REESE LIBRARY 
 
 UNIVERSITY OF CALIFORNIA. 
 z/)<ressiouNo. 'tf 6 J<. Class M<. 
 
MODEEN 
 GAS AND OIL ENGINES. 
 
A PRACTICAL TREATISE 
 
 ON 
 
 MODERN GAS MD OIL ENGINES 
 
 BY 
 
 FREDERICK GROVER, A.M.InstC.E., M.I.Meeh.E., 
 
 \A 
 
 Consulting Engineer, Leeds. 
 
 (SECOND EDITION.) 
 
 PRICE FOUR SHILLINGS AND SIXPENCE NET. 
 
 1897. 
 THE TECHNICAL PUBLISHING CO. LIMITED, 
 
 31, WHITWORTII STREET, MANCHESTER. 
 
 JOHN HEYWOOD, 
 
 AND 30, SHOE LANE, LONDON; AND RIDGEFIELD, MANCHILSTEB; 
 And all Booksellers. 
 
 .0. VAN NOSTRAND COMPANY, 
 NEW YORK. 
 

 
. 
 
 PREFACE. 
 
 IN writing these pages I have endeavoured to supply to the 
 average mechanical draughtsman the. information necessary to 
 enable him to apply his art to the design of gas engines. A real 
 knowledge of the elements of machines can only be acquired in 
 the fitting and erecting shops. There the eye is trained and a 
 sense of proportion developed which is always helpful in the 
 design of new patterns. Unless, however, a draughtsman 
 thoroughly understands the principles which underlie his art, 
 he is liable to errors which prove themselves as costly as those 
 made by the mere theorist. 
 
 In developing the conception of the work before me, I have 
 first described the general arrangement of a gas-engine plant, 
 then the types of modern gas engines, and have afterwards 
 attempted to explain how their leading dimensions may be 
 calculated. 
 
 The importance of systematic testing is now felt by many 
 engineers. I have therefore described in detail the apparatus 
 required and the calculations necessary to make a complete 
 gas-engine trial. In connection with this subject a chapter is 
 added on the practical analysis of gases. 
 
 The first part of the book concludes with a description of a 
 series of experiments made to determine the effects of 
 products of combustion when present in explosive mixtures of 
 coal gas and air. Some doubt has been expressed as to the 
 accuracy of the conclusions drawn from these experiments, on 
 the ground that the mixing of the gases was imperfect. Having 
 regard to the fact that at least one minute of time elapsed 
 between filling the vessel and igniting the charge, whereas in 
 a gas engine running at 180 revolutions per minute only one- 
 third of a second elapses, I think it improbable that the diffusion 
 of products of combustion in a gas-engine cylinder is more 
 
IV. PREFACE. 
 
 perfect than in my experimental apparatus, notwithstanding 
 the wide difference between the conditions. 
 
 In the second part of the book a brief description is given of 
 the physical properties of oils, of oil-engine vaporisers, and a few 
 special points in connection with oil-engine testing. 
 
 In conclusion, I desire to express my thanks to the following 
 firms for their kindness in assisting me : Messrs. Crossley; 
 Andrews; Dick, Kerr, and Co.; Tangye, Fielding, Wells Brothers, 
 Barker, Dougill, Crosby, Globe Engineering Co., Priestman, 
 and Elliott Brothers. I also desire to acknowledge the assist- 
 ance I have derived from the works of Professor William 
 Robinson, Professor Unwin, Mr. Bryan Donkin, Mr. Dugald 
 Clerk, and from Professor Capper's report on the Royal 
 Agricultural Society's Trials, held at Cambridge. 
 
 F. G. 
 
 Leeds, 
 
 July, 1897. 
 
CONTENTS. 
 
 CHAPTER I. 
 Introduction 
 
 CHAPTER II. 
 Arrangement of Engine-room .............................................. 7 
 
 CHAPTER III. 
 
 Types of Gas Engines: Atkinson's "Differential" and "Cycle" Engines 
 Comparative Cycles of Gas Engines The Crossley Engine -The 
 " Stockport " Gas Engines .......................................... 19 
 
 CHAPTER IV. 
 The " Griffin " Engine ...................................................... 46 
 
 CHAPTER V. 
 The Tangye Engine The Fielding Gas Engine The "Premier " Gas Engine 53 
 
 CHAPTER VI. 
 Self-starters ................................................................ 71 
 
 CHAPTER VII. 
 Two-cycle and other Engines .............................................. 77 
 
 CHAPTER VIII. 
 
