 " water per stroke at cut-off (10.8 per cent.) ..... ........ 0.0377 k. 
 
 " water entrained ............................ . ........... 0.0134 k. 
 
 " water condensed at cut-off ........ . .................... 0.0243 k. 
 
 Heat given to iron ............................................... 12.24 c. 
 
 Weight of fluid in large cylinder ................................. 0.3478 k. 
 
 " dry steam at end of stroke ............................. 0.3324 k. 
 
 " water at end of stroke (4.5 per cent.) .................. 0.0154 k. 
 
 Internal heat at end of admission, U ............................ 192.68 c. 
 
 " stroke, U, ........................ ........ 198.80 c. 
 
 U Ui ........................... . .......................... 6.12 c. 
 
 Heat furnished by jacket ......................................... 13.54 c. 
 
 " iron above.. . 12.24 c. 
 
 during expansion 19.66 c. 
 
 " absorbed during expansion, total work 14.50 c. 
 
 " lost, external radiation 3.50 c. 
 
 Re - 1.66 c.; per cent, of heat furnished - 6 ~ = 0.75. 
 
 220.14 
 
 Check on Re. : 
 
 Internal heat at end of stroke, U^ 198.80 c. 
 
 Back pressure work +5.10 c. 
 
 Heat retained in cushion 10.03 c. 
 
 " " condenser. . . 7.81 c. 
 
 186.06 c. 
 gained by injection water 183.41 c. 
 
 Re 2.65 c. 
 
 The other method ga"Ve 1.66 c. 
 
 6 * - 45 per cent - 
 
THE A L SA TIA JV EXPERIMENTS, ETC. 245 
 
 The error appears to be 2 ' 6 ^ 1>66 = 1.9. As the check on Re gives 
 2.65, the injection has gained less heat than was rejected.] 
 
 Experiment II. on the Munster engine and I. on the horizontal engine 
 differ little as to proportions of final water and heat lost by the cooling 
 
 due the condenser Re. We have stated the difference 7 A 2 ^~l|^ 7 i 5 = 4.7 
 
 7.290 
 
 per cent, between the cost of a total H. P. It is due, part to the differ- 
 ence between 6 and 7 expansions, but more to the strong compression in 
 the vertical engine which partially annuls the effect of the clearance. 
 
 The experiments, of which the analyses will follow, do not offer the 
 precision of the two preceding series, of which the consumption checks 
 within one per cent., nearly; also we will neglect the weight of fluid in the 
 clearance when establishing the internal heats U l and U and the differ- 
 ence U 9 ZTi. I did not proceed thus until I had rendered an account of 
 the error which is committed. 
 
 With engine 267 H. P., U 7, = 16.58 c.; when the clearance is taken 
 into account, U U l = 19.80 c.; when it is neglected, there is an error of 
 
 X M - 1M = 0.6 per cent. For the horizontal engine, U U, will be 
 505.4)3 
 
 3.44 in place of 3.63 c., an error of M 3 + 3._44 = 0.1 per cent. 
 
 loo. 51 
 
 Our second manner of procedure is thus justified above all in the 
 practical experiments which check within 3 per cent, only, but we add 
 again that this approximation is very satisfying and conducts us to some 
 very remarkable results. [The error is always one way, and the compari- 
 sons are very accurate] . 
 
 MALMERSPACH ENGINE, EXPERIMENT A 201 H. P., EXPANSION, 6. 
 
 Heat account and cooling by condenser Re: 
 
 Weight of fluid in small cylinder 0.6135 k. 
 
 " " dry steam at cut-off 0.5350 k. 
 
 " " water < " (12.8 per cent.) 0.0785 k. 
 
 " " " entrained 0.0312k. 
 
 " " " condensed at cut-off 0.0473k. 
 
 Heat given to iron 24.09 c. 
 
 Weight of fluid in large cylinder 6135 k. 
 
 (< " dry steam at end of stroke 0.5429 k. 
 
 " "water " " " (11.5 per cent.) 0.0706k. 
 
 Internal heat at end of admission, U 384.86 c. 
 
 " stroke, ^ 327.69 c. 
 
 U U,.. 7.17 c. 
 
246 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 Heat furnished by jackets 16.38 c. 
 
 " iron. . 24.09 c. 
 
 during expansion ... 47.64 c. 
 
 " absorbed " " by total work 34.60 c. 
 
 " lost by external radiation 4.60 c. 
 
 Q A A 
 
 Rc = 8.44 c.. being - = 2.1 per cent, of the heat furnished. 
 
 402.13 
 
 Check on .Rc: 
 
 Internal heat at end of stroke, t/i 327.69 c. 
 
 Back pressure work 8.61 c. 
 
 Heat retained after condensation. . . . 13.94 c. 
 
 322.26 c. 
 Heat gained by injection water 328.04 c. 
 
 .Rc 5.68 c. 
 
 By the other method 8.44 c. 
 
 Error, 8 ^^ 8 = 0.7 per cent. 
 
 This engine differs from the horizontal one by a larger proportion of 
 terminal water, 11.5 per cent, in place of 5.3 per cent. The cost of a total 
 
 H. P. is also greater by I^^J' 290 = 1.5 per cent. 
 
 The expansion 6 and the conditions of regulation are the same, but 
 we see that the jackets are not working in the same manner as those of the 
 horizontal engine condensing 10 per cent, of the steam, while the second 
 is only 5.1 per cent.; there is then from this fact a loss which changes the 
 internal heat and increases the heat lost by the cooling due the con- 
 denser. 
 
 Compared with Experiment II., 267 H. P., Experiment B gives us 4.7 
 per cent, less condensation in the jackets and a less cushion, which brings 
 
 the Malmerspach engine "~ ' ' = 6.1 per cent, worse than the 
 
 Munster engine. All these considerations should indicate where the 
 economy really is rather than to a large expansion commencing in the 
 small cylinder. 
 
 MALMEKSPACH ENGINE EXPERIMENT C., 215 H. P.; EXPANSION 13. 
 
 Account of heat etc., per stroke: 
 
 Weight of fluid in small cylinder 0.5642 k. 
 
 " " dry steam at cut-off 0.4303 k. 
 
 " " water " " (23.7 per cent.) 0.1339 k. 
 
 " " " entrained.. 0.0304k. 
 
 condensed at cut-off. . 0.1035 k. 
 
THE ALSATIAN EXPERIMENTS, ETC. 247 
 
 Heat given to iron 51.30 c. 
 
 Weight of fluid in large cylinder 0.5642 k. 
 
 " " dry steam at end of stroke 0.4632 k. 
 
 " " water " " (17.9 per cent.) 0.1010 k. 
 
 Internal heat at end of admission, U 283.55 c. 
 
 " " " " stroke, U 1: 282.31 c. 
 
 U H! 1.24 C. 
 
 Heat furnished by jackets 22.25 c. 
 
 " iron.. 51.34 c. 
 
 Total heat furnished during expansion 47.83 c. 
 
 Heat in total work done 38.79 c. 
 
 " " external radiation 4.60 c. 
 
 Re - 32.44c, being 31 ' 44 = 8.3 percent, of the entire heat furnished. 
 376.42 
 
 The check on Re is not as exact as before; but as we remember that in 
 this experiment the error was 5 per cent., while in the others it was about 
 2.5 per cent., this is not surprising. 
 
 Internal heat at end of stroke, U 283.31 c. 
 
 Back pressure work 8.48 c. 
 
 Heat retained in condensed water . . . . 13.09 c. 
 
 277.70 c. 
 " gained by injection water 295.39 c. 
 
 Re 17.69 c. 
 
 The other method gave 31.44; 31 ' 44 ~ 17 - 69 = 3.7 per cent, error. 
 
 376.42 
 
 The other experiments were much closer. 
 
 The cut-off in the small cylinder being much earlier, it is desirable 
 to calculate the internal heat Z7 2 at the end of the stroke in the small 
 cylinder. 
 
 Weight of fluid in small cylinder 0.5642 k. 
 
 " dry steam at end of its stroke 0.4901 k. 
 
 Internal heat, Z7 2 305.84 c. 
 
 MALMERSPACH ENGINE EXPERIMENT. D., 213 H. P.; EXPANSION, 13. 
 
 Account of heat, etc., per stroke: 
 
 Weight of fluid in small cylinder '. 0.5753 k. 
 
 " " dry steam at cut-off. . 0.4328k. 
 
 " water (24.7 per cent.) 0.1425 k. 
 
 " " entrained.. 0.0310k. 
 
 ; condensed at cut-off ... 0.1115k. 
 
 Heat given to iron 55.31 c. 
 
248 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 Weight of fluid in large cylinder 0.5753 k. 
 
 " " dry steam at end of stroke 0.4628 k. 
 
 ' water (19.5 per cent ) 0.1121 k. 
 
 Internal heat at cut-off, U 286.44 c. 
 
 " end of stroke, U t 283.20 c. 
 
 U U l 3.24 c. 
 
 Heat furnished by jacket 22.89 c. 
 
 " iron.. 55.31 c. 
 
 during expansion 81.44 c. 
 
 " taken by total work of expansion 39.04 c. 
 
 " " " external radiation 4.06 c. 
 
 .Re 37.80 c. 
 
 or? on 
 
 Per cent, of entire heat -' ~ n = 9.9 per ce nt. 
 
 OO4:.' 10 
 
 The check on Re: 
 
 Internal heat at end of stroke, ^ 283.20 c. 
 
 Back pressure work 7.97 c. 
 
 Heat retained after condensation . . . 15.54 c. 
 
 275.63 c. 
 
 " gained by injection water 307.38 c. 
 
 Re 31.75 c. 
 
 Error, 37 ' 80 ~ 81 - 7S =1.6 per cent. 
 384.03 
 
 At the end of stroke in small cylinder, C/",: 
 
 Weight of fluid 0.5753 k. 
 
 " dry steam 0.4928 k. 
 
 " water (1.43 per cent.) 0.0825 k. 
 
 Internal heat at end of stroke in small cylinder 308.79 c. 
 
 These two experiments with the same expansion, 13, are made, one 
 on the right engine, the other on the left, and give results little different. 
 
 Thecost per total H. P. of C. is only 6 983 g ~ 6 878 = 1.5 per cent. 
 
 6 . 982 
 better than D. 
 
 The proportions of water for C and D respectively are at cut-off, 23.7 
 and 24.7 per cent.; at the end of the stroke in small cylinder, 13.1 and 
 14.3 per cent., and at the end of stroke in large cylinder, 17.9 and 17.5 per 
 cent., following nearly the 1J per cent, difference. 
 
 Re differs 8.3 and 9.9 per cent., only 1.6 per cent. 
 
 We will take for comparison experiments B, expansion 6; C, expansion 
 13, and E, expansion 28, all made on the left-hand engine. 
 
 The last two experiments throw light upon a fact which should, above 
 all, attract our attention. It is the great evaporation which takes place in 
 the small cylinder during the first expansion. The introduction at full 
 pressure has been carried to nearly half the stroke of the small cylinder, 
 
THE ALSA TIAN EXPERIMENTS, K Tf. 249 
 
 without preventing the evaporation of 10 per cent, of the fluid originally 
 condensed, and a rapid augmentation of the internal heat of steam, 22.29 c. 
 for C, and 22.35 D. (U E^). We see also that during expansion in the 
 large cylinder a portion of the vapor, existing at the end of the stroke, 
 in the small cylinder, has been condensed about 5 percent., and the in- 
 ternal heat of the steam (UfUJ diminished 23.53 c. C, and 25.59 c. D. 
 
 The number of calories returned by the internal heat during the stroke 
 of the large piston is nearly the same as the work of expansion, 25 c. in 
 the large cylinder; we could then conclude that the steam jacket has done 
 nothing in this second expansion, in a word, does not perform its office. 
 
 Arriving at this conclusion will be denying one of the elementary prin- 
 ciples of physics, for we know that the greater the difference of tempera- 
 tures the more energetic the transfers of heat. During the stroke of the 
 large piston the temperature of the steam in the jackets is much higher 
 than that of the steam in the cylinders, it is then during this period that the 
 transfer of heat should be best made that the jacket should furnish more; 
 this is really what takes place. 
 
 We will show later in treating of expansion how the passage of calo- 
 ries is made, and what is their occupation; we shall see that this pheno- 
 menon, which appears at first to be entirely abnormal, explains itself 
 naturally; we shall see it become a very simple consequence of the princi- 
 ple of the transmission of heat which it seems at first to contradict. 
 
 We give then, to terminate the series of analyses, the two experiments 
 E and F, made with expansions 28 and 25. We regret that we cannot join 
 thereto the analytical story of the engine at St. Eemy. Its consumption 
 checks within 3 per cent, nearly, and it agrees closely with the Malmers- 
 pach engine, but it lacks the exact elements necessary in the indicator 
 diagrams. 
 
 MALMERSPACH ENGINE, EXPERIMENT E., 143 H. P.; EXPANSION, 28. 
 
 Account of heat, etc., per stroke: 
 
 Weight of fluid in small cylinder 0.3679 k. 
 
 " dry steam at cut-off. . 0.2214 k. 
 
 water (40 per cent.) 0.1465 k. 
 
 " entrained.. 0.0200k. 
 
 ; condensed 0.1265 k. 
 
 Heat given to iron 63.08 c. 
 
 Weight of fluid in large cylinder 0.3679 k. 
 
 *' dry steam at end of stroke .. 0.3030k. 
 
 water (7.6 per cent.) 0.0640 k. 
 
 Internal heat at end of admission, U 157.44 c. 
 
 " " " stroke, U^ 183.32 c. 
 
 U U,.. 25.88 c. 
 
250 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 Heat furnished by jacket 16.09 c. 
 
 " " iron above.. 63.08 c. 
 
 Total heat furnished during expansion 53.29 c. 
 
 Heat in total work " " " 29.35 c. 
 
 " " external radiation 4.60 c. 
 
 Re 19.34 c. 
 
 Relatively to entire heat furnished ' - = 7.8 per cent. 
 
 H l. I ( 
 
 Check on Re: 
 
 Internal heat at end of stroke, U v 183.32 c. 
 
 Back pressure work 6.63 c. 
 
 Heat retained in condensed steam. . 6.99 c. 
 
 182.96 c. 
 
 Heat gained by injection water ; 198.21 c. 
 
 Re ! 15.25 c. 
 
 Differing from the other "^t^Jf'" = L6 P er cent - 
 
 CTg at end of stroke small cylinder: 
 
 Weight of fluid 0.3679 k. 
 
 " " steam end of stroke. . 0.2976 k. 
 
 " water ' (19.1 per cent.) 0.0702 k. 
 
 Internal heat " " U z 186.81 c. 
 
 MALMERSPACH ENGINE, EXPERIMENT F., 149 H. P.; EXPANSION 25. 
 
 Account of heat, etc., per stroke: 
 
 Weight of fluid in small cylinder 0.3862 k. 
 
 dry steam at cut-off 0.2471 k. 
 
 " water " " (36.1 per cent.) 0.1391 k. 
 
 " entrained 0.0210 k. 
 
 " " condensed 0.1181 k. 
 
 Heat given to iron 58.83 c. 
 
 Weight of fluid in large cylinder 0.3862 k. 
 
 dry steam at end of stroke .3180 k. 
 
 water " " (17.8 per cent.) 0.0682k. 
 
 Internal heat at cut-off, Z7 172.10 c. 
 
 " " end of stroke, t/i 192.45 c. 
 
 U U, 20.35 c. 
 
 Heat furnished by jacket 17.38 c. 
 
 " " iron above.. 58.83 c. 
 
 Total heat furnished during expansion 55.86 c. 
 
THE ALSATIAN EXPERIMENTS, ETC. 251 
 
 Heat furnished during external radiation 4.60 c. 
 
 absorbed by total work 29.93 c. 
 
 RC 21.33 c. 
 
 91 *\ 
 
 being _- = 8.2 per cent, of entire heat per stroke. 
 259.53 
 
 Check on Re: 
 
 Internal heat at end of stroke, 7^ 192.45 c. 
 
 Back pressure work 346.43 c. 
 
 Heat retained after condensation - 7.50 c. 
 
 191.38 c. 
 " gained by injection water 210.33 c. 
 
 Re 18.95 c. 
 
 Differing 21.3318.95 =o.9percent. 
 2o9.53 
 
 Weight of fluid in small cylinder 0.3862 k. 
 
 " dry steam at end of stroke 0.3147 k. 
 
 " water (18.6 per cent.) 0.0715 k. 
 
 Internal heat, Z7 2 197.46 c. 
 
 These two experiments E and F, give results which accord perfectly 
 with expansions 28 and 25. We should also note in order the profound 
 modifications to which the steam is submitted when the expansion is 
 changed from 13 to 28. The internal heat which in C diminishes 1.24 c. 
 during the expansion, changes to an increase of 25.88 c. for E, while for D 
 and F there is for one a diminution of 3.24 c., and for the other an acces- 
 sion of 20.35 c. between expansion 13 and 25. 
 
 INFLUENCE OF VARIABLE EXPANSIONS UPON THE WORKING OF STEAM IN 
 "WOOLF" ENGINES THEIR UTILITY FROM THE POINT OF VIEW OF CON- 
 SUMPTION. 
 
 The exposition of the very complex phenomena which absorb us, the 
 study of which should be made as clear and as easily grasped as possible 
 induces us to give in Table III., page 254, a summary of the principal 
 results which form the basis of our discussion. 
 
 The action of the iron upon the fluid which it incloses is so well estab- 
 lished, and the result of Hirn's labors on heat engines is such that it 
 naturally follows that variable expansions modify the nature even of the 
 work of steam. We introduce into the cylinders different weights of 
 steam, different quantities of heat, therefore it is not astonishing to see 
 during expansion variations in the direction and amount of the changes 
 of heat. But that which should be useful in practice is the experimental 
 determination of these changes, followed by the results of their analysis, 
 and their justification. We shall fall perchance on facts at first inadmis- 
 sible like those we found for expansion 13, and find the natural explana- 
 tion in the most profound study of the phenomenon, a purely physical 
 study. 
 
252 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 The paradoxical fact which we will recall is presented then in experi- 
 ments C and D upon each of the Malmerspach engines working with 
 expansion 13, and with full pressure more than one-half of the stroke in 
 the small cylinder. During the expansion in the large cylinder a portion 
 of the steam existing at the end of the stroke of the small piston is con- 
 densed Experiment C., 4.8 per cent.; Experiment D., 5.2 per cent. The 
 internal heat has diminished Z7 2 U^= + 23.53c. and 25.59. But the work 
 of expansion in the large cylinder has demanded and absorbed 24.9 c. and 
 25 c., that is to say, nearly the same amounts for this period of work. 
 There was no heat furnished from outside; the jacket appears to have 
 yielded nothing; was it not working during this period of expansion? 
 
 This hypothesis is inadmissible, for we have seen that it contradicts a 
 well-known principle, of physics relative to the transmission of heat; the 
 exchange of heat across the sides of the cylinder should be more rapid 
 with the greatest difference of temperature between the two surfaces, 
 for one of them is in contact with the jacket steam at boiler pressure, 
 and the other possesses the temperature of the cylinder steam at a much 
 lower pressure. We would remark that the difference of temperature is 
 not the only factor which can accelerate this transfer; the layer of water, 
 which covers the internal surface has also its influence; it augments the 
 rapidity with which heat is brought to the inner surface; it provokes a 
 proportionally greater action in the jacket, as we will prove by the figures 
 of Table III. 
 
 How can it be, then, since the jacket is in the best possible condition 
 to furnish heat, that it appears to be inactive? This apparent anomaly 
 has a very simple cause the action of the surfaces at the commencement 
 of the stroke of the large piston. 
 
 I established in my paper of 1878 that at the first tenth of the stroke 
 of the large piston, a moment when the dry steam is nearly equally divided 
 between the small and large cylinders, there is one-half at least of the 
 fluid deposited as water upon the walls of the large cylinder; if we mark, 
 then, that at the end of the first tenth of the stroke of the large piston the 
 large cylinder contains more water than steam, that is, that the first part 
 of this stroke the condensation has been very considerable, while the 
 cooling due the condenser has been only 1.3 per cent. loss. 
 
 The same fact is presented for experiments and D in stronger pro- 
 portions yet; since by the cooling due the condenser the heat taken from 
 the iron during exhaust is 8.3 and 9.9. It is upon this water which covers 
 the surface that the jacket acts, and since it cannot evaporate a sufficient 
 quantity, the internal heat at the end of the stroke is less than at the end 
 of the stroke of the small piston. Such is the particular circumstance 
 which the five experiments made upon the coupled engines at Malmers- 
 pach present to us; it is not the first time we have had occasion to remark 
 it. I have already noted it in the experiments made on the Munster en- 
 gines, 1876, working with little compression in the clearance spaces. 
 
 Let us indicate the modifications which a variable expansion brings 
 to the transformation of steam and to the action of the jacket. 
 
THE AL8A TIAN EXPERIMENTS, ETC. 253 
 
 The three experiments, B, C and E, the first the result of a very slight 
 expansion in the small cylinder where the steam has been admitted 
 nearly all the stroke, with a total expansion of 6; the Indicated H. P. only 
 reaches 201. To do this feeble work the steam pressure had to be throt- 
 tled, for we have with thirteen expansions and full pressure, a work of 215 
 H. P., a greater load in spite of the less introduction. 
 
 In these conditions the weight of steam condensed during admission 
 is small, 12.8 per cent., less the water carried, over 5 per cent. = 7.8 per 
 cent, deposited upon the surface. The jacket yields proportionately less 
 heat, for it only condenses 5 per cent, of the steam brought to the engine. 
 We remark that in spite of the condensation which took place at the first 
 stroke of the large piston, the proportions of water are within 1.3 percent, 
 the same at the beginning and end of the expansion, and that the cooling 
 due the condenser is small enough, 2.1 per cent, of the entire heat brought 
 the engine. In spite of these conditions, which appear advantageous 
 enough, the consumption per total H. P. is 7,404 k., while those of C and 
 E are 6,878 k., and 6,731 k., that is to say, by 7 and 9 per cent. 
 
 Is this the gain realized by the expansion? This we shall see. 
 
 Experiment C, with introduction of half stroke in the small cylinder, 
 presents a condensation of 23.7 5 = 18.7 per cent, during the admis- 
 sion, but the proportion of water which is found is partly evaporated 
 
 Passing to the results of experiment E, expansion 28, this modification 
 is much more marked. The proportion of water at the end of an intro- 
 duction of stroke of small cylinder is 40 per cent., and 21 per cent, 
 evaporates during the first expansion, and the internal heat increases 
 29.37 c.; in short, we see that as the expansion in the small cylinder is 
 increased the transfers of heat are increased. 
 
 On the other hand, the action in the large cylinder follows another 
 law. We have seen, that with nearly full stroke introduction in the small 
 cylinder, experiment B, the internal heat remains nearly stationary, 
 diminishing by only 7.17 c., between the.ends of the stroke of the small and 
 large pistons. The same fact is found in experiment E, 28 expansions. 
 But we have seen above that the intermediate experiments C and D show 
 us a considerable fall of internal heat, a fall sufficient to furnish to the 
 work of expansion the number of calories which it requires. Here there 
 is, then, as in the small cylinder, an increase in the transfers of heat from 
 experiments B to C, but a decrease follows the minimum, which appears 
 toward expansion 13, to which is due that the internal heat t/i is increased 
 relatively to the final internal heat U. 2 of the small cylinder. 
 
 By considering only the phenomena of the total expansion from the 
 moment that it commences in the small cylinder to the end of the stroke 
 in the large cylinder, we see that the differences of internal heat are con- 
 tinually reversed; this shows that the transfers of heat are greater and 
 greater as the expansion is increased. Thus for experiment B the final 
 internal heat U l is 7.17 c. less than U at the end of admission; for experi- 
 ment C this difference is only 1.24 c.; while for experiment E the differ- 
 ence is reversed, and the final internal heat is 25.88 c. larger than at the 
 
254 
 
 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 1 
 
 HORIZONTAL 
 
 ENGINE. 
 
 rH 
 1 1 
 
 CN O rH t^ O 
 CO CO D O O O 00 
 
 <N 
 
 
 So 
 
 jjrs-irv'-g 
 
 ^0 
 
 
 rHO 
 
 
 
 (35 5O <N O O 
 
 * CO 
 
 CO O 
 
 
 82 
 
 t-^ 05 CO rH 
 3 'M CO rH rH rH 
 
 to co 
 
 
 
 MUNSTER ENGINE. 
 
 M 
 
 rnSiScOcloOOrH 
 
 3 
 
 
 83 
 
 t-t-t-t-CCt-OSCNOOO 
 
 -* rH CO rH 
 
 | 
 1 
 
 
 ri 
 
 ri 
 
 cfslcoiooS 
 
 88 
 
 
 jgg 
 
 -ft***'*"* 
 
 10 ec 
 
 
 O rH 
 
 a 
 
 islcoSco^rH 
 
 8S 
 
 
 28 
 
 CO <M C- rH ^ rH S 
 
 T 
 
 
 eo co 
 
 MALMERSPA.CH ENGINE. 
 
 M 
 
 -* O 00 O CO rH X 
 
 o IH t-^ o co ia o > us <M' 
 
 d + 
 
 
 co' oi 
 
 d 
 
 on o 10 
 
 00 tO rH rH ( t- t- rH O5 <M CM W5 -* CO 
 
 rH rH rH CO <M rH rH <M CS) CO 
 
 H 
 
 JH CO OS OOt-OJ-* 
 t- O t- O >O O rH <O 00 CO M< CO CO 
 
 co co to co co <M o ko oo" o O5 t-' id 01 co' oi t^ 
 
 :::::::::::: : : 
 
 PER SINGLE STROKE. 
 
 1 ! ; !- ii 
 
 ii-l-i i; 
 
 : fe :S g : :^bb ; : 
 
 of expansions 
 cated horse-power on pistons 
 steam per hour per total horse-power, kilo 
 pressure work, per cent, of total work. . . 
 steam per hour per indicated horse-power, 
 tianical efficiency, percent 
 steam per hour per net horse- power, kilog 
 cent, priming or water entrained 
 steam condensed in jackets 
 water in steam at cut-off 
 end small cylinder 
 end large cylinder 
 age of internal heat during expansion Uo 
 " in small cylinder 77o 
 
 in large cylinder 17% 
 JKC or cooling due condenser in calories 
 Re in per cent, total heat brought to engine. . . 
 
 lUIHril i 
 
THE ALSATIAN EXPERIMENTS, ETC. 255 
 
 end of admission. The qualities of heat furnished by the jacket, the 
 proportions of steam condensed in the small cylinder during admission, 
 which transforms the inner surface of the cylinder into a reservoir of heat, 
 and the internal heat, all these values are intimately connected together, 
 as we have already said many times: they depend the one on the other; 
 and are in some sort the various manifestations of the same thing. We 
 also see them changing in kind the condensation in the jacket and small 
 cylinder during admission; increase with expansion, furnishing thus 
 more heat in proportion than is measured by the expansion and consider- 
 ably more. 
 
 This law appears to be general in "Woolf" engines, not only with 
 variable expansions, but for fixed expansions and variable powers obtained 
 by throttling the steam. The experiments upon the "Munster" engine 
 show us that with 185 H. P. the final internal heat U^ is 31.89 c. more than 
 the initial internal heat U at the end of the admission, while this differ- 
 ence is only 17.40 c. for the experiment with 347 H. P. Meanwhile exam- 
 ining more closely the figures of the two series of experiments at 
 Malmerspach and Munster, we discover one point which merits being 
 brought into light. The differences U U t are close together for experi- 
 ments I. and II., 17.40 c. and 16.58 c., more alike than B and C, 7.17 c. and 
 1.24 c., which differ much from E, 25.88 c. This leads us to believe that 
 in the limits of expansion and throttling of experiments I., II., B, C, the 
 differences of internal heat remain nearly stationary; they only commence 
 to increase rapidly when we go beyond 13 expansions and throttle below 
 267 H. P. of experiment II. 
 
 Finally we see for all the "Munster" eperiments of 1877 the final 
 internal heat U^ is greater than the initial internal heat U . If this fact is 
 the reverse of what was produced upon the same engine in 1876, it is that 
 at that time the engine was differently regulated, the compression in the 
 clearance space being much less. The increase of compression diminished 
 greatly the lead influence of the clearance. This realized an economy 
 and the result betrays itself by an increase in the final internal heat. It is 
 not necessary to believe that this is always the characteristic sign of a 
 better working of the engine. We have in effect experiments C and E, of 
 which the consumptions vary only 2.1 per cent, when the differences of 
 internal heat U U are 1.24 c., and 25.88 c. On the other hand, experi- 
 ments I. and II. give U U^ 17.40 c. and 16.58 c. . and meanwhile the 
 consumptions vary 2.4 per cent., the least being for experiment II. 
 throttled to 267 H. P. 
 
 Summing the effects of a greater total expansion, commencing in the 
 small cylinder and placing parallel to them the effects of the same kind 
 produced by throttling, reducing the indicated work in the same propor- 
 tion that the change of expansion did, the best experiments are C and E, 
 expansions 13 and 28, and experiments II. and III., with the same expan- 
 sion 7. The indicated horse-powers are nearly in the same ratio, 
 215 - 267 , 
 
256 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 "When the expansion changes from 13 to 28 the condensation in the 
 small cylinder during admission increases from 23 to 40 per cent.; the 
 heat given to the iron is increased. The jacket also gives a little more 
 heat, but the greater part of this heat is restored in the small cylinder 
 itself. It furnishes there more heat than is needed for the work of expan- 
 sion, for the evaporation of 10 and 21 per cent, of the water deposited 
 upon the surface. This is proved by the increase of internal heat by 22 
 and 29 c. at the end of the stroke of the small piston. In the same circum- 
 stances, throttling with nearly full admission in the small cylinder, 
 experiments II. and III., the communication between cylinder and boiler 
 is open also. The proportions of steam condensed only vary from 14 to 
 21 per cent., which show a very different kind of working. The steam 
 passes then to the large cylinder; it is in this that the transfers of heat 
 are found which did not take place in the small cylinder. The internal 
 heat is increased during the expansion 16 and 31 calories; 9 and 14 per 
 cent, of the water is evaporated. On the other hand, with 13 and 28 
 expansions, commenced with cut-offs at one-half and one- fifth in the small 
 cylinder, the reverse action took place. The energetic changes are less 
 in the large cylinder than in the small cylinder. This is a fact which will 
 serve us later in comparing single and double cylinder engines. 
 
 Experiment B seems to contradict this law, for the internal heat 
 diminishes 7 calories from the beginning to the end of the expansion, and 
 the proportions of water are nearly the same, 12 and 11 per cent. This is 
 due to two causes, the influence of a distribution with very little com- 
 pression, and the less effect of the jacket, where there is only 5 per cent, 
 of the weight of the steam condensed; but it is none the less true, that the 
 heat furnished by the jacket has been during the expansion in the large 
 cylinder. Before attacking the question from the practical industrial side 
 let us seek to render an account of the advantages of expansion, consid- 
 ering only the work of the steam; that is to say, abstracting imperfections 
 of vacuum and frictions of the engine. 
 
 In principle, theory has conducted M. G. Zeuner to recommend very 
 prolonged expansions; M. G. A. Hirn, on the contrary, considering the 
 practical conditions imposed on the engine, adopts moderate expansions. 
 We shall see that these two opinions which appear contradictory are both 
 sanctioned by our experimental researches when properly analyzed. The 
 work of M. Zeuner supposed that the cylinder was non-conducting, and 
 the closed cycle perfect, but it is clear that the cycle is interrupted by the 
 conditions in which we place our "Woolf " engines. We should naturally 
 expect to see the benefits of expansion considerably diminished by the 
 transfers of heat to the internal surfaces; but the difference is great enough 
 to allow us to affirm the law. 
 
 Commencing by comparing the cost per total H. P. of experiments I., 
 II., III., C and E, experiment I., 347 H. P., expansion 7, has been made 
 with an initial pressure near that of C and E, but it is experiment II., 
 made with a lower pressure, which presents the minimum cost, 6.945 k. 
 This is due to some particular circumstance which we have not been able 
 
THE AL SA TIAN EXPERIMENTS, ETC. 257 
 
 to bring out;* and which causes the cooling due the condenser to be least 
 for experiment I., while to compensate this action we take for the con- 
 sumption for 7 expansions the mean of I. and II., 7.028 k. The cut-offs for 
 expansions 7, 13 and 28 are in the ratio of 1, ^ and J; the costs are 7.028 k., 
 6.878k.; and 7.731k., a decrease of 2.2 per cent.; or an expansion four 
 times as great procures 4.4 per cent. gain. The law of M. Zeuner is then 
 found verified experimentally in "Woolf " engines. It is to be noted that 
 the disturbing effect of the surfaces does not reverse this law, for in the 
 method used to determine the cost this influence is fully accounted for. 
 The cost for experiment III., 185 H. P., 7 expansions, is 5 per cent, inferior 
 to that of I., II. Whence we conclude that it should have a marked advan- 
 vantage of about 10 per cent, in replacing by more expansion a too great 
 lowering of initial pressure. 
 
 We have left to one side experiment B, made with the same engine as 
 C and E, for two motives, which appeared to us ought to exclude it; in the 
 first place the jacket did not yield enough heat. And then the valves are 
 set with a compression not sufficient to help the clearance. But every- 
 body knows that one of the effects of a prolonged expansion is to exten- 
 uate in part the pernicious effect of the considerable waste spaces of the 
 "Woolf" engine. 
 
 The cost of a net H. P. brings us the modifications of the more or less 
 perfect vacuum which falsifies the preceding law; thus experiments I. and 
 II. differ in reverse order 1.4 per cent., and C and E, which vary 1.5 per 
 cent. 
 
 Finally the cost of a net H. P., upon which the combined effects of 
 poor vacuum and friction give the following results: Experiment I., 9.864 
 k.; II., 10.357k.; III., 12.411 k.; C, 9.465k.; E, 10.019 k. The least is exper- 
 iment C, expansion 13; it only varies 5 per cent, from E, expansion 28, 
 and 4 per cent, from I, expansion 7, full pressure. This justifies the prop- 
 osition of M. Him. There remains to calculate the dimensions to be given 
 to engines and the frictions which result, the foreknowledge of the inter- 
 rupted cycle and the moderate expansion. In a word he says: "For 
 reasons of practical fact the steam engine with broken cycle and 
 moderate expansion works better, notwithstanding its faults, than the en- 
 gine working with the perfect cycle," without regard to first cost. 
 
 Experiment III., 185 H. P., expansion 7, pressure much reduced, differs 
 24 per cent, from the corresponding experiment, E, expansion 28, full 
 pressure. This difference is due to the vacuum and friction. If the hori- 
 zontal engine occupies the last rank, it owes it to the small compression, 
 its vacuum and frictions. Its inferiority is only 2 per cent, relatively to 
 the corresponding experiment II. and B, and 6 per cent, referred to the 
 experiment with greatest power, I., 347 H. P. 
 
 The practical consequences of these collected researches upon the 
 "Woolf" engine may be stated thus: 
 
 1. From the total point of view, considering only the best utilization 
 of the heat brought to the engine, we can reduce the maximum work with 
 5 kilos, pressure. For example (70 pounds), expansion 7 to i, by 
 
258 STEAM USING; OH, STEAM ENGINE PRACTICE. 
 
 diminishing the initial pressure, and there results a loss of 5 per cent, of 
 the cost of a total H. P.; while by reducing the introduction in the small 
 cylinder, we can gain 4 per cent, of the same cost. 
 
 2. When one is obliged to reduce the work one-half, whether because 
 of a change of load or because used with water power, we can make a 
 practical gain of 10 per cent, at least by replacing the throttle by a varia- 
 ble cut-off in the small cylinder. We suppose it well understood that the 
 back pressure work remains the same in the two cases. The friction is 
 naturally the same, since the work is the same. 
 
 3. The engine working near its full load, it is possible to vary the 
 work 10 per cent, more or less, without any notable change in the eco- 
 nomic re'gime due to expansion or change of pressure. There follows the 
 disposition we have already remarked; an expansion variable by hand and a 
 governor throttle valve, which will answer all requirements, and is the 
 most simple and durable. 
 
 INFLUENCE OF EXPANSION IN SINGLE CYLINDEE ENGINES. 
 
 We proceed with this study as we did with the "Woolf" engines, 
 checking first the consumption of the different experiments of which we 
 make use. These experiments are upon four valve engines with and with- 
 out jackets. They consist first of experiments made upon a horizontal 
 "Corliss" by the Mechanical Committee of the Industrial Society in April 
 and May, 1878. 
 