 French Engines : Simplex Gas Engines Lenoir Engine Charon Engine 
 Niel Engine Lalbin Engine. German Engines: Benz Engine 
 Koerting-Lieckf eldt Engine The Daimler Gas Engine The Capitaine 
 Gas Engine ........................................................ 86 
 
 CHAPTER IX. 
 Testing Gas Engines : Indicating .......................................... 92 
 
 CHAPTER X. 
 Indicators for Gas Engines : Crosby Indicator The Tabor Indicator The 
 
 Wayne Indicator The Simplex Indicator .......................... 99 
 
 CHAPTER XI. 
 Reducing Gear for Gas Engines ............................................ 109 
 
 CHAPTER XII. 
 Gas-engine Trials .......................................................... 1H 
 
VI. CONTENTS. 
 
 CHAPTER XIII. PAGE 
 The Practical Analysis of Coal Gas 12(3 
 
 CHAPTER XIV. 
 Calculations Required in Working Out Results of Engine Trials 139 
 
 CHAPTER XV. 
 Calculations on the Indicator Diagram 144 
 
 CHAPTER XVI. 
 Gas-engine Trial, Otto Cycle : Coal Gas Used 154 
 
 CHAPTER XVII. 
 Gas-engine Design 159 
 
 CHAPTER XVIII. 
 Producer Gas, and its Application to Gas Engines 176 
 
 CHAPTER XIX. 
 The Effects of the Products of Combustion upon Explosive Mixtures of Coal 
 
 Gas and Air 183 
 
 Appendices to Part 1 200 
 
 PART II. PETROLEUM ENGINES. 
 
 CHAPTER XX. 
 Petroleum 205 
 
 CHAPTER XXI. 
 Oil Engines : Priestman's, Griffin's, Hornsby-Akroyd, Crossley's, Wells', 
 
 Trusty , 215 
 
 CHAPTER XXII. 
 Oil Engines: Campbell's, Britannia, Capitaine, Fielding, The Daimler Motor. 231 
 
 CHAPTER XXIII. 
 Oil Engine Testing 241 
 
MODERN GAS ENGINES. 
 
 CHAPTER I. 
 INTRODUCTION. 
 
 IN writing upon the subject of modern gas engines, it may 
 at first be thought that an historical preface is out of place ; 
 but in glancing through the records of progress made 
 during the last 35 years, it is remarkable how many of the 
 improvements we are apt to regard as new are merely 
 the embodiment of suggestions made by those engaged in 
 bringing the gas engine into practical use. In more than 
 one instance has an important advance been entirely lost 
 sight of until, by more recent investigation, it has been 
 again introduced and submitted to practical tests. These 
 remarks indicate the importance of a concise knowledge of 
 the failures and successes of early inventors, and, if need 
 be, justify the introduction of historical matter. 
 
 It is probable that ever since gunpowder has existed 
 inventors have sought to utilise its potential energy for the 
 use as well as for the destruction of their fellow-men. In 
 support of this assumption, we find records of attempts to 
 explode gunpowder in a cylinder as early as 1678. A 
 practical form of engine might have been evolved had it 
 not been for the mechanical difficulties to be overcome in 
 obtaining continuous and regular explosions with such a 
 substance as gunpowder. It is not surprising, therefore, 
 that at a time when the steam engine was beginning to 
 occupy men's minds, the combustion engine was entirely 
 
2 INTRODUCTION. 
 
 neglected ; nor is it surprising that so long as an explosive 
 substance could only readily be obtained in the form of 
 solid matter, mechanical difficulties impeded all progress. 
 When, however, Dr. Watson discovered that gas could be 
 distilled from coal, and when, in the year 1792, Murdoch, a 
 Cornish engineer, demonstrated the practical application 
 of coal gas for lighting purposes, then we find the internal 
 combustion engine again the subject of experiment. About 
 this time, it will be remembered, the steam engine had 
 become a powerful assistant in the mining industry, and 
 was undergoing important development, brought about by 
 the world-famed W'att. Naturally the majority of mecha- 
 nicians were absorbed in furthering its possibilities. 
 Nevertheless, from the year 1791 to 1801 three patents are 
 on record setting forth the use of explosive mixtures of 
 coal gas and air in motors. The first, by John Barber, 
 explains how a wheel, with vanes upon its circumference, 
 may be driven by releasing the pressure of an explosion 
 through an orifice placed in proximity to the vanes. The 
 second patent, taken out by Street in 1794, mentions the 
 use of cylinder and piston, the latter being driven out- 
 wards by the pressure of the explosion. Flame ignition is 
 first made use of in this engine. In 1801 another patent 
 appeared under the name of LeboD, setting forth the 
 advantages of compressing the gas and air before entering 
 the explosion cylinder. Thus we see that as early as 1801 
 two of the principles had been explained, upon the merits 
 of which later engines became a practical success. It 
 appears curious that so much time should elapse between 
 the publication of these specifications and the production 
 of a really useful gas motor ; but it is highly probable that 
 the two factors contributing to this end were, firstly, the 
 crude mechanical appliances for manufacturing purposes ; 
 and secondly, the tendency to follow the details of small 
 steam-engine design, which in the matter of valves, glands, 
 and pistons are entirely unsuitable for the peculiar condi- 
 tions of gas engine work. 
 