 The experimental process is that of M. Hirn often described in our 
 Bulletins, and the direct results of observations are given in that for Dec., 
 1878. I have checked and analyzed them, and added the experiments 
 made in 1873 and 1875, upon the vertical engine with four valves of M. 
 Hirn, without a jacket, and with and without superheating; it was also 
 tried with a brake in 1865 by the Mechanical Committee. We will com- 
 mence with the results of the "Corliss:" 
 
 Corliss Engine by Berger, Andre & Co. 
 
 Tried with brake by the Mechanical Committee; check on direct con- 
 sumption. 
 
 I. Indicated on piston 105 H. P. 
 
 Revolutions per minute 50.41. 
 
 Net on brake 92 H. P. 
 
 Mechanical efficiency, per cent 88 
 
 Back pressure work in per cent, total work 10 
 
 (2.1 Tbs.) 0.148 k. 
 
 Expansion 11 
 
 (Boiler pressure 68 Ibs. above atmosphere.) 
 
THE ALSATIAN EXPERIMENTS, ETC. 25S 
 
 Per single stroke : 
 
 Heat brought by dry steam 0.1306 k. x 654.24 c. = 85.44 c. 
 
 " " water entrained 0.0043 k. x 158.18 c. = 0.68 c. 
 
 " steam to jacket.. ... .0.0094 k. x 496.06 c. = 4.66 c. 
 
 " to engine 90.78 c. 
 
 " kept to condenser. 0.1349 k. x 23.5 c. = 3.17 c. 
 
 expended, Q 87.61 c. 
 
 " gained by injection water, Q l 5.4304 k. x 13.3 c. = 72.22 c. 
 
 15.39 c. 
 
 " in work done 11.03 c. 
 
 " radiated 1.50 c. 
 
 12.53 c. 
 
 Error, 15 - 39 - 12 ' 53 = 3.1 per cent. 
 90.78 
 
 Total heat is 90.78 c., it represents: 
 
 Dry steam per stroke 90 - 78 =0.1387 k. 
 
 6o4.24 
 
 " hour per total H. P 7.188 k. 
 
 ind. H. P 7.983 k. 
 
 net H. P 9.071 k. 
 
 II. Indicated on pistons 137 H.P. 
 
 Revolutions per minute 51.12 
 
 Net H. P 135 
 
 Mechanical efficiency, per cent 91 
 
 Back pressure work in per cent, total work 8.8 
 
 (2^ Ibs.) 0.169 k. 
 
 Expansion 8 
 
 (Boiler pressure 66 Ibs. above atmosphere). 
 
 Heat brought by dry steam 0.1678 k. x 653.82 c. = 109.71 c. 
 
 " entrained water 0.0075 k. x 156.74 c. = 1.16 c. 
 
 " jacket . .0.0112 k. x 497.08 c. = 5.57 c. 
 
 Total. 116.44 c. 
 
 Heat retained to condenser. . . .0.1752 k. x 27.65 c. = 4.84 c. 
 
 Q ......................................................... 111.60 c. 
 
 Heat gained by injection water ............. 9.534 k. x 17.45 c. = 93.18 c. 
 
 o ....................................................... 18.42 c. 
 
 Heat in work done ........................................ 14.25 c. 
 
 " " external radiation ................. ............... 1.50 c. 
 
 - 15.75 c. 
 
 Error, ~' = 2.3 per cent. 
 
260 STEAM USING; Oil, STEAM ENGINE PRACTICE. 
 
 Heat per stroke, 116.44 represents: 
 
 Dry steam per stroke ................................... 116 : 44 L = 0.1781 k. 
 
 653.82 
 
 " hour per total H. P ................... .......... 7.236 k. 
 
 " " ind. H.P ............................ 7.939k. 
 
 " " " " "net H. P ........................... 8.724 k. 
 
 III. Indicated on piston .............. . ...... ............. 158 H. P. 
 
 Revolutions per minute ........... ............................ 49.54 
 
 Net on brake .................................................. 142 
 
 Mechanical efficiency, per cent ................................ 92 
 
 Back pressure work in per cent, of total work ................. 8.1 
 
 .......................................... (2.6 Ibs.) 0.184k. 
 
 Expansion ..................................................... 6 
 
 (Boiler pressure 68 Ibs. above atmosphere.) 
 Heat brought by dry steam ...... . ......... 0.2015 k. x 654.24 c. = 131.83 c. 
 
 " entrained water .......... 0.0112 k. x 158.18 c. = 1.77 c. 
 
 " to jacket ................... 0.0114 k. x 0.0114 c. = 5.65 c. 
 
 Total ....................................................... 139.25 c. 
 
 Heat retained by condenser .................. 0.2127 k. x 29.57 c. = 6.29 c. 
 
 # ............................................................ 132.96 c. 
 
 Heat gained by injection water Q l ........... 5.7334 k. x 19.47 c. = 111.63 c. 
 
 Q Q l ............................. .......................... 21.33 c. 
 
 Heat in work done ....................................... . 16.99 c. 
 
 " " external radiation ................................ 1.50 c. 
 
 - 18.49 c. 
 
 Error, -JrL = 2.4 per cent. 
 
 Heat per single stroke = 139.25 represents 
 Dry steam per stroke ................................... o = ' 2128 k< 
 
 004.^4 
 
 " hour per total H. P ............................... 7.307 k. 
 
 " " " " ind. H. P .............................. 7.955 k. 
 
 " " " " " net H.P ............................ ...8.646k. 
 
 These three experiments check within 3 per cent., a very satisfactory 
 result but, let us add, one difficult enough to obtain. 
 
 It is essential, as we have seen from a long practical experience in 
 these researches, to attend to the most minute precautions to attain results 
 which check to about 1 per cent., and we consider as purely accidental, 
 circumstances which give closer figures, and believe the attempt to get 
 them an illusion. 
 
 The consumptions of the "Corliss" engine with expansion of 6, 8, and 
 11, show us the limited influence of expansion upon the cost. I have 
 already had occasion to note this fact when the "Woolf " engines were in 
 question. There results from this that within the wide limits where we 
 
THE ALSA TIAN EXPERIMENTS, ETC. 261 
 
 have operated, and for jacketed engines, the intrinsic work of the steam is 
 scarcely influenced by a more or less prolonged expansion, this being 
 defined by the cost of a total H. P. 
 
 The "Woolf" engines gained 4.5 per cent, by change from 7 to 28 ex- 
 pansions. The "Corliss" gains 1.6 per cent, by change from 6 to 11. 
 
 On the other hand, for the "Corliss" as well as the "Woolf," the Differ- 
 ences which exist in cost change sign when we pass from the total work to 
 the indicated work in which the back pressure makes itself felt; these 
 differences are still more increased when we take the cost of a net horse- 
 power, that is to say, the industrial consumption, which shows us that 
 expansion 6 is 4.5 per cent, better than 11 for the "Corliss" engines, and 
 for "Woolf" engines expansion 13 is 5.5 per cent, better than 28. 
 
 That which we have noted concerns only jacketed engines. Does ex- 
 pansion act differently without this adjunct? 
 
 The experiments made upon the vertical engine of M. Hirn will 
 show us. 
 
 Hirn Engine. 
 
 Saturated steam, variable expansions. Check on consumption. 
 
 I. Indicated on pistons 107 H. P. 
 
 Revolutions per minute 30.41 
 
 Net on brake , 95 H. P. 
 
 Mechanical efficiency, per cent 89 
 
 Back pressure work in per cent, total work 11.8 
 
 (3 R)s.) 0.213k. 
 
 Expansion 7 
 
 (Boiler pressure, 56 K>s. above atmosphere.) 
 
 Heat brought by dry steam 0.2604 k. x 652.48 c. = 169.90 c. 
 
 " entrained water. . 0.0030 k. x 152.25 c. = 0.46 c. 
 
 Total 170.36 c. 
 
 retained to condenser. . . .0.2634 k. x 32.25 c. = 8.49 c. 
 
 Q .............. ................... 161.87 c. 
 
 Heat gained by injection water Q l .......... 8.8131 k. x 15.72 c. = 140.38 c. 
 
 Q, Qi ............... -... ....................... 21.49 c. 
 
 Heat in work done ....................................... 18.77 c. 
 
 " " external radiation ..... ...... ..................... 2.50 c. 
 
 - 21.27 c. 
 
 Heat per stroke, 170.36 represents 
 
 Dry steam per stroke ................................... rrb = -' 2611 k - 
 
 652.48 
 
 " hour per total H. P .............................. 7.822k. 
 
 " " " ind. H. P ............................... 8.837 k. 
 
 11 " " net H. P.. . 9.929k. 
 
262 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 II. Indicated on piston 146 H. P. 
 
 Revolutions per minute 30.55 
 
 Net on brake 134 H. P. 
 
 Mechanical efficiency, per cent 92 
 
 Back pressure work in per cent, total work 9.2 
 
 " " (3. libs.) 0.225k. 
 
 Expansion 4 
 
 (Boiler pressure 52 Ibs. above atmosphere.) 
 
 Heat brought by steam 0.3695 k. x 651.70 c. = 248.80 c. 
 
 " entrained.. . .0.0037 k. x 149.61 c. = 0.65 c. 
 
 Total 241.36 c. 
 
 Heat retained to condenser. . . .0.3732 k. x 33.65 c. = 12.56 c. 
 
 Q n 228.79 c. 
 
 Heat gained by injection water, Q l 92917 k x 21.82 c. = 202.75 c. 
 
 Qo Qi 26.04 c. 
 
 " in work done 25.27 c. 
 
 " in external radiation 2.50 c. 
 
 27.77 c. 
 
 Error ^ "^ = 0.7 per cent. 
 Heat per stroke, 241.35 c. represents 
 
 Dry steam per stroke,. . . . ^'^ = 0.3703 k. 
 
 651.70 
 
 " " " hour per total H. P 8.449 k. 
 
 " " " " ind. H. P 9.307 k. 
 
 ' " " " net H. P 10.341k. 
 
 These figures seem to prove that expansion has much greater effect 
 when there is no jacket. Expansion 7 is 7.4 per .cent, better than 4 for the 
 total and 4 per cent, better for the net H. P. Is this kept up with super- 
 heated steam? 
 
 Superheated Steam Him Engine. 
 
 I. Indicated on piston 113 H. P. 
 
 Revolutions per minute 29.98 
 
 Net on brake 102 H. P. 
 
 Mechanical efficiency, per cent 90 
 
 Back pressure work in per cent, of total work 9.7 
 
 (2.6 Ibs.) 0.188 k. 
 
 Expansion 7 
 
 (Boiler pressure 56 Ibs. above atmosphere.) 
 Temperature of steam 196 c. (superheated 44.73 c.) 
 
 Heat brought by dry steam 0.2240 k.x 652.48 c.= 146.15 c. 
 
 " " "superheating 0.2240 k.x 0.5x44.73 c.= 5.01 c. 
 
 Total. . . . 151.16 c. 
 
THE A LSA TIA N EXPERIMENTS, ETC. 263 
 
 Heat retained to condenser .. . .0.2240 k.x 30.42 c.= 6.81 c. 
 
 Q 144.35 c. 
 
 Heat gained by injection water, Q^ 8.7384 k.x 14.05 c. = 122.77 c. 
 
 Qo-Qi - 21.58 c. 
 
 Heat in work done 19.97 c. 
 
 " " external radiation 2.50 c. 
 
 22. T C . 
 
 Error 21 " 58 ~ 22 ' 47 =0.6 per cent. 
 
 lol.lo 
 
 Heat total per stroke, 151.16 c. represents 
 Dry steam per stroke 1 ^' 16 = 0.2317 k. 
 
 Dry steam per hour per total H. P 6.655 k. 
 
 "ind. H.P 7.370k. 
 
 " net H.P 8.188k. 
 
 II. Indicated on piston 154 H. P. 
 
 Revolutions per minute 30.174 
 
 Net on brake 143 H. P. 
 
 Mechanical efficiency, per cent 93 
 
 Back pressure work in per cent, of total work 8.3 
 
 Back pressure (3 Ibs.) 0.215 k. 
 
 Expansion 4 
 
 Temperature of steam 231c. (superheated 80.85c.) 
 
 (Boiler pressure 55 Ibs. above atmosphere.) 
 
 Heat brought by dry steam 0.3065 k. x 652.79 c. = 199.92 c. 
 
 " superheating 0.3065 k. x 0.5 x 80.85 c. = 12.39 c. 
 
 Total , 212.31 c. 
 
 Heat retained to condenser. . . .0.3065 k. x 31.3 c. = 9.52 c. 
 
 # ..................................... .................. 202.72 c. 
 
 Heat gained by injection water, & ........... 9.350 k. x 18.7 c. = 174.84 c. 
 
 QQ, ..................................................... 27.88 c. 
 
 Heat in work done ...................................... 27.08 c. 
 
 " in external radiation ................................. 2.50 c. 
 
 - 29.58 c. 
 
 Error ^"^" = 0.8 per cent. 
 /Heat per stroke, 212.31 c., represents 
 Dry steam per stroke.. ................. = - 32 ^ k. 
 
 DO 4. 4 7 
 
 Dry steam per hour per total H.P ................................ 7.000 k. 
 
 " ind. H.P ................................ 7.633k. 
 
 " net H.P ................................ 8.207k. 
 
 III. -Indicated on piston .................................... 125 H. P. 
 
 Revolutions per minute ....................................... 30.306 
 
 Net on brake . ..114H. P. 
 
264 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 Mechanical efficiency, per cent 91 
 
 Back pressure work in per cent, total work 8.9 
 
 (2.7 Ibs.) 0.190 k. 
 
 Expansion 2 
 
 Temperature of the steam 223 c. (superheated 73 c.) 
 
 (Boiler pressure, 55 Ibs. above atmosphere). 
 
 Heat brought by dry steam 0.2822 k. x 552.25 c. = 184.06 c. 
 
 superheating .0.2822 k. x 0.5 x 73.00 c. = 10.30 c. 
 
 Total 194.36 c. 
 
 Heat retained to condenser. . . .0.2822 k. x 34.26 c. = 9.95 c. 
 
 Q 184.41 c. 
 
 Heat gained by injection water, & 8.5983 k. x 18.76 c. = 161.20 c. 
 
 <2o-#i 23.11 c. 
 
 Heat in work done 21.86 c. 
 
 " in external radiation 2.50 c. 
 
 24.36 c. 
 
 Error 2 A"- 206 ercent 
 
 + 94.36 
 
 Heat total per stroke 194.36 c. represents: 
 Dry steam per stroke _ 194 ' 36 = 0.2979 k. 
 
 Dry steam per hour per total H. P 7.874 k. 
 
 " " ind. H. P 8.655k. 
 
 " " net H. P 9.511k. 
 
 The experiments with superheating, expansion, 4 and 7, differ 4.9 per 
 cent., without superheating 7 per cent, of the cost of a total H. P. Hold- 
 ing account of the superheating temperature 196 c. in the first case, we 
 can say that between 4 and 7 expansions there is 7 per cent, difference 
 with common and superheated steam. Introduction of J probably passes 
 below the limit above which the cost is nearly constant. 
 
 The superheater arranged to work with the normal conditions, returns 
 less when the consumption is reduced, as is the case when expanding 7 
 times. It can produce a temperature of about 220 c. when the draught 
 is not influenced by atmospheric conditions. We had a difference of 4 
 calories, and without, we should have had a gain of 3 per cent. more. 
 
 The experiment which best proves the influence of expansion when it 
 passes certain limits is III., cut-off J, with superheating to 233, giving 125 
 H. P. The cost of a total H. P. is 15.2 per cent, more than I., and this dif- 
 erence would have been greater if I. had the same temperature of super- 
 heating as III., 223, instead of 196. 
 
 But we say this experiment III. was made with a throttling of 1 atmos- 
 phere between beginning and end of admission. 
 
 We have seen that throttling is far from producing the effects and hav- 
 ing the influence generally attributed to it. It only brought the "Woolf " 
 engines, when the normal load was diminished one-half, an increase of 5 
 
THE ALSATIAN EXPERIMENTS, ETC. 265 
 
 per cent, in the cost of a total H. P. Finally, an experiment of Hirn's, 
 verified, gave, with the same introduction, one-half, and a load reduced 
 to 100 H. P. gave the same cost as with 125, that is, with 20 per cent, less 
 power. 
 
 Stating, then, a loss by throttle of only 5 per cent., there remains the 
 fact that, with two expansions, there is an increase of 10 per cent, on the 
 cost of a total H. P. 
 
 We have rapidly examined the principal facts which stand in relief on 
 a first approach to the figures of consumption. A deeper study of the 
 phenomena demands a verified analysis, which we will give. 
 
 CORLISS ENGINE EXPERIMENT I., 105 H. P.; EXPANSION 11. 
 
 Account of heat, etc., per single stroke: 
 
 Weight of fluid at cut-off 0.1370 k. 
 
 " " dry steam at cut-off 0.0848 k. 
 
 " " water " " 38.3 per cent 0.0522k. 
 
 " " " entrained.. . 0.0043k. 
 
 " " 
 
 " condensing admission .......................... 0.0479 k. 
 
 Heat given to iron ........................ 0.0479 k. + 499.77 c. = 23.93 c. 
 
 Weight dry steam at end of stroke ............................... 0.1072 k. 
 
 " water " 21.7 per cent .................. 0.0298k. 
 
 Internal heat at cut-off, U ...................................... 59.55 c. 
 
 " " end of stroke, U l .............................. 65.87 c. 
 
 U Z7i ................................................. 6.32 c. 
 
 Heat furnished by jacket ................... . .................... 22.28 c. 
 
 " iron above.. 23.93 c. 
 
 Heat furnished during expansion 4.66 c. 
 
 " absorbed by total work of expansion 9.57 c. 
 
 " lost by external radiation 1.5 c. 
 
 Re .- 22.28 11.07 = 11.21 c. 
 
 Re, per cent, of total heat ^^ = 12.03 c. 
 yu.To 
 
 Check on Re. 
 
 Internal heat at end of stroke 65.87 c. 
 
 Back pressure work 1.21 c. 
 
 Heat retained in condenser 3.17 c. 
 
 " - " cushion.. 1.12 c. 
 
 62.79"c. 
 Heat gained by injection water 72.22 c. 
 
 Re 9.43 c. 
 
 Error. 1L21 9o ^ 8 9 - 43 = 1.9 per cent. 
 
266 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 CORLISS ENGINE EXPERIMENT II., 137 H. P.; EXPANSION, 8. 
 
 Heat account, etc., per single stroke: 
 
 Weight of fluid present 0.1775 k. 
 
 " dry steam at cut-off. . ...... 0.1211 k. 
 
 " " water at cut-off (31.7 per cent.) 0.0564 k. 
 
 " " entrained 0.0074 k. 
 
 " " " condensed during admission 0.0490k. 
 
 Heat given to iron 0.0490 k. x 499.93 c. = 24.49 c. 
 
 Weight of dry steam at end of stroke 0.1434 k. 
 
 " " water at end of stroke (19.2 per cent.) 0.0341 k. 
 
 Internal heat at cut-off U 82.25 c. 
 
 " " end of stroke Z7 X : 88.17 c. 
 
 U ~ U, 5.92c. 
 
 Heat furnished by jacket 5.57 c. 
 
 " " " iron (above). . 24.49 c. 
 
 during expansion 24.14 c. 
 
 " absorbed by total work during expansion 11.50 c. 
 
 " lost by external radiation 1.50 c. 
 
 Re 24.14 13.00 = 11.14 c. 
 
 1114 
 Rc in per cent, of total heat furnished = 9.8. 
 
 Check on Rc: 
 
 Internal heat at end of stroke 88.17 c. 
 
 Back pressure work 1.38 c. 
 
 Heat retained in condenser 4.84 c. 
 
 <k " " cushion., . . 1.23 c. 
 
 83.48 c. 
 " gained by injection water 93.18 c. 
 
 Rc 9.70 c. 
 
 Error, 11 ' 14 - 9 ' 70 = 1.2 per cent. 
 
 -L i. t) 4:4: 
 
 CORLISS ENGINE EXPERIMENT III, 158 H. P.; EXPANSION 6. 
 
 Account of heat, etc., per single stroke: 
 
 Weight of fluid 0.2151 k. 
 
 " " dry steam at cut-off 0.1608 k. 
 
 " water " " (25.3 per cent.) 0.0543 k. 
 
 " " entrained . . 0.0112 k. 
 
 condensed at cut-off. . 0.0431 k, 
 
THE ALSATIAN EXPEEIMENTS, ETC. 267 
 
 Heat given to iron 0.0431 x 498.64 = 21.49 c. 
 
 Weight of dry steam at end of stroke 0.1753 k. 
 
 "water " " (18.5 per cent.) 0.0398k. 
 
 Internal heat at cut-off U 106.30 c. 
 
 end of stroke U^ .108.13 c. 
 
 ETo Z7i 1-83 c. 
 
 Heat furnished by jacket 5.65 c. 
 
 " iron (above) , 21.49 c. 
 
 during expansion 25.31 c. 
 
 absorbed by total work of expansion 12.66 c. 
 
 " lost by external radiation 1.50 c. 
 
 Re 25.31 14.16 = 11.15 c. 
 
 Re. in per cent, of total heat furnished ' a =8. 
 
 139.25 
 
 Check on Re. 
 
 Internal heat at end of stroke ETj 108.13 c. 
 
 Back pressure work 1.51 c. 
 
 Heat retained in condenser 6.29 c. 
 
 " cushion . . .1.21 c. 
 
 102.07 c. 
 
 " gained by injection water 111.63 c. 
 
 Re 9.56 c. 
 
 Error, n - 5 l- 2 f 6 =1.1 per cent. 
 
 The results of the analysis of these three experiments upon the "Cor- 
 liss" engine at different loads follow in order as remarkable as regular. 
 The increase of final internal heat U lt compared with initial internal heat 
 I7 varies with the water present at cut-off and with the expansion. 
 
 There is an important circumstance which we shall utilize later in 
 comparing single and double cylinder engines. In spite of the radical 
 difference in the principles of construction of the "Corliss" and "Woolf" 
 engines, we see that with 13 expansions for the "Woolf" and 6 for the 
 "Corliss" the weight of water at the beginning and end of the expansion 
 and the cooling due to the condenser are very close. 
 "Woolf" engine, initial water, 23.7 per cent.; final 17.9 per cent.; .Rc.8.13 
 
 per cent. Expansion 13. 
 "Corliss" engine, initial water, 25.3 per cent.; final 18.5 per cent., .Re. 8 per 
 
 cent. Expansion 6. 
 
 With expansions 28 and 11 there are greater differences. 
 "Woolf" engine, initial water, 40 per cent.; final 17.6 per cent., Re. 7.8 per 
 
 cent. Expansion 28. 
 "Corliss" engine, initial water, 38.3 per cent.; final 21.7 per cent., Re. 1.23 
 
 percent. Expansion 11. 
 
268 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 HIRN ENGINE EXPERIMENT I., 107 H. P.; EXPANSION 7. 
 
 Account of lieat, etc., per single stroke: 
 
 Weight of fluid 0.2634 k. 
 
 " " dry steam at cut-off. . . 0.1656 k. 
 
 water " 37 per cent 0.0978 k. 
 
 " entrained.. . 0.0030k. 
 
 " condensed at cut-off 0.0948 k. 
 
 Heat given to iron 0.0948 k. x 507.35 c. = 48.10 c. 
 
 Weight of dry steam at end of stroke 0.1707 k. 
 
 " water " 35.2 per cent 0.0927 k. 
 
 Internal heat at cut-off U 114.19 c. 
 
 " " " end of stroke U^ .. . 109.07 c. 
 
 U Z7, 5.12 c. 
 
 Heat furnished from iron (above) . 48.10 c. 
 
 during expansion ............................... 53.22 c. 
 
 " absorbed by total work ........................ ............. 13.70 c. 
 
 " lost by external radiation ................................. 2.50 c. 
 
 Re ............................................... 53.22 16.20 = 37.02 c. 
 
 tyrj f\n 
 
 Re in per cent, of heat furnished 21.6 c. 
 
 170.36 
 
 Check on Re. 
 
 Internal heat at end of stroke C^ ................................ 109.07 c. 
 
 Back pressure work ............ ............................... 2.43 c. 
 
 Heat retained in condenser .................................. 8.49 c. 
 
 Re ........................................................... 37.37 c. 
 
 HIKN ENGINE SATURATED STEAM, 146 H. P., EXPANSION 4. 
 
 Account of heat, etc., per single stroke; 
 Weight of fluid .................................................. 0.3722 k. 
 
 " dry steam at cut-off ................................. 0.2571 k. 
 
 " water (31 per cent.) .................................. 0.1161 k. 
 
 " entrained. . . 0.0037 k. 
 
 " condensed at cut-off 0.1124 k. 
 
 Heat given to iron 0.1124 x 513.76 c. = 57.73 k. 
 
 Weight of dry steam at end of stroke 0.2792 k. 
 
 " water (25.2 per cent) 0.0940 k. 
 
 Internal heat at cut-off, Z7 173.92 c. 
 
 end of stroke, E/i . 175.79 c. 
 
 U 9 U, 1.87 c. 
 
THE ALSA T1AN EXPERIMENTS, ETC. 269 
 
 Heat furnished from iron (above) 57.73 c. 
 
 during expansion 55.86 c. 
 
 " absorbed by total work of expansion 15.83 c. 
 
 " lost by external radiation 2.50 c. 
 
 Re 55.86 18.33 = 37.53 c. 
 
 Re in per cent, of heat furnished, ------- = 15.4 
 
 Check on Re: 
 
 Internal heat at end of stroke, C7, 175.79 c. 
 
 Back pressure work f 2.58 c. 
 
 Heat retained after condensing 12.56 c. 
 
 165.80 c. 
 " gained by injection water 202.75 c. 
 
 Re 36.95 c. 
 
 Error. 37 - 53 ~ 36 - 95 = .1 per cent. 
 
 These two experiments offer a peculiarity which is worthy of atten- 
 tion; we find equal weights of water at the end of the stroke 0.0927 k. and 
 0.0940 k. and the values of Re, 37.02 c. and 37.53 c. The cooling due the 
 condenser is, we know, the heat carried by the water on the surface which 
 it covers, re- evaporating and going to the condenser; the agreement with 
 the weight is remarkable. 
 
 This fact is not found with the "Corliss" engine, for Re is the same, 
 11.21 c. and 11.15 c., while the weights are 0.0298 and 0.0398 k. for expan- 
 sion 11 and 6. We believe then this is due to the jacket, for the same 
 phenomenon reversed is met in the "Woolf" engine; the weight of water 
 present at the end of the stroke of the large piston is 0.0413 k. and 0.0431 
 k. nearly the same for experiments 267 and 185 H. P., while Re is 6.65 c. 
 and 13.48 c. respectively. 
 
 HIRN ENGINE STEAM SUPERHEATED TO 196 C.; EXPERIMENT I. 113 H. P.; 
 
 EXPANSION 7. 
 
 Account of heat, etc., per single stroke. 
 
 Weight of fluid : 0.2240 k. 
 
 " " dry steam at cut-off 0.1688 k. 
 
 " " water " " (24.6 per cent.) 0.0552k. 
 
 Heat given to iron 0.0552 x 506.5 = 27.96 c. 
 
 Weight of dry steam at end of stroke 0.1761 k. 
 
 " " water " (21. 38 per cent) 0.0479k. 
 
 Internal heat at end of cut-off, U 110.22 c. 
 
 end of stroke, Uj, 108.56 c. 
 
 Z7 U^. 1.66 c. 
 
27O STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 Heat furnished by superheating 5.99 c. 
 
 " iron (above) 27.96 c. 
 
 during expansion 35.61 c. 
 
 " absorbed by total work of expansion 14.31 c. 
 
 " lost by external radiation 2.50 c. 
 
 Re 35.61 16.81 = 18.80 c. 
 
 1 ft ftO 
 
 Re in per cent, of total heat furnished - ~ = 12.5 c. 
 
 151.16 
 
 Check on Re: 
 
 Internal heat at end of stroke, C/i 108.56 c. 
 
 Back pressure work 2.14 c. 
 
 Heat retained after condensation. . . 6.81 c. 
 
 103.89 c. 
 Heat gained by injection water 122.77 c. 
 
 Re * 18.88 c. 
 
 Error, 18 ^:^M = 0.05 percent. 
 151.16 
 
 HIEN ENGINE STEAM SUPERHEATED TO 231 C.; EXPEKIMENT II., 154 H. P.; 
 
 EXPANSION 4. 
 
 Account of heat, etc., per single stroke: 
 
 Weight of fluid 0.3065 k. 
 
 " " dry steam at cut-off 0.2866 k. 
 
 " water " " (6.5 per cent.) 0.0199 k. 
 
 Heat given to iron ' 0.099 X 504.42 10.04 c 
 
 Weight of dry steam at end of stroke 0.2698 k. 
 
 " " water " " " (12 per cent.) 0.0367k. 
 
 Internal heat at cut-off, U 176.81 c. 
 
 " " at end of stroke, U l 164.44 c. 
 
 U U, 12.37C. 
 
 Heat furnished by superheating 13.22 c. 
 
 " iron (above) .. 10.04 c. 
 
 during expansion 35.63 c. 
 
 " absorbed by total work of expansion 16.52 c. 
 
 " lost by external radiation 2.50 c. 
 
 Re 35.63 19.02 = 16.61 c. 
 
 Re in per cent of heat furnished ^~ = 7.8 c. 
 
 212.31 
 
 Check on Re: 
 
 Internal heat at end of stroke, l/i 164.44 c. 
 
 Back pressure work 2.45 c. 
 
THE ALSATIAN EXPERIMENTS, ETC. 271 
 
 Heat retained after condensing ................................ 9.59 c. 
 
 157.30 c. 
 " gained by injection water ................................. 174.84 c. 
 
 Re .......................................................... 17.54 c. 
 
 Error - 7 : 5 ^ 0.4 per cent. 
 
 HIBN ENGINE STEAM SUPERHEATED TO 223 C.; EXPERIMENT III., 125 
 H. P.; EXPANSION 2. 
 
 Account of heat, etc., per single stroke. 
 Weight of fluid ................................................. 0.2822 k. 
 
 Weight of dry steam at cut-off .................................. 0.2866 k. 
 
 " water at cut-off (superheated) ....................... 0.0044 k. 
 
 " dry steam at end of stroke ........................... 0.2449 k. 
 
 " water at end of stroke (13.2 per cent.) ............... 0.0373 k. 
 
 Internal heat at cut-off, U ...................................... 169.93 c. 
 
 end of stroke, U, ................................ 149.42 c. 
 
 U /! ...................................................... 20.52 c. 
 
 Heat furnished by superheating ................................ 11.56 c. 
 
 during expansion ................. .....' ......... 32.68 c. 
 
 " absorbed by total work of expansion ...................... 9.24 c. 
 
 " lost by external radiation ............. . .................. 2.50 c. 
 
 Re ........................................ . ....... 32.08 11.74 = 20.34 c. 
 
 OQ OJ_ 
 
 Re in per cent, of heat furnished - = 10.5 c. 
 
 194. 36 
 
 Check on Re 
 
 Internal heat at end of stroke ............................... ____ 149.42 c. 
 
 Back pressure work ............................................. 2.17 c. 
 
 Heat retained after condensing .................................. 9.49 c. 
 
 141.64 c. 
 
 " gained by injection water ....... ......................... 161.30 c. 
 
 Re ........................................................... ... 19.66 c. 
 
 The three values of Re, 18.80 c., 16.61 c. and 20.34 c., vary little; we can 
 therefore state that these figures do not agree with the weight of water at 
 the end of the stroke. Thus for experiments II. and III. the. final weights 
 of water are 0.0367 k. and 0.0373 k., very near each other, while Re is 16.61 
 c. and 20.34 c. We recognize here the effect of superheating, for under 
 the same conditions saturated steam gave equal weights of water and 
 equal values of Re. It is to be remarked that the highest cooling by the 
 
272 STEAM USING; OH, STEAM ENGINE PRACTICE. 
 
 condenser corresponds to the lowest work and the longest admission ^ 
 superheated. 
 
 With jacket, on the contrary, the largest value of Re is with the long- 
 est expansion, 11, of the "Corliss," or to the least work, 185 H. P., of the 
 "Wool!" engine. The reason is that the jacket gives up heat during 
 exhaust proportionally to the difference of temperature, while with super- 
 heating and saturated steam the heat is stored in the surface. Also we see 
 that the difference 16 to 20 c. is much less than for the "Woolf " engine, 6 
 to 13 c., while for equal values of Re, 11 c., we have in the "Corliss" 0.0298 
 and 0.0398 k., varying 3 to 4, while with superheating the values of Re 
 vary 4 to 5. 
 
 Finally, we see the proportion of steam condensed during admission 
 diminish rapidly, with the expansion passing from 24.6 to 6.5 to in exper- 
 iments I., II. and III., while in this last case we find, if we calculate by 
 volume and weight the steam at cut-off, we get more than that given by the 
 direct measurement. The steam occupying the volume at cut-off possesses 
 an excess of heat, which is betrayed by the greater pressure than would 
 be given by the same weight of saturated steam occupying the same space. 
 
 It is then incorrect to say that superheated steam falls to the condition 
 of saturated steam always, and partially condenses on arriving in the 
 cylinder; it is also wrong to attribute to the initial condensation all the 
 loss of pressure observed between the boilers and cylinders. These losses 
 of pressure are mainly due to restrictions in the pipes and passages; this 
 is proved by this experiment where the steam at cut-off is superheated and 
 where the condensation cannot be the cause of such loss of pressure. 
 
 Influences of various expansions upon single cylinder engines their utility 
 from the point of view of cost. 
 
 Uniting in a single table (on opposite page) the results of our analyses 
 we shall render the discussion easier. 
 
 The general effect of a prolonged expansion appears the same whether 
 it be a question of one cylinder or two; whether jacketed or not. It is 
 to be remarked that the phenomenon is much more regular when the 
 expansion is produced in the same inclosure, the order of transfers of heat 
 is continuous, while the large cylinder of the "Woolf" engines gives birth 
 to singular modifications in the thermic action of the surfaces. This 
 peculiarity we have had occasion to note for the experiment with expan- 
 sion 13, and introduction of one- half the small cylinder. 
 
 The increase of expansion always causes an increase in the proportion 
 of steam condensed during admission, and a decrease in the amount of 
 dry steam per total H. P. per hour; but this double action is more or less 
 energetic as the engine is provided with a jacket or not, and according as 
 saturated or superheated steam is used, which the figures of Table IV will 
 prove to us. 
 
 Notwithstanding the reputation of the "Corliss" and its derivatives 
 we have studied them as a single cylinder jacketed, with four, valves. 
 
TBS ALSATIAN EXPERIMENTS, ETC. 
 
 273 
 
 TABLE IV. 
 
 HIBN WITHOUT JACKET. 
 
 
 COBLISS, WITH JACKET. 
 
 
 Saturated. 
 
 Superheated. 
 
 No. of expansions 
 Force indicated on pis- 
 tons, H. P 
 
 I. 
 
 11 
 105 
 7.188 
 10 
 7.983 
 88 
 9.071 
 3 
 6.5 
 38.3 
 21.7 
 6.32 
 11.21 
 12.03 
 
 II. 
 
 8 
 137 
 7.236 
 8.8 
 7.939 
 91 
 8.724 
 4 
 6 
 31.7 
 19.2 
 5.92 
 11.14 
 9.8 
 
 III. 
 
 6 
 158 
 7.307 
 8.1 
 7.955 
 92 
 8.646 
 5.2 
 5.3 
 25.3 
 18.5 
 -1.83 
 11.15 
 8 
 
 I. 
 
 7 
 107 
 7.822 
 11.8 
 8.837 
 89 
 9.929 
 1.1 
 
 37 
 35.2 
 +5.12 
 37.92 
 21.6 
 
 n. 
 
 4 
 
 146 
 8.449 
 9.2 
 9.307 
 92 
 10.341 
 1.0 
 
 31 
 25.2 
 
 1.87 
 37.53 
 15.4 
 
 I. 
 
 196 
 
 7 
 
 113 
 6.655 
 9.7 
 7.370 
 90 
 8.188 
 
 24.6 
 21.4 
 +1.66 
 18.80 
 12.5 
 
 n. 
 
 231 
 4 
 
 154 
 7.000 
 8.3 
 7.633 
 93 
 8.287 
 
 6.5 
 12 
 + 12.37 
 16.61 
 
 7.8 
 
 in. 
 
 223 
 2 
 
 125 
 7.874 
 8.9' 
 8.655 
 91 
 9.511 
 
 * 
 13.2 
 +20.52 
 20.34 
 10.5 
 
 Dry steam per hour per 
 total H. P., ks 
 
 Back pressure work per 
 cent, of total work. . . 
 Dry steam per hour per 
 Ind. H. P.,ks 
 
 Mechanical efficiency, 
 per cent 
 
 Dry steam per hour per 
 net H. P., ks ... 
 
 Per cent, water carried 
 over 
 
 Per cent, water con- 
 densed in jacket 
 
 Per cent, water contain- 
 ed at cut-off 
 
 Percent, water contain- 
 ed at end of stroke. . . 
 Uo Ui during expan- 
 sion, cs . . 
 
 Re loss by cooling due 
 condenser, cs . 
 
 Re in per cent, of total 
 heat furnished 
 
 
 Without doubt the arrangement with circular valves and the releasing gear 
 are very ingenious and offer a certain interest, but practically we much 
 prefer fiat valves with right line movements moved by cams or eccentrics; 
 they always close tight in consequence of the very nature of their work- 
 ing. We have seen "Woolf" engines with the valve seat in perfect order 
 after twenty years' working. The Hirn engine with four slide valves, 
 with seats smooth and close to the cylinder, presents no defects after 
 twenty-five years. This reservation made, we will pass to the examina- 
 tion of the three cases of the single -cylinder jacketed engine with four 
 valves. 
 
 The expansion changing from 6 to 14, the proportion of initially con- 
 densed water increases from 25 to 38 per cent., a change of 13 per cent. 
 At the end of the stroke the difference is less marked, from 18 per cent, to 
 21 per cent. This proves a greater evaporation for the greater expansion. 
 The heat required is drawn from the jacket and surface which has re- 
 ceived it during the admission. But I do not insist upon this series of 
 phenomena discovered by M. Hirn and described by him, for I have 
 treated them at sufficient length in an elementary form in my paper of 
 
 *Superheated. 
 
274 STEAM USING; OB, STEAM ENGINE PRACTICE. 
 
 25th October, 1876 (Bulletins for March, April, May, 1877), concerning the 
 experiments executed under his direction. 
 
 That which we should remark, however, is that this progress crosses 
 the difference Z7 C/i between the internal initial and final heats, the lat- 
 ter becoming greater as the expansion is increased. The " Woolf " engines 
 with variable expansion and jacket have already presented the same phe- 
 nomenon. Is it then one of the effects of the expansion? or is it 
 an effect of the jacket? The study of the Hirn engine permits us to 
 decide. 
 
 We have seen that the proportion of water at the end of the stroke 
 for the "Corliss" differs 3 per cent., while the weight of water is 0.0298 k. 
 for 11 expansions, 0.0341 k. for 9, and 0.0398 for 6; the value Re is also a 
 function of this weight of water, since the loss Re is the amount of heat 
 required to evaporate a portion of water on the surface and return it to 
 the condenser in the shape of steam. How is it then that the three values 
 of .Re are equal to 11 calories, when the weights are as 1, 0.85, 0.75? The 
 same phenomenon presents itself with the "Woolf" engines, but in a man- 
 ner reversed: in that we have different coolings with the same weights at 
 the end of the stroke. Thus for B and F, expansion 6 and 25, the weights 
 are 0.0706 and 0.0685 k. of contained water, while the values of Re are 
 8.44 c. and 21.33 c. Is this still the effect of expansion? But then in the 
 same expansion equal weights of water should give equal values of Re., 
 and we should not have with the Munster engine, expanding 7 times, 
 values of Re, 6.65 c. and 13.48 c., when the weight of terminal water is 
 0.0413 k. and 0.0430 k. for 267 and 185 H. P. Still, a question to which the 
 experiments on the Hirn engine will give a solution. 
 
 The intrinsic values of the different expansions are fixed as we have 
 said by the cost of a total H. P. The economy realized by the work of 
 steam is 2.15 per cent, between expansion 6 and 11 for the "Corliss" en- 
 gine ; but this benefit does not exist industrially, the most economic 
 being expansions 4.6 per cent, better than 11. We can state here again, 
 for single -cylinder engines, the fact that our analyses established for 
 "Woolf" engines jacketed that prolonged expansions are far from being 
 economical in practice. 
 
 The Hirn engine, without jacket, consuming saturated steam with ex- 
 pansions of 4 and 7, offers initial condensations of 31 and 37 per cent. ; with 
 the latter the terminal value is nearly the same, 35 per cent., while the 
 former has evaporated 6 per cent., which is the reverse of jacketed en- 
 gines. In these, a large expansion gives a larger evaporation ; the super- 
 heating also gives a large evaporation with a less condensation for the ex- 
 pansions which augment it. With expansions of 2, 4 and 7, the steam re- 
 mains superheated for the first, and for the two others condenses at the 
 end of admission, 6 and 24 per cent, at the end of the stroke. The propor- 
 tions of contained water are 13, 12 and 21 per cent. For the first case the 
 superheated steam becomes saturated and then condenses 13 per cent, 
 during the expansion, for the second the proportion of water increases 6, 
 and for the last only 3 per cent, to the end of the stroke. 
 
THE AL8A TIAN EXPERIMENTS, ETC. 275 
 
 With regard to the surface condensation and evaporation, the super- 
 heating acts partly as a jacket. Its influence upon the internal heat Z7 T 
 and cooling due the condenser Re should present the same analogies, for 
 these quantities of heat are in intimate relation with the weights of water 
 present at the commencement and end of the expansion ; but the study of 
 U and Re will show us that the action of superheating is less energetic 
 upon them than a jacket, but that it conducts to an economy at least 
 equal if not greater upon the consumption of steam. 
 
 If we work the Hirn engine with saturated steam, the internal heat U^ 
 grows less with regard to U when the expansion is increased. The dif- 
 ference Z7i U = 1.87 c. with 4, and becomes 5.12 c. for 7 expansions. 
 The jacketed engines gave a reversed difference; the final internal heat is 
 there increased with regard to the initial, when the expansion was pro- 
 longed. The jacket is then the only cause of this accession of internal 
 final heat following the order of expansion. This is also the manner of 
 action with superheating, for we see the difference U l U pass from 20.52 
 c. to 12.37 c., and to 1.66 c. when the expansions are 2, 4 and 7. 
 
 The remarkable action of the jacket upon different expansions, and 
 the analogous but less strongly marked effect of superheating, are more 
 especially characterized by the successive values which the cooling due 
 the condenser Re acquires. The Hirn engine without superheating, with 
 expansions 4 and 7, presents the same weight of water, 0.0940 k., and 0.0927 
 k. at the end of the stroke, the values of Re are 37.53 c. and 37.02 c. Note 
 well that it is not necessarily thus, that the thermic conditions of the sur- 
 faces could and should differ with expansions 4 and 7. This remarkable 
 fact proves to us that without a jacket and with saturated steam the weight 
 of final water decides the value of Re. If, however, we give to the steam 
 an excess of heat and superheat it, we see that with equal weights of water 
 at the end of the stroke, 0.0373 k. and 0.0367 k. with expansions 2 and 4, 
 we have values of jRc, 20.34 c. and 16.61 c., differing 3.7 c., when with 
 saturated steam they were equal; but it is with jacketed engines that the 
 difference is greatest. Experiments B and F on the Malmerspach engine, 
 expansions 6 and 25, have, as we have seen, values of Re, 8.44c. and 21.33 c. 
 for weights 0.0706 k . and 0.0682 k. The Munster engine offers Re, 6.65 c. 
 and 13.48 c. for 267 and 185 H. P., the weight of water being 0.0413 k. and 
 0.0431 k. Finally, the "Corliss," with .Re = 11 c. for expansions 6 and 11, 
 has terminal water 0.0398 k. and 0.0298 k. 
 
 There is a particular mode of working, which, singular and energetic 
 as it seems, should find its natural explanation when carefully studied. 
 The cooling by the condenser Re is, as we have many times said, the 
 heat drawn from the surface by the evaporation of a part of the water 
 which covers it, and the passage to the condenser of the water as steam. 
 This heat furnished by the iron surface is given in two ways, according as 
 the engine is jacketed or not. If saturated steam arrives in a cylinder 
 without this improvement, it partially condenses during admission, giving 
 up thereby a certain quantity of heat to the metal, which restores it during 
 expansion and exhaust. We have stated that in this case and in the limits in 
 
276 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 which we have operated, the values of Re are equal for equal weights of ter- 
 minal water. Substituting superheated steam for saturated steam, the heat 
 furnished the iron surface during admission will be differently restored 
 during the exhaust. Equal weights of terminal water give different values 
 of Re, a loss of 3.7 c. more corresponding to the greater expansion. But it 
 is with the jacket that the loss Re increases with the expansion for equal 
 weights of terminal water; here the heat is notonly stored in the metal sur- 
 face during admission, it is furnished from without across the thickness of 
 metal by steam from the boiler condensing on the cylinder. This conden- 
 sation is greatest when the difference of temperature is greatest between the 
 steam in the cylinder and the jacket, that is to say, during exhaust. This 
 transfer of heat accelerates the evaporation of the water which lines the 
 interior surface, augmenting the value of the loss Re. Such is the cause 
 which in most cases renders the jacket less effective than superheating 
 moderately; that its action appears at first sight more energetic we should 
 remember that it makes the difference of heat that it gives the expansion 
 and the waste heat that it gives the condenser a larger quantity than is 
 given by superheating. 
 
 The influence of a greater or less expansion appears to be more with 
 un jacketed engines, at least this is the case with the Him engine between 
 4 and 7 expansions. With saturated steam the costs of a total H. P. are 
 8.449 and 7.822 k. which differ 7.4 per cent. The same costs with super- 
 heated steam are 7.000 and 6.665k.; at the same expansions they differ 
 4.9 per cent., being less than for saturated steam. But we should hold 
 account also that the steam has not been obtained at the same tempera - 
 ture in these two experiments. If the superheating had been the same, 
 the difference would have been 2 per cent, more, or 7 per cent., whether 
 the engine works with or without superheating. Is it then only an effect 
 of the absence of a jacket, or should we consider it as caused by an intro- 
 duction already too great in the cylinder? We are brought to believe 
 that these causes both act, for with superheating and introduction of one- 
 half there is a difference of 15.2 per cent, more than in the experiment 
 with 7 expansions, a difference which would have been still greater if the 
 temperature of the steam in this last experiment had been 223 C. instead 
 of 196 C. 
 
 We see from the cost of a total H. P. which fixes the value of the in- 
 trinsic work of steam, and to which we could give the name of generic 
 cost, offers a marked advantage for the prolonged expansion 7, as well for 
 saturated and superheated steam. This advantage diminishes in practice, 
 the back pressure and friction not varying with the useful work. That is 
 with superheated steam the cost of a net H. P. is the same 8.188 k. and 
 8.207 k., when the generic costs differ 4.9 per cent, for expansions 7 and 4 
 with saturated steam, and the same expansions the generic economy 7.4 
 per cent, becomes a practical economy of 4 per cent. We found with the 
 "Corliss" that the generic economy of 2.15 became a practical loss of 4.6 
 when the expansion changed from 6 to 11. 
 
 These results bring us to the conclusion we have already stated 
 
THE ALSA TIAN EXPERIMENTS, ETC. 277 
 
 for "Woolf" engines jacketed, they confirm experimentally the theoretic 
 advantage of large expansions established by M. Zeuner; at the same 
 time the practical advantage of moderate expansions, preconceived by M. 
 dim. 
 
 COMPARISON OF EXPERIMENTAL VALUES FOR DOUBLE AND SINGLE CYLINDER 
 ENGINES CONCLUSIONS. 
 
 Before stating the figures which serve for this study let us recall some 
 principles which it is indispensable to keep well in mind during the 
 course of the discussion. 
 
 We have determined for each experiment the number of calories 
 brought to the cylinder per stroke, as well as the consumption of dry 
 saturated steam that it represented. This calculation had for its object 
 to bring all the engines to the same unit of comparison. Thus all the en- 
 gines which we have examined, whether jacketed or not, whether using 
 superheated steam or not, consume a certain number of calories per stroke, 
 which can be represented by a weight of dry steam produced by the 
 boiler at the pressure of the experiment. By operating in this manner 
 we preserve a rigorously exact unit, while approaching the old valuation 
 which established the consumption of steam taken by the cylinder with 
 more or less water and even superheated. 
 
 It is not necessary to say that this weight of dry steam is not that 
 which we study in the cylinder; in all the analyses we naturally hold ac- 
 count of the weight really existing in the engine. 
 
 This consumption of dry steam is then presented in three forms, em- 
 bracing each a particular order of facts. The cost per hour of a total H. 
 P., of an indicated H. P., and of a net H. P. 
 
 The first supposes a perfect vacuum, and we get rid of the construction 
 of the engine, and sometimes of the system adopted; in a word, we have 
 the intrinsic value of the work of the steam. In comparing them, we are 
 in a manner studying the use of the heat of the heat brought to the 
 engine. 
 
 The second holds account of the back pressure work, which varies 
 from one experiment to the other upon the same engine with different 
 loads. It is the cost referred to the measure of the area of the indicator 
 diagrams. 
 
 Finally the cost of a net H. P. takes into account the friction of the 
 engine and is the industrial value, more or less economic, of the engine ; 
 it corresponds to the force measured by the brake. 
 
 To each of these three kinds of consumption corresponds a method 
 of working characteristic of engines with one or two cylinders. This 
 mode of working we will define, and join therewith the different results 
 which indicate the kind and value of the transfers of heat. 
 
 In the series of our eighteen experiments we find two experi- 
 ments of which the figures are closely alike ; they can only be distinguished 
 
278 STEAM USING; OK, STEAM ENGINE PRACTICE. 
 
 by the cost, and are eminently suited for an exact comparison of single 
 and double cylinder engines. 
 
 These are the experiments upon the " Woolf," expanding 13, and the 
 "Corliss," expanding 6. The proportions of contained water differ little, 
 25.3 for the beginning and 18.5 for end, for the "Corliss " ; 23.7 and 17.9 per 
 cent, for the " Woolf." 
 
 The internal initial and final heats are nearly constant, 106.30 c.-~ 
 108.13 c. = 1.83 c. = U U x and 283.55 c. 282.31 c. = 1.24 c. The 
 " Woolf " presents a peculiarity which we have already noticed. Dur- 
 ing the expansion in the " Corliss " engine a continuous evaporation of 
 6.8 per cent, took place. In the " Woolf " engine the phenomenon is not 
 so simple. We have in the total expansion an evaporation of 5.8 per cent.; 
 but, because of the arrangement with two cylinders and the introduction 
 to half stroke in the small cylinder, the expansion is broken into two por- 
 tions, in each of which the transfers of heat in a manner differ. 
 
 The expansion in the small cylinder during the last half of the stroke 
 gives rise to an evaporation of 10.3 per cent, of the water inclosed with 
 the steam, the internal heat is augmented TJ U l = 283.55 c. 305.84 c. = 
 22.29 c., much more rapidly than in the corresponding experiment with 
 the "Corliss," and is with 13.1 water. When the mixture passes to the large 
 cylinder at the end of the stroke, the weight of the water is 17.9 per cent., 
 there has been 4.8 per cent, of the fluid condensed in the second part of 
 the expansion. 
 
 The modification of the thermic phenomena by an expansion com- 
 menced at half the small cylinder is radical, as we see ; it seems to prove 
 the inaction of the jacket. We ha TT e seen that the jacket counteracts in 
 part the energetic condensations which are produced at the commence- 
 ment of the stroke of the large piston. 
 
 The fact which we note, joined to the values of the cooling, by the 
 condenser, for the two experiments should reduce to their just value the 
 edifying considerations, outside of experimental grounds, which show 
 the superiority of " Woolf " engines. 
 
 The too widely diffused opinion still is that for stationary as for mar- 
 ine engines the double- cylinder jacketed engine is actually that which 
 gives the best economic return in the production of motive power. 
 
 " In these engines the steam acts first with or without expansion in the 
 small cylinder, then with expansion in the large cylinder. This latter is 
 then alone in communication with the condenser. By this disposition a 
 portion of the force produced escapes the cooling action of the condenser 
 and the cylinder condensation. 
 
 " Further, because of the jackets in which the two cylinders are placed 
 in the midst of steam from the boiler direct, the expanding steam in the 
 large cylinder is at a much lower temperature than that of the jacket, and 
 consequently is easily warmed and the cylinder condensation notably 
 lessened." 
 
 These deductions, which appear so logical that I myself believed them 
 rational enough (1874-75), receive the formal condemnation of experience, 
 
THE ALSA TIAN EXPERIMENTS, ETC. 279 
 
 This proves once more that it is impossible, as M. Hirn has said, to estab- 
 lish a priori anything that shall be tolerably exact concerning steam en- 
 gines. "A correct theory can only be established a posteriori, that is to say, 
 after the experimental study of each special system of engines." We shall 
 see at the same time that it is important to collect together the discover- 
 ies that M. Hirn calls by the name of the "practical theory of the steam 
 engine." This theory will enable us to establish on a certain basis and 
 with the most minute details the comparative experimental values of 
 the different systems of steam engines. 
 
 It is generally admitted, we will say, that the arrangement with two 
 cylinders ought to withdraw a portion of the force produced from the 
 cooling action of the condenser; how is it then that we have Re 8 and 8.3 
 per cent, for the " Woolf " and the " Corliss " ? This same arrangement, 
 where the expansion is made in the large .cylinder with a steam jacket 
 which warms it easily, should diminish the condensation. Whence comes 
 then the condensation of 5 per cent., which we have found, when, on the 
 contrary, the "Corliss" and the small cylinder present evaporations of 7 
 and 10 per cent, during the expansion. Two hypotheses which appeared 
 to have a certain basis are annulled by the analysis; they should put us 
 on our own guard against all researches which are not purely experimen- 
 tal and verified. 
 
 The two experiments which we compare have the same proportion of 
 initial and final water, 25 and 18 per cent., the same loss by the condenser, 
 8 per cent., for both the internal, initial and final heat are nearly con- 
 stant. 
 
 The consumption of dry steam should then fix for us the relative 
 value of single and double cylinder engines with jackets, but they should 
 not be admitted without preliminary discussion. 
 
 The consumption of dry steam per total H. P. gives, as we have said, 
 the intrinsic value of the work of the steam; it is 6.878 k. for the " Woolf," 
 expanding 13, and 7.307 k. for the "Corliss," expanding 6, being 5.9 per 
 cent, in favor of the " Woolf." This figure fixes at 6 per cent., the intrin- 
 sic economy realized by expansion in a separate cylinder; if it was not 
 submitted to two important corrections in passing from the domain of 
 theory to the realm of industrial practice, it would seem to give reason to 
 the partisans of this system. 
 
 But before making these two corrections, we should notice that these 
 two experiments have been made, with all their similar features, with 
 different expansions. An expansion of 13 gave with the "Corliss" an econ- 
 omy of 1.5 per cent., and we can affirm without committing sensible error, 
 that an expansion of 13 would give 2 per cent, improvement, which re- 
 duces to 4= per cent, the generic economy of expansion in the large cylin- 
 der. It is not large,' as we see, and this disappears completely, giving 
 place to a considerable increase in practice. 
 
 In a note presented to our society (session Aug. 26, 1874) and appear- 
 ing in the Bulletin in 1875 I studied the variations of back pressures in 
 steam cylinders, with the values of negative work in terms of total work. 
 
28O STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 This study proved that the negative work absorbed 15 per cent, of the 
 total work in "Woolf " engines, and 7 per cent, only in "Corliss" engines, 
 the back pressure work being the same 0.210 to 0.217 in the two cases (2.9 
 and 3.0 tbs.). These results were established where the proportions of 
 total work absorbed by the back pressure remained constant. The two 
 experiments which occupy us have not the same vacuum, that of the "Cor- 
 liss" is a little better than the "Woolf," 0.184 to 0.226 k., making 8 per cent, 
 instead of 7 per cent, for we are within the stated limits; with a back pres- 
 sure of 0.226 k., the negative work of the "Corliss" would have been 9 per 
 cent. It is thus proved that with the same vacuum of 0.226 k. by the fact 
 of adding a second cylinder to prolong the expansion we create a serious 
 loss; this loss destroys the genuine economy that we have established for 
 double-cylinder engines from the consumption per total H. P. 
 
 We shall not be astonished then that the "Corliss" leads 2.3 per cent.. 
 Avhen we compare the cost per indicated H. P., over the "Woolf." 
 
 Finally, the last brake experiments by our Mechanical Committee have 
 proved that there is a notable difference in the power absorbed in friction 
 of the moving parts in two horizontal engines, of which one had a single 
 cylinder while the other had two cylinders, "Woolf" system. This infe- 
 riority adds its influence to the negative work and brings the two follow- 
 ing consumptions pernet H. P.; "Corliss," 6 expansions, 8.646 k.; "Woolf," 
 13 expansions, 9.465 k. Such are the weights of dry steam valued on 
 a Prony brake; they represent the quantities of heat expended in the 
 
 industrial cost of the motor, their difference is ^-~ 5 ^* 8.7 per cent. 
 
 9.465 
 
 in favor of the engine with one cylinder, coming to the point of the prin- 
 ciple which I stated at the session of January 30, 1878, concerning jacketed 
 engines, viz.: 
 
 That it is always possible to construct a vertical beam engine with one 
 cylinder and four valves, consuming saturated steam that shall be at least as 
 economical as the "Woolf" beam engine. 
 
 We have compared the two experiments which are alike in most of the 
 transfers of heat, etc. There are also others which should conduct us to 
 more rigorous conclusions. Let us see, meanwhile, what will result from 
 the whole of our verified experiments, to which we will add the results 
 upon the compound of the French Navy, which I published in my last 
 memoir. 
 
 Our readers may perhaps find that we fall into fastidious repetitions; 
 but the figures inscribed in Table V. shows us once more that it is impos- 
 sible to draw conclusions if one is not in possession of a sufficient series 
 of experimental studies, not only of one system or other, but even from 
 one experiment to another upon the same engine. What is the use of 
 seeking the law of expansion, of substituting a curve of the hyperbolic 
 family, of comparing adiabatic curves and others with the indicator dia- 
 gram, when it is impossible to deduce from one verified experiment the 
 results which we shall obtain when we impose other conditions upon the 
 engine? The practical theory, the verified analysis of numerous expert 
 
THE ALSATIAN EXPERIMENTS, ETC. 
 
 281 
 
 TABLE V. 
 
 
 
 
 Si, 
 
 g- 
 
 
 
 * 
 
 *' 
 
 
 
 Is" 
 
 ftS * 
 
 og- 
 
 
 Pi 
 
 M 
 
 pansion. 
 
 ii 
 
 1 
 
 ft 
 
 o 
 
 8l ||a' 
 SlliJfJ 
 
 3 ft o to bOi 
 "S -^ 2 . 
 
 li 
 
 2 s 
 II 
 
 |M 
 
 >M 
 
 ference 
 iternal h 
 /o - Uj. 
 ft lories. 
 
 >, 
 
 S ft 
 
 III 
 
 
 "2 
 
 H 
 
 ^*f-H -2 
 
 
 <v 
 
 &w 
 
 *-* ^SJ Q 
 
 c o"^ 
 
 
 d 
 
 M 
 
 13 
 
 
 
 n ~ 
 
 ft 
 
 
 ft 
 
 ft 
 
 
 DOUBLE CYLINDERS. 
 
 
 
 
 
 
 
 
 j 
 
 
 I. "Woolt" vertical jack- ( 
 
 143 
 
 28 
 
 6.731 
 
 0.181 
 
 18.6 
 
 8.273 
 
 83 
 
 10.019 25.88 
 
 7.8 
 
 eted 1 
 
 215 
 
 13 
 
 6.878 
 
 0.226 
 
 15.6 
 
 8.149 
 
 86 
 
 9.465 +1.24 
 
 8.3 
 
 II. " Woolt" vertical jack- J 
 
 185 
 
 7 
 
 7.384 
 
 0.234 
 
 24.1 
 
 9.730 
 
 78 
 
 12.411 31.89 
 
 3.5 
 
 eted .. 1 
 
 347 
 
 7 
 
 7.112 
 
 0.293 
 
 17.4 
 
 8.614 
 
 87 
 
 9.864 17.40 3.4 
 
 III. "Wooir horizontal! 
 
 130 6 
 
 7.290 0.253 
 
 20.1 
 
 9.120 
 
 86 
 
 10.563 3.63 
 
 1.2 
 
 jacketed .... 1 
 
 180 6 
 
 7.328 0.295 
 
 17.4 
 
 8.878 89 
 
 9.975 6.12 
 
 0.7 
 
 Compound vertical jacket 
 
 
 
 
 
 
 
 
 
 
 ed 
 
 690 
 
 5 
 
 7.510 
 
 0.216 13.4. 
 
 8.671 
 
 89 
 
 9.975 +40.32 
 
 5.7 
 
 SINGLE CYLINDERS. 
 
 
 
 
 
 
 
 
 
 
 "Corliss" horizontal jack- j 
 
 105 
 
 11 
 
 7.188 
 
 0.148 
 
 10.0 
 
 7.983 88 
 
 9.071 6.32! 12.0 
 
 eted 1 
 
 158 
 
 6 7.307 
 
 0.184 
 
 8.1 
 
 7.955 92 
 
 8.640 1.831 8.0 
 
 Him vertical unjacketed, 
 
 
 
 
 
 
 
 superheated 196 
 
 113 T 
 
 6.655 
 
 0.188 
 
 9.7 
 
 7.370 90 8.188 +1.66 12.5 
 
 Hirn vertical unjacketed. 
 
 
 
 
 
 
 
 superheated 231 
 
 154 4 
 
 7.000 
 
 0.215 
 
 8.3 
 
 7.633 
 
 93 8.207 +12.37 7.8 
 
 Hirn vertical unjacketed. 
 
 
 
 
 
 j | 
 
 superheated 223 1251 2 
 
 7.8741 0.190 
 
 8.9 
 
 8.655 
 
 91 
 
 9.511 +20 52! 10.5 
 
 Hirn vertical unjacketed, 
 
 
 
 
 
 
 . . _ 
 
 superheated 220 
 
 99 2 
 
 7.763 0.178 
 
 10.4 
 
 8.063 
 
 88 9.844 +16.82 14.1 
 
 Hirn vertical unjacketed, 
 
 
 
 
 
 
 
 
 saturated steam 107 
 
 7 
 
 7.822 
 
 0.213 
 
 11.8 
 
 8.837 
 
 89 
 
 9.929 
 
 +5.12 
 
 21.6 
 
 Hirn vertical unjacketed, 
 
 
 
 
 
 
 
 
 
 
 
 saturated steam 
 
 146 
 
 4 
 
 8.449 0.225 
 
 9.7 
 
 9.307 
 
 92 
 
 10.341 
 
 1.87 
 
 15.4 
 
 ments, can alone give us the solution of problems to which they still 
 apply formulae which are only experimental in name. This law which M. 
 Hirn has stated, and which my preceding memoirs have confirmed, makes 
 us consider the formulae called experimental of expansion and work as 
 algebraic distractions, perhaps interesting, but having, from a practical 
 point of view, a utility hardly worth contesting. 
 
 For us the true theory of the steam engine created by M. Hirn com- 
 pletes itself every day that new experiments, entirely checked and ana- 
 lyzed, are made public and added to those which we have published after 
 him. 
 
 Have I not compared the results of pure experience in determining the 
 influence of the vacuum, of throttling, of expansion, of the addition of a 
 separate cylinder? And still in this case before placing the principle let 
 us leave a sufficient margin for the minor irregularities which it is difficult 
 to avoid practically. 
 
 All that we have said concerns only those researches which are en- 
 titled "experimental researches." By the side of the practical theory the 
 generic theory of the employment of steam as an intermediate in the em- 
 ployment of force retains its special value. The fine work of Zeuner, for 
 example, conceived in the purely theoretic manner, remains a model 
 
282 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 of the method to be followed when the hypothesis is that the 
 surfaces transmit no heat; but what it is necessary to shun with the 
 greatest care is the mixture of generic theory with the practical theory 
 of which the starting points are diametrically opposed one to the other.. 
 
 The successive analyses have already given all the remarable peculiar- 
 ities; in each of the experiments we have used them to determine the 
 influence of expansion upon the transformations of the steam. We shall 
 also ask of Table \ a collected view of the costs and the circumstances 
 which influence them. 
 
 A part of the interesting results which we shall note concern the "Cor- 
 liss, " expansion 6, and the horizontal "Woolf" with the same expansion. 
 They present the relative value qf single and double cylinder engines in a 
 different form from that which we have examined when we consider only 
 the parallelism between the transformations of steam. Thus, with the 
 same expansions the cost of a total H. P. corresponds, 7.307 k. for the 
 "Corliss" at 158 H. P., and 7.290 and 7.328 k. for the "Woolf" with 130 and 
 180 H. P. But we have already said that this latter had an insufficient 
 compression in the clearance; if this compression had been properly regu- 
 lated the cost would descend to 7 kilos, of steam per total H. P. We 
 should then find an intrinsic difference of 4 per cent, in favor of expan- 
 sion in a separate cylinder, a difference which confirms the principle 
 that we have stated concerning the generic economy of this kind of 
 engine. 
 
 What becomes of this benefit of 4 per cent, when we pass from the 
 total work to the indicated and the net work; that is to say, when we 
 occupy ourselves with only the industrial work usefully realized. 
 
 The negative work of the "Woolf" engine absorbs 20 per cent, at 130 
 H. P. and 17 per cent, at 180 H. P., while the "Corliss" only loses 8 per 
 cent., which brings the "Corliss" 6 per cent, ahead of the "Woolf "under 
 its best conditions at 180 H. P. If we compare the cost of a net H. P., 
 
 valued by a brake, we find 9 ' 97 ^ ~ 8 ' 64 ~ = 13.5 per cent., which a suffi- 
 
 t/.y /o 
 
 cient compression will reduce 4 per cent., and there remains a practical 
 gain in favor of the "Corliss" of 9.5 per cent., working at the same expan- 
 sions. 
 
 We see that the cost of a total H. P., 7.307 k. and 7.328, are equal for 
 the "Corliss" and horizontal "Woolf" for 6 expansions, 158 and 180 H. P. 
 A well-known construction would give the same vacuum, 0.184 k., to the 
 "Woolf" as the "Corliss;" but the negative work would be 8 per cent, and 
 11 per cent, at least. A difference of 3 per cent, should be added for fric- 
 tion, being 92 and 89 per cent, respectively. 
 
 The total, 6 per cent., could be reduced by 4 per cent, by proper 
 cushion and there remains at the end of our account an inferiority of 2 per 
 cent, for the "Woolf" working at the same expansion as the single -cylinder 
 " Corliss. " In conditions easily realized in practice the consumption of the 
 "Woolf" per net H. P. would be 8.8 k. This is the lowest we should expect 
 from this kind of engine. In my preceding paper I had fixed this con- 
 
THE ALSATIAN EXPERIMENTS, ETC. 
 
 28.3 
 
 sumption at 9 k. deduced from the vertical "Woolf " and the "Him" engine 
 with superheated steam. 
 
 With different expansions, 13 and 6, but with equal transformations of 
 steam, the costs per total H. P. vary 5.9 percent, against the "Corliss" with 
 the same vacuum 0.184 k.: for the two engines the back pressure work 
 would have been 8 per cent, for the "Corliss" and 13 per cent, for the 
 "Woolf," being 5 per cent, the other way; there remains only about 1 per 
 cent, advantage, comparing by the indicated H. P. Industrially this 
 advantage disappears and there is a loss of 5 per cent, against the "Woolf," 
 the friction being such that we obtain 86 per cent, being 6 per cent, worse 
 than the "Corliss" at 92 per cent, mechanical efficiency. 
 
 We have then the right to conclude that the single-cylinder engine is 
 at least the equal of the double-cylinder, although to this day with different 
 consumption more favor is shown the double-cylinder engines. 
 
 A number of other interesting results are presented by our table. I will 
 point out those given by superheating without going into details on the sub- 
 ject, the question having been many times treated in our Bulletins, then the 
 effect of throttling reducing from 347 to 185 H. P. For the other experi- 
 ments the small difference in generic cost is the other way and may be 
 neglected. 
 
 We can push all these comparisons very far into the most intimate 
 details of the working of the engine, following the elementary 
 method, which consists simply of examining all the results of our verified 
 experiments. We believe that we are rendering a service to our readers in 
 leaving to each the particular study of the experiment which interests him. 
 Our duty should be to point out the path to be followed; to give a sufficient 
 number of verified experiments to permit the engineer to find the points 
 for future researches. My end in writing this paper has been above all to 
 confirm by a sufficient number of verified analyses the principle of the 
 equality of double and single cylinder engines in point of consumption, to 
 determine the degree of influence of a more or less prolonged expansion, 
 to verify again the influence of a reduction of passageway or throttling the 
 steam entering the cylinders. 
 
 
 
 1 
 
 X 
 
 1 
 
 1 
 
 1 
 
 ! 
 
 CONSUMPTION OF DRY STEAM 
 
 d 
 
 o 
 
 Pi 
 
 l^ 
 
 Cfl 
 
 Pi 
 
 
 Pi 
 
 s 5 
 
 PER HOUR IN DIFFERENT 
 
 "3 
 
 W UJ 
 
 | 
 
 ' oj 
 
 C43 
 
 hi 
 
 & Pn 
 
 ENGINES. 
 
 
 
 13 
 
 !i 
 
 e3 
 
 11 
 
 ! 
 
 IB 
 
 
 H 
 
 EH 
 
 
 ,] 
 
 W 
 
 ^ 
 
 u 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Woolf engines jacketed 
 
 13 
 
 6 8 
 
 14 
 
 7 9 
 
 87 
 
 9 8 
 
 1.14 
 
 
 7 
 
 7.1 
 
 13 
 
 8 2 
 
 88 
 
 9.3 
 
 1.16 
 
 U M M 
 
 6 
 
 7.3 
 
 12 
 
 1.3 
 
 90 
 
 9.2 
 
 1.15 
 
 Compound engines, jacketed. . . 
 
 5 
 
 7.5 
 
 12 
 
 8.5 
 
 90 
 
 9.5 
 
 1.19 
 
 
 
 
 
 
 
 
 
 Corliss, jacketed 
 
 6 
 
 7.3 
 
 8 
 
 7.9 
 
 92 
 
 8.6 
 
 1.08 
 
 Him, superheated 
 
 7 
 
 6.6 
 
 9 
 
 7.2 
 
 90 
 
 8.0 
 
 1.00 
 
 
 
 
 
 
 
 
 
284 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 All these actions we have seen are in a restricted circle; we can fix for 
 the builders the lower limit of consumption, a limit which we should seek to 
 attain by good designs. We can give at the same time the corresponding 
 consumption of coal, but we warn the reader that this valuation should 
 never be taken as a unit of comparison, the calorific power of coal being as 
 variable as their localities. We will admit, however, that from the numer- 
 ous experiments made by the Society of Mulhouse upon boilers, a weight 
 of dry steam, eight times the weight of the coal can be produced by 
 medium coal (Kon champ). This is the figure on which we will give the fuel 
 required. 
 
 The consumptions of dry steam per hour for indicated and net horse - 
 power, are deduced from the total horse-power found experimentally 
 They have a real base, a definite point of departure essentially certain in 
 practice. 
 
 The back pressure work varies from 0.200 k. to 0.220 k. for the double- 
 cylinder engines and 0.180 to 0.200 k. for single cylinders, values which it 
 is always possible to secure by proper arrangements of the exhaust. 
 
 Concerning the mechanical efficiency, three of them have been ob- 
 tained directly by brake. The others have been deduced from these same 
 experiments by comparing the indicator diagrams. We have seen that 
 there cannot be an error of 1 per cent. The net power is then exact. It 
 is that which represents the industrial work. 
 
 In this form, which only differs from Table V. by the uniformity of the 
 back pressure work, we perceive clearly the influence of an expansion pro- 
 longed from 7 to 13, a generic gain of 4 per cent., which industrially is 2 
 per cent. The compound engine comes after the " Woolf." 
 
 The horizontal "Woolf" and horizontal "Corliss" have the same generic 
 consumption for the same expansion. This disappears and the "Corliss" is 
 6.5 per cent, better than the "Woolf," industrially. 
 
 Finally the steam per total horse -power for the "Woolf" expanding 13 
 times is within 3 per cent, of the Hirn unjacketed with steam heated to 
 200 C., which becomes 12 per cent, in favor of the superheating, when we 
 compare the net horse-power, while the "Corliss" is only 7 per cent behind 
 the Hirn. 
 
 In presence of these results have we not the right to ask what dis- 
 coveries have been made since the day that Watt created the engine? 
 
 Without speaking of the well-balanced mechanism of the beam engine 
 we can say that this man of genius had produced the heat engine complete 
 in its three parts; the separate condenser, the steam jacket, an expansion 
 prolonged as useful, leaving to those who came after to seek the attain- 
 ment of high pressures, the construction of boilers at his time prohibiting 
 him from following the complete realization of the principles of expansion 
 which he had stated. 
 
 Also is there not in the series of improvements in the construction of 
 engines that we have been studying, a marked discovery in the history of 
 the heat engine, bringing another order of ideas, born nearly at the same 
 time as thermodynamics, of which it is one of the best applications? 
 
THE ALSA TIAN EXPERIMENTS, ETC. 285 
 
 Each of my readers has always understood that I refer to the discovery 
 made by Hirn and stated in our Bulletins for 1855, for it was at that time 
 that he demonstrated the thermic influence of the metallic surface of the 
 inside of the cylinders upon the inclosed steam. This discovery, analyzed, 
 discussed and completed by twenty years of continual research in which 
 he has kindly associated me, has conducted him to a method of analysis 
 that he has called the "Practical Theory of the Steam Engine. " We have 
 not insisted upon the exact character and almost elementary simplicity 
 which is one of the remarkable features of this practical theory. As to 
 its importance, it is, I think, placed beyond doubt by the value of the 
 facts that it has permitted us to define in the course of this paper. 
 
CHAPTER VI. 
 
 STEAM HEATING.* 
 INTRODUCTION. 
 
 The subject of Steam Heating will be presented in this Chapter under 
 the following heads: 
 
 A. The theory of steam heating, the laws of the transmission of heat 
 and the coefficients used in reducing this theory to practice. 
 
 B. A description of the various systems in use in the United States, 
 with a note on the magnitude of the works employed up to 1881. 
 
 C. A detailed description of the apparatus used and the experiments 
 made under the direction of the writer. 
 
 D. The project of heating a cotton mill with steam, and a comparison 
 of the cost when condensing and non- condensing engines are used to drive 
 the mill. 
 
 A. The Theory of Steam Heating. 
 
 The transfer of heat takes place by three processes: Radiation, Con- 
 duction, Convection. 
 
 In heating a room, for example, with an open grate fire, the heat in 
 the room is mostly radiant heat. The heat of the fire is given to the gas 
 and air in the flue and is carried away by convection. With radiation we 
 have little to do, as in all heating apparatus the heat is brought to and 
 delivered from some intervening solid body by fluids and passed through 
 the solid. 
 
 The rate of conduction through a plate is expressed by a well-known 
 formula (Eankine's "Steam Engine, "p. 260.) 
 
 mf m 
 
 q where T and F are the temperatures of the fluids on 
 
 a + a' + px 
 
 either side of the plate, a and a* are the coefficients of conductivity from 
 the fluids to the plate, and x the thickness of the plate, p the conductivity 
 of the plate itself. 
 
 For British units with x in inches p = 0.0043 heat units per square foot 
 per hour for iron plates, a quantity which, with ordinary thicknesses, is so 
 small that it may be neglected for iron plates. 
 
 The terms <rand a' depend entirely upon the nature of the surfaces and 
 upon the rapidity with which the fluids are circulated and the heat brought 
 to and removed from the plate or surfaces in question. What is required 
 
 *A paper read before the Engineers' Club of St. Louis, June, 1882, by the Author. 
 
STEAM HEATING. 287 
 
 for us as engineers is to determine the limits which appear in practice, 
 leaving to the physicist the general investigations of the laws of the sub- 
 ject. 
 
 It has been suggested (v. Rankine's "Steam Engine," p. 260) that for 
 
 _ irrv T}^ 
 
 iron plates <r + </ may be put equal r _^~, and then q = -' -'- and 
 
 a = ^ ) 2 , and he further states that for air and water in a furnace 
 
 q 
 and boiler a is from 160 to 200 for q in British units. 
 
 When the fluids are steam and air, we should expect from the greater 
 mobility of steam than water, a greater rate of conduction, and we are not 
 disappointed. 
 
 From many experiments in the open air, the steam condensed per 
 square foot of pipe surface of | inch thick wrought iron is found to be 1| 
 pounds per square foot per hour when T' = 220 F. and T = 20 F. 
 
 In British heat units, the heat given up per pound of steam is 969 
 units; but as the water temperature is not given and is usually from 180 F. 
 to 200 F., we shall make no large error by taking in this and in other in- 
 vestigations the heat delivered by condensing 1 pound of steam as 1,000 
 heat units; we then have. 
 
 = <JL~_r? = 20 x 20 = 29 nearly. 
 q 137o 
 
 Experiments made by the writer, which will be described hereafter, 
 give values of a as great as 100. As experiments were made under the 
 usual usage, and are more nearly in accordance with practice, the results, 
 moreover, giving for the larger values of a a less rapid transmission of 
 heat, they are obviously the safer to follow in designing heating surface. 
 The great difference is to be attributed entirely to difference in the circu- 
 lation of the air by which the condensation was effected. 
 
 By experiments made in the United States Navy, seepage 63 of "Trea- 
 tise on Boilers," by Engineer-in- Chief Wm. H. Shock, the transfer of heat 
 is stated to be in proportion to the difference in temperature, instead of 
 the square of the difference as taken by the writer. 
 
 The experiments there given appear to give the same quantities as 
 those noted above in the open air, but would be reduced to q = 2.6 (T' T) 
 by the writer's experiments under the practical conditions commonly 
 found, and this seems to agree more closely with the usual practice in the 
 United States, and we shall therefore use this formula in preference to the 
 other and more scientific one. 
 
 The heat required in any given building will depend upon the heat 
 transmitted to the external air around the building and the amount of air 
 carried through the building or the ventilation. The former effect is 
 
 m' rp 
 
 measured by the same formula, q = - , where of course the con- 
 
 G + (7 + pX 
 
 stants have different values, and the latter, by the quantity of air used in 
 ventilation and the amount of heat given to the air. 
 
 The constants a, a f and p vary very much with the materials of the 
 
288 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 walls and roofs, and the kinds of surfaces; but the most important cause 
 of variation is the rapidity with ^vhich the heat is removed from the out- 
 side by the action of the wind, and the variation found here is so great 
 that the minor changes become less important. The effect of ventilation 
 is easily computed from the weight and specific heat of air when the quan- 
 tities are known. 
 
 The values of p and a + a' are given by some writers on heating and 
 ventilation, but our safest course, as engineers, will be to seek the practi- 
 cal limits given by experience, and we find the limits pretty wide. 
 
 According to D. K. Clark ("Manual for Mechanical Engineers," p. 488), 
 M. Peclet found that for T' T = 36 F. each square foot of wall 
 would transmit 26 British units per hour, and that the glass windows 
 passed heat at the rate of 30 units in place of 26. The same author gives, 
 on the authority of Mr. Hood (Idem., p. 481), the rate of 1.4 units per 
 square foot per hour per degree of F. of T f T for glass windows, or for 
 36 q = 50.4 units, and the further statement that q varies with the square 
 root of the velocity of the wind. From experiments by the author on large 
 buildings, the transmission varied from 0.67 to 1.25 units per square foot 
 per hour per degree F. of T' Tfor the whole surface of walls and win- 
 dows. The larger values were produced by wind action, and are the ones 
 that should be taken in practice. The effects of a liberal ventilation are 
 included in the above results. The experiments will be given later. 
 
 Summing up these results we find that the transfer of heat from steam 
 to air may be expressed as q = a (T' T) where T' and Tare the temper- 
 atures of the steam and air on each side of a thin iron surface, and q is the 
 rate of heat units per unit of surface per hour. 
 
 For T' and T in degrees F. and q in British heat units, a = 2.6 for 1 
 square foot per hour. 
 
 For T f and T in degrees centigrade and q in calories per square metre 
 per hour, a = 12.48. 
 
 For the external surface of a building, including walls, windows and 
 roof together, and taking no account of the material for the maximum 
 transfer of heat, q = c (T f T). 
 
 For q in British heat units per square foot per hour c = 1.25 for F 3 . 
 
 For q in calories per square metre per hour, c = 6 for 0. 
 
 For the surface of a steam boiler as ordinarily constructed ^ -^ 2 
 
 a 
 
 For British heat units and F. per square foot per hour, a = 200. 
 For French heat units centigrade degrees and meters, per square meter 
 per hour in calories, a = 23.14. 
 
 1 = 0.0432. 
 a 
 
 For keeping any building permanently warm we must have a steady 
 flow of heat from the furnace to the boiler, from the boiler through the 
 heaters to the air in the building and from the walls to the external air. 
 The same number of heat units per hour must be transferred in each case, 
 whence it becomes easily possible to find the heating surface and boiler 
 
STEAM HEATING. 
 
 289 
 
 surface to warm any given building, taking, of course, the most unfavor- 
 able cases, and allowing for losses between the boilers and heaters, and for 
 the ventilation. 
 
 Another method of expressing the transfer from heaters to building is 
 by the number of units of volume which can be warmed by a unit of 
 heating surface, and this, of course, varies with the proportions of the 
 building and the range of external temperature, but the application must 
 be to buildings of ordinary proportions, and though more commonly used, 
 is really less reliable in its results. 
 
 With the practice of the Dubuque Steam Supply Company, at Dubuque, 
 Iowa, we find that with the external air ranging to F. or 18 C., 1 
 square foot of heating surface warms a number of cubic feet as follows, in 
 columns 2 and 4. 
 
 WHEN HEATEBS ABE IN 
 SAME ROOMS. 
 
 
 
 
 
 
 
 Cubic ft. per 
 sq. ft. 
 
 Cubic metre 
 per square 
 metre. 
 
 Cubic ft. per 
 sq. foot. 
 
 Cubic metre 
 per square 
 metre. 
 
 Dwellings 
 
 50 
 
 15 
 
 40 
 
 12 
 
 Stores wholesale 
 
 125 
 
 37 
 
 100 
 
 30 
 
 % " retail 
 
 100 
 
 30 
 
 80 
 
 24 
 
 Banks ) 
 
 
 
 
 
 Offices V 
 
 70 
 
 21 
 
 60 
 
 18 
 
 Drufj stores ) 
 
 
 
 
 
 Drv goods 
 
 80 
 
 24 
 
 70 
 
 21 
 
 Hotels, large 
 
 125 
 
 37 
 
 100 
 
 30 
 
 Churches 
 
 200 
 
 60 
 
 150 
 
 45 
 
 
 
 
 
 
 WHEN HEATEBS ABE IN 
 BASEMENTS AND WABM 
 
 To reduce these numbers to cubic metres warmed per square metre 
 we multiply by 0.3, obtaining columns 3 and 5. 
 
 B. Various Systems in the United States. 
 
 In the United States the application of steam for heating was begun 
 in 1842, by J. J. Walworth and Joseph Nason, and the first building heated 
 by steam was a cotton mill in Portsmouth, N. H. The exhaust steam from 
 the engine was used very successfully, and from that time to the present 
 there has been a steady increase in the number and magnitude of the 
 works constructed for this purpose. This has been owing to the severity 
 of the climate and the large number of new buildings erected in the rapid 
 growth of the country. The business of the Walworth Manufacturing 
 Company, of Boston, Mass., in constructing steam heating plant is now 
 $1,500,000 per annum, and there is in the United States a business estimated 
 by competent authorities at $6,000,000 per annum, which has for the past 
 30 years averaged $2,000,000 per year. In other words, there is now in- 
 vested in the United States about $60,000,000 in steam heating apparatus, 
 
290 STEAM USING; OB, STEAM ENGINE PRACTICE. 
 
 so that in this subject there is no lack of precedents for many kinds of 
 apparatus. 
 
 There are two classes of plant used, as was indicated by the-columns 
 2 and 4 for the table above. 
 
 When the heaters are placed in the rooms to be heated, the heating is 
 said to be direct. When the heaters are used to warm air in separate 
 chambers, which air is then transferred through flues to the rooms to be 
 warmed, the heating is called indirect. 
 
 The choice between these systems is to be governed by other condi- 
 tions. Where the air is not renewed frequently, or where the space to be 
 warmed is large in plan but not in height, the direct system appears pre- 
 ferable; but where large ventilation is required, or the building is lofty, 
 the indirect method offers many advantages. It appears necessary to pro- 
 vide more heating surface by the latter method; but the labor and 
 expense of fittings is often much reduced thereby, while with improved 
 arrangements there is little difference in the surface required. Usually 
 the movement of the air by which the indirect system heats a building is 
 effected by gravity, but in some instances fans have been employed, re- 
 quiring, of course, power to drive them. 
 
 In regard to the different forms of heaters used in the United States 
 there is great variety, from the single line of large pipe and the manifold 
 lines, where the steam flows through several pipes side by side, to the 
 elegantly finished work used in all the large hotels, and the complicated 
 heaters used with the indirect method. 
 
 Within the last three years preceding 1881 steam heating has assumed 
 a new form, by the use of long lines of pipes underground, thus placing 
 the subject upon the basis of a gas or water supply. Companies for this 
 purpose have been formed and works put in operation in many places in 
 the United States. The table opposite gives some information concern- 
 ing these works. This list is being rapidly extended, and after passing 
 its experimental stage the subject can be placed upon a good financial 
 basis. At the present time, although the matter is a success from the 
 physical standpoint, yet rates and charges are still in an unsettled con- 
 dition, and the owners by no means satisfied in most of these places. 
 
 From experiments with pipe laid and protected, as such companies 
 protect them, a loss of heat of the pipes is found at 50 British heat units 
 per square foot per hour with steam at 258 F., the ground at say 58, 0.05 
 pounds per square foot per hour being the condensed water by weight. 
 Experiments by the writer gave the same value, but in long lines a further 
 loss takes place by leakage and by imperfect traps, a very essential part 
 of the system. In fact the experience at Dubuque is that about one-half 
 of the steam made is wasted, according to the statement of the superin- 
 tendent. 
 
 Steam for heating is carried at all pressures in the United States, and 
 while, as well known, no economy in fuel can result from the use of high 
 pressures, yet a smaller plant can be made to do the work, and, in fact, 
 the first remedy in cold weather for cold rooms is to raise the steam pres- 
 
STEAM HEATING. 
 
 291 
 
 sure until the increased energy of transmission of the heaters produces 
 the desired result. 
 
 City and State. 
 Lockport, New York 
 
 Pipe Underground. 
 .3 miles. 
 800 feet 6 inches 
 
 Dubuque, Iowa 
 
 Auburn, New York 
 Detroit, Michigan 
 
 Milwaukee, Wisconsin. 
 
 
 1,960 
 3,818 
 4,441 
 205 
 2,073 
 2,508 
 643 
 1 mile 
 
 5 
 4 
 3 
 2 
 Ifc " 
 154 " 
 1 
 
 et 6 inches 
 8 
 smaller than 6 in 
 ' 10 inches 
 
 >rk. 
 
 1 
 
 5,000 fe 
 
 2,000 
 5,000 
 ' 585 
 
 "1 
 
 I 2,540 " 8 
 
 I 5,156 " 6 
 
 <! 875 " 5 
 
 |5,195 " 4 
 
 1,300 " 3 
 
 (.6,900 " smaller pipe. 
 
 Remarks. 
 
 Heats 3,500,000 cubic 
 feet space; has 100 
 consumers 
 
 Boiler capacity 25,000 
 Ibs. water per hour, 
 from 40 F., at 280 F. 
 evaporation. 
 
 Heats 6,000,000 cub. ft. 
 
 Heats 5,000,000 cubic 
 feet and runs 10 en- 
 gines. 
 
 Boilers for steam heating are of all sorts and kinds, and the only point 
 which is vital in designing them is to keep in mind the range of action to 
 which they will be subjected. For the work of a boiler in making steam 
 for heating is more like that of a locomotive boiler than anything else. 
 Every degree change of temperature, and every change in the wind is felt 
 by the men with the shovels, and quickness of steaming and capacity of 
 furnace for burning fuel is essential. Grate surface enough must be pro- 
 vided to do the w r ork in the coldest weather; and this grate will be too 
 large for economic evaporation in milder weather; and while large boiler 
 surface is a good thing it is not judicious to invest in a boiler large enough 
 to work at a high economy of evaporation during a few days only in a year 
 of the hardest work. It is more economical to crowd the boilers at the ex- 
 pense of the fuel at such times, and a boiler must be provided which can 
 be crowded hard. 
 
 C. Apparatus and Experiments Made by the Author. 
 
 In the winter of 1878 the writer placed before the directors of 
 Washington University in the City of St. Louis, Missouri, a plan for heat- 
 ing a portion of the buildings belonging to the University by steam, 
 which plan was adopted by them and built. 
 
 The central group of buildings consists of the Academy, the Museum 
 of Fine Arts, the Manual Training School, the University, Laboratory, and 
 Gymnasium; the three latter are called collectively the University. To 
 the west is the Mary Institute, with its own boiler, and to the east the Law 
 School, occupying the old building erected for the Mary Institute; the 
 future of this building being uncertain, it has not been connected for 
 steamheating. The Mary Institute heating apparatus is mainly indirect, 
 
292 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 and consists of beaters hung in the basement in small air chambers con- 
 nected by flues to the different rooms. The air heated is taken either from 
 the basement or from out-doors as desired. The operation is quite satis- 
 factory; the steam is at present supplied by two boilers in the building, 
 but it is probable that it will be connected to the central group and oper- 
 ated from the main boiler house in a short time. It has now been in use 
 for four years. The condensed water returns directly to the boilers. 
 The heaters of cast-iron are corrugated castings short and placed hori- 
 zontally. 
 
 The Academy and Museum of Fine Arts are heated by fittings put in 
 by the Walworth Manufacturing Company. In the Academy cold air is 
 taken through openings in the walls close to the heaters, and the foul air 
 passes through flues from the rooms to the top of the building and 
 escapes. The steam and return mains are led around the basement, and 
 vertical steam and return pipes rise through the buildings with one heater 
 connected to them on each floor. As little horizontal pipe is used as pos- 
 sible, and that is kept in the basement. The steam pipes rise all the way 
 and the return pipes fall all the way. The horizontal steam main must 
 be kept dry and the return main full of water in order to prevent 
 what is known as "snapping" a phenomenon sure to attract attention 
 when it does occur, and which is sometimes dangerous to the joints of 
 the pipes. 
 
 "Snapping" is caused in this way: when any of the condensed water 
 finds its way into steam that is warmer than itself, it causes a sudden con- 
 densation, and the steam closes up so rapidly that a shock and violent 
 sound result. With large pipes and well-defined currents of steam the 
 water condensed remains at the temperature of the steam and is swept 
 along with it. To illustrate more fully, if, in the Academy building, a 
 heater on the upper floor is shut off, steam will condense in the upper 
 portion of the vertical stand-pipe, and there being no current in the upper 
 portion and not much below, the water stands in drops on the iron cool- 
 ing and accumulating till it falls into the hotter steam below. A sound 
 like the crack of a rifle is the result. The remedy is to open the heater 
 on the summit of this pair of pipes, or to connect the stand-pipes them- 
 selves; a better but more expensive way is to put in a pair of vertical pipes 
 for each heater from the mains in the basement, with the valves at the 
 bottom. The greater portion of the University buildings is heated by 
 tubular heaters in the basement, the warm air being led to the various 
 rooms by flues in the walls and moved only by gravity. The old heaters 
 were of cast-iron heated by fires made in them, and the new heaters were 
 applied to the same system of flues. The heaters were described in a 
 paper by Mr. Chas. F. White before the Club, and printed in The American 
 Engineer. Certain rooms in the upper floors are unprovided with wall 
 flues and are warmed by direct heaters; as these heaters are connected to 
 a pair of horizontal pipes of considerable length, the condensed water does 
 not drain properly. The drying tables and sand baths in the chemical 
 and physical laboratories and one small heater is placed in a small room 
 
STEAM HEATING. 
 
 293 
 
 on the first floor of the south wing, and one on the upper or second floor 
 of the Laboratory building. The Gymnasium is to have heaters of the 
 kind in the upper floor, with horizontal mains in the basement. At the 
 west end of the University building, the heating surface is not sufficient 
 to make up the loss from the walls, and the upper floors draw off the heat 
 from the lower rooms. This could be prevented by controlling the area of 
 the flues which carry away the warm air, but will be remedied by placing 
 heaters in the lower room. With the exception of two rooms the heating 
 is ample in the coldest weather. 
 
 The Manual Training School is heated in part by direct and in part by 
 indirect radiation. The two lower floors with the four workshops are 
 warmed by 4 lines of 2 -inch pipe along the foot of the walls under the 
 benches, and the upper or school-room floor by a pair of tubular heaters in 
 the basement, and one room has two lines of pipes on three sides. The 
 steam is taken either from the steam main or from the exhaust steam of the 
 n on -condensing engine used for running the tools in the workshops. The 
 air from the heaters in the basement is conveyed to the upper floor by metal 
 flues passing through the floors. 
 
 The dimensions of the buildings and surfaces met with are given 
 for the central group in the following table with French and English 
 units: 
 
 APPROXIMATE DIMENSIONS OF BUILDINGS IN THE CENTBAL GBOUP. 
 
 
 Academy. 
 
 Manual 
 Training 
 School. 
 
 Museum 
 of 
 Fine Arts. 
 
 Gym- 
 nasium 
 
 Uni- 
 versity. 
 
 Volume, c. f 
 
 450,000 
 
 225,000 
 
 500,000 
 
 100,000 
 
 750,000 
 
 Volume, c. m 
 
 12,600 
 
 6,300 
 
 14,000 
 
 2,800 
 
 21,000 
 
 External surface in sq. ft. .. 
 External surface in metres. 
 Heating surface in sq. ft 
 Heating surface in metres. . 
 Volume c. f . to 1 sq. ft. heat 
 ing surface. 
 
 36,000 
 3,348 
 3,500 
 323 
 
 129 
 
 20,000 
 2,418 
 1,870 
 174 
 
 120 
 
 36,000 
 3,348 
 3,300 
 307 
 
 152 
 
 13,000 
 1,200 
 500 
 46 
 
 200 
 
 80,000 
 7,440 
 6,000 
 558 
 
 125 
 
 To 1 sq. ft. of wall 
 
 12 8 
 
 11.2 
 
 13 9 
 
 7 7 
 
 9 4 
 
 Cubic metres to 1 sq. metre 
 heating surface 
 
 30 
 
 36 
 
 45 
 
 60 
 
 38 
 
 To 1 sq metre wall 
 
 3.8 
 
 3.3 
 
 4.2 
 
 2.3 
 
 2.8 
 
 External surface to heating 
 
 10 3 
 
 10.7 
 
 10.9 
 
 26. 
 
 13.3 
 
 
 
 
 
 
 
 The boilers are three in number, set independently, two being used at 
 once. They have each 768 square feet of heating surface, and 24 square 
 feet of grate. The stack is 12.25 square feet aperture at the top, and is 105 
 feet above the grate bars. 
 
 All fuel is weighed, and all water, fresh or return, is measured either 
 by metre or weight as desired. The coal used is bituminous, from the 
 neighboring mines in Illinois, and has a chemical composition capable of 
 
294 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 evaporating 12 units of water by 1 unit of coal by weight, from and at 
 212 F.; ash, 12 per cent. 
 
 Experiments of March, 1880. Fuel and water weighed for one week; 
 one boiler; maximum coal per square foot of grate, 38 pounds; mean eva- 
 poration from and at 212 F., 6.49 pounds; priming, 2 per cent. 
 
 Experiments of October, 1880. Coal weighed and water by meter 
 (Worthington) for one week; two boilers; mean evaporation for the week, 
 7.1 pounds; maximum for 24 hours, 7.9 pounds. 
 
 As the meter was a piston meter, the results are not likely to be in ex- 
 cess of the truth on that account. The weight was charged each time, full 
 and empty barrel of water, and the time noted. 
 
 The priming by the method of Hirn. The work done was exceedingly 
 varied in the March experiments; it was found that at that time the work 
 from 6 A. M. to 12 noon was double that done in the other 18 hours. 
 
 Experiments upon the transmission of heat have been made upon the 
 Academy and University buildings only, the quantity of water condensed 
 being noted, with the steam pressure, the temperature of the external air, 
 the air in the buildings and of the return water. 
 
 With external temperatures from 10 F. to 45 F., and the buildings 
 from 60 F. to 75 F. } steam from 260 F. to 280 F., the values already given 
 for a were found. The duration of the experiments was from three 
 to eight hours, taken during the ordinary operation of the works. The 
 return water was usually from 190 F. to 210 F. The experiment for 
 underground condensation was made in the same way, and lasted five 
 hours.* 
 
 The construction of the boilers was shown in the paper by Mr. "White 
 above referred to. 
 
 There is one element, not yet mentioned, and that is the time in which 
 the buildings must be warmed. In the experience of the writer at the 
 University buildings the Academy can be warmed in cold weather in three 
 hours, and in fact the steam is only supplied for twelve hours out of the 
 twenty- four. The University building with the indirect heaters has to 
 be kept warm all the time, and in cold weather takes ten or twelve hours 
 to get warm throughout. For rapid heating, the direct system with ample 
 surface appears best adapted; but for steady heating, with purity of air, 
 the indirect is to be preferred. 
 
 D. Designing a System. 
 
 Suppose that it is required to design the heating apparatus for a 
 large building. This will include the boilers, the heaters and their dis- 
 position, and the choice of a system direct or indirect, and the use of ex- 
 haust steam from an engine. 
 
 *In the experiments with the buildings the ample ventilation was not interfered 
 with. The other buildings have just been completed, and the experiments made w}th. 
 them are only preliminary. 
 
STEAM HEATIXi*. 295 
 
 To fix our ideas let us consider the case to be that of a cotton mill with 
 the following dimensions: 
 
 In English measure, 328 feet long, 40 wide and three stories of 13 feet 
 in height, making 40 feet say as the height, rectangular in plan, having 
 a volume of 52-1,800 cubic feet and a surface of 42,560 square feet. 
 
 In French measure, 100 metres long, 12 metres wide and 12 metres 
 high, having a volume of 14,400 cubic metres and a surface of 3,888 square 
 metres. 
 
 Such a mill will contain 10,500 spindles, 225 looms and the proper pro- 
 portions of cards, with the other machinery belonging to the manufacture 
 of cotton. 
 
 Let the lowest external temperature be assumed at 5 F. = 15. C., 
 and let the minimum internal temperature be assumed at 59 F. = 15 C., 
 values which would be likely to suit most localities in the United States, 
 and even if exceeded would not be very often passed. The range of wall 
 transmission will then be from 59 to 5 = 54 F. = 30 C., and the quan- 
 tity of heat transferred may reach 54 x 1.25 units per square foot per hour 
 or 30 x 6.0 calories per square metre per hour 67.5 English or 180 French 
 units. 
 
 67.5 x 42,560 = 2,872,800 heat units per hour, or 
 180 x 3,888 = 699,840 calories per hour. 
 
 To decide upon the amount of heating surface the temperature of the 
 steam must be known. The choice of the direct or indirect methods will 
 probably be made from the form of the building, which covers a large area 
 and is not very high, and the magnitude of the rooms, supposed to be not 
 more than two to a floor, as direct surface; and- the kind of heaters by 
 economy only, to be pipes laid along the foot of the walls, or rather along 
 the walls near the floors. 
 
 The temperature of the steam will much depend upon where it comes 
 from, from a separate boiler or from the exhaust of an engine, and we will 
 examine three cases: 
 
 1. A separate boiler. 
 
 2. A non-condensing engine. 
 
 3. A condensing engine. 
 
 With a separate boiler we are not limited as to pressure except by con- 
 venience, and we will assume 50 pounds per square inch above the at- 
 mosphere, 65 pounds absolute, or 4J atmospheres, the temperature of 
 the steam being 297 F., or 148 C. 297 59 = 238 F. = T T = 
 
 148 _ 15 = 133 C., q=( T '~~ T ^ = 566.44 heat units per square foot per 
 
 hour = 1,528 calories per square metre per hour. 
 
 The slight difference in these results is due to a neglect of decimals and 
 is of no practical value. 
 
 The boiler, to give this amount of heat, will have to evaporate 2,873 
 pounds of water per hour, or say 3,000 pounds, and will require, at 4 
 pounds of water evaporated per 1 square foot per hour, an amount of heat- 
 
296 STEAM USING; OR, STEAM ENGINE PRACTICE. 
 
 ing surface = 750 square feet. As this is the maximum capacity, we find 
 that 24 square feet of grate, with coal evaporating 6 pounds of water 
 per 1 pound coal, burning 500 pounds per hour, or about 22 pounds coal 
 per square foot of grate, is an ample provision. A smaller grate, with 
 careful firing, would give better results for fuel, but would not be as easy 
 to work on cold days. 
 
 The cost of such a boiler set in the United States would not be far 
 from $1,300, including everything ready to une, but not counting any out- 
 lay for buildings to put it in. The cost of the pipe in place to make 4,460 
 square feet of surface for 1-inch or 2-inch pipe would be about 40 cents per 
 square foot, or $1,856. Total cost, $1,856 + $1,300 = $3,156, and including 
 contingencies, say $3,200 at 5 francs per dollar, 16,000 francs, of which the 
 boiler cost 6,500 and the pipes 9,100 francs. 
 
 Suppose the steam had been at 212 F. or 100 C., 212 59 = 153 F., 
 in place of 238 for T' T, but the transmission is now reduced, and the 
 surface must now be increased 1.55 times to 7,192 square feet, say 7,200 or 
 to a cost of 13,950 francs; total, say 31,000 francs. The steam temperature 
 is now at the lowest possible in a non- condensing steam engine; and if 
 such an engine, using 2,800 pounds of steam per hour, be at hand, the only 
 outlay involved is the pipe surface of 14,000 francs, say; but with the 
 more active circulation of the steam there will be found an increase in the 
 radiating effect, which, however, we will not consider here. 
 
 We have then a decrease in the cost of the plant of 400 dollars, or 2,000 
 francs, which is equivalent to say 40 dollars or 200 francs per year interest 
 and repairs, and we should, of course, recommend the use of exhaust steam. 
 
 There remains to be considered the use of a condensing engine, with the 
 exhaust steam at say 100 F. = 38 C. 
 
 It would be possible, by the use of a very large amount of pipe surface 
 and a very carefully arranged drainage, to return to the air pump; but the 
 great cost of pipe surface and the practical troubles of making pipe joints 
 hold in vacuum would cause us to reject this idea as impracticable, and there 
 remains the use of a separate boiler, or the running of our condensing 
 engine as a non -condensing engine during the winter months. 
 
 For ordinary condensing engines the increase of fuel would be in pro- 
 portion to the increase of work on the forward stroke, rendered necessary 
 by the increase of back pressure, and in such cases it would be desirable 
 to use a separate boiler, as the fuel used would be great; for example, the 
 engine using 400 I. H. P., with 4 pounds of coal per H. P. per hour, or 1,600 
 pounds of coal per hour. The increase of fuel may be measured by the 
 increase of forward work, which, of course, depends upon the engine used; 
 but if the forward mean pressure had been 30 pounds and the mean back 
 pressure 3 pounds per square inch when running condensing, the forward 
 pressure will now be raised to 43 pounds, and the back pressure to 16 
 pounds, the fuel from 1,600 to 2,292 pounds, or say 700 pounds per hour 
 increase, between three and four francs per hour, say 40 francs per day for 
 the maximum work in heating, but this must be kept up for 100 days, or a 
 cost per year of 4,000 francs. 
 
STEAM HEATING. 297 
 
 When we compare this with our separate boiler we have an excess of 
 2,000 francs to begin with, but also the cost of the fuel in addition, which 
 cannot be taken less than 1,800 per year, as estimated below, making 
 together, say 2,000 francs, and we should in this case use the separate boiler 
 and engine. 
 
 When, however, we have the mill in good order, and have the best 
 engine in use, that of M. Hirn, using superheated steam, we find that 
 the power is not over 250 horse-power and the fuel 2 pounds per hour 
 per I. H. P., or 500 pounds per hour in place of 1,600, the increase 
 500f* _ 500 = 217 pounds, which costs about one franc per hour, or say 
 $200 to $230, or 1,000 to 1,200 francs per year, on the above basis of 100 
 days of 12 hours. 
 
 The cost of the fuel alone for the separate boiler on the basis given 
 above of a maximum of 500 pounds, and say an average of 300 pounds of 
 coal per hour, would be from 30 to 40 cents, or 1-fc to 2 francs, say for the 
 100 days of 12 hours; $360, or 1,800 francs, as estimated above, which, with 
 the cost of the boiler, would leave us $160, or 800 francs per year in favor 
 of this plan. 
 
 We find then that with non- condensing engines and the best class of 
 condensing engines, the use of exhaust steam is desirable, while with ordi- 
 nary condensing engines a separate boiler is to be preferred. 
 
 The pipe surface for 7,200 square feet can be arranged to take the 
 exhaust steam from the engine through an 8-inch or the 10-inch exhaust 
 pipe to the top of the building, and to open into 2-inch pipes on each floor, 
 in two directions, uniting in a descending 10-inch pipe carried to a hot 
 water tank, and the exhaust to be either circulated or turned loose into the 
 air without going through the building. 
 
 Seventy, two hundred feet would require 14,400 feet of two-inch pipe 
 and as we can easily place 300 feet in one line and 600 feet in the two lines, 
 we have twenty four lines, or four lines of 2-inch pipe on each of the two 
 main walls. The end walls and vertical pipes will make up the amount. 
 Twenty-four lines of 2-inch pipe will present the same resistance that one 
 line of 10-inch, and we need fear no great increase of back pressure above 
 that assumed. (Two-inch pipe is 0.05 m. diameter.) The wall pipes should 
 fall uniformly 1 in 200. Each line of pipe should have its own valve con- 
 nections, and when most of them are closed the steam allowed to flow 
 directly into the air, as well as through the building, to avoid back pres- 
 sure. One and one- quarter inch pipe is usually preferred to larger sizes, as 
 being easier to bend without fracture. 
 
 In the above recommendation of the use of exhaust steam by discard- 
 ing the condenser of an engine, it is, of course, supposed that both boilers 
 and engine can do the required work under the new conditions with entire 
 safety and satisfactorily. This can only be ascertained in the particular 
 instance by a careful and complete examination of the conditions under 
 which the engines are now working, and the change should not be made 
 until the result is clearly foreseen, for the example supposed the best state of 
 things, and for all but the best kind of condensing engines and mills in the 
 
298 
 
 STEAM USING; OR, STEAM ENGIXE PRACTICE. 
 
 best order, we shall not find it advantageous to make so novel a departure, 
 and shall use the separate engine and boiler. 
 
 To get the pipe surface three lines of 2 inch pipe in each floor may 
 then be employed. 
 
 In the United States the cost of a given surface of pipe is about the 
 same for 1-inch and 2-inch pipe, the labor making up the cost to the same 
 amount. With the larger sizes the cost increases, but for the use of exhaust 
 steam the larger pipe is to be preferred, unless more than one line be used 
 at once in the example above. 
 
 [NOTE. The conclusions of the above paper have been borne out by the 
 experience of the winters 1880-81 and 1881-82. C. A. SMITH.] 
 
 UNIVEESITY 
 
RETURN 
 TO ^ 
 
 CIRCULATION DEPARTMENT 
 
 202 Main Librar 
 
 LOAN PERIOD 
 HOME USE 
 
 ALL BOOKS MAY BE RECALLED AFTER 7 DAYS 
 
 Renewals and Recharges may be made 4 days prior to the due date. 
 
 Books may be Renewed by calling 642-3405. 
 
 DUE AS STAMPED BELOW 
 
 1 9, 1989 
 
 
 
 AUTO, w.wu 
 
 : 
 
 
 icr 1 ? uo; 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 UNIVERSITY OF CALIFORNIA, BERKELEY 
 FORM NO. DD6 BERKELEY, CA 94720 
 
U C BERKELEY LIBRARIES , 28 84 
 
 mini 
 
 03210^737