INTRODUCTION. 3 
 
 It is curious to note how the phases of construction of the 
 steam engine, during a period extending over nearly 200 
 years, have been practically repeated in the case of the gas 
 engine in about half the time. Just as in the steam engine, 
 when low-pressure steam was condensed in the cylinder so 
 that unbalanced atmospheric pressure could be utilised, so 
 in the gas engine many attempts were made and these 
 were not without a measure of success to utilise the 
 unbalanced atmospheric pressure due to the contraction in 
 volume of an explosive mixture, rather than the pressure 
 produced during the explosion. In 1823 an English patent 
 was taken out by Brown, giving the details of such an 
 atmospheric engine. The pressure of the explosion was 
 permitted to escape through valves in the piston, which, 
 however, closed when the pressure of the atmosphere ex- 
 ceeded that of the gas. A little water injected into the 
 cylinder assisted in cooling the products of combustion, and 
 so reduced the pressure rapidly below that of the atmosphere. 
 A few of these engines were constructed for commercial 
 purposes, but owing to the difficulty in keeping them going, 
 and to the expense in gas, they were finally abandoned. 
 From 1823 to 1838 but little was done to popularise the gas 
 motor, although during that lapse of time the water jacket 
 was added to the cylinder for cooling purposes, and some 
 attempts were made to govern the speed of an .engine by 
 controlling the admission of gas. 
 
 In 1838 Barnett made practical use of the principle of 
 compression by constructing a single-acting motor cylinder, 
 at each side of which was placed a gas and air pump. The 
 three pistons moved simultaneously in the same directions ; 
 the gas and air pumps delivered into a receiver, from which 
 communication was made to the motor cylinder as the 
 charge was ignited. Barnett afterwards constructed double- 
 acting engines, but with less success. 
 
 Between the years 1838 and 1860 nothing of much practical 
 value was added to the gas motor. There were a considerable 
 number of patents registered during this period, describing 
 
4 INTRODUCTION. 
 
 engines working with oxygen and hydrogen gases, but no 
 real progress was made until 1860. About this time it 
 became recognised that double-acting cylinders were not 
 suitable for gas engines, on account of the necessary absence 
 of compression, the intensity of heat, the difficulties of 
 packing and keeping a narrow piston in order. All these 
 points were fully exemplified by the working of the Lenoir 
 engine, which made its appearance about 1860. Although 
 exhibiting those weak points of design, the Lenoir engine, 
 owing to commercial agencies, became popular. The details 
 of construction resemble somewhat those of an ordinary 
 steam engine : a slide valve, S ports for admission and escape 
 of the gases, electric ignition, water- jacketed cylinder and 
 covers, were amongst the special features of the engine, 
 Notwithstanding the consumption of gas amounted to about 
 100 cubic feet per horse power per hour (a fact probably not 
 known to purchasers), this engine had a large sale. The 
 defects of working ruined the future of the engine; and 
 although it was subject to alteration in the hands of Hugon 
 by the introduction of flame instead of electric ignition, it 
 never recovered the reaction of opinion set up by public 
 disappointment. 
 
 In 1843 Joule determined the mechanical equivalent of 
 heat, and from that time attention was drawn to the 
 question of thermal efficiency of motors, although it was 
 some years before this treatment of the subject was 
 thoroughly understood. In 1860 Barsanti and Matteucci, 
 appreciating the convertibility of heat into motion, pointed 
 out the necessity of rapid expansion after explosion, in 
 order that the heat might be utilised in doing work, instead 
 of the greater part being transmitted through the walls of 
 the cylinder to waste itself in heating the jacket water. 
 The engine constructed by Barsanti and Matteucci never 
 got beyond the experimental stage, in spite of the correct 
 principle upon which it was worked. In 1863 Otto and 
 Langen, having also conceived the necessity for rapid 
 expansion, devised a motor somewhat similar to Barsanti's, 
 
TRK r 
 
 UNIVERSITY 
 
 INTRODUCTION. 5 
 
 making use of a free piston, which was shot upwards in a 
 vertical cylinder by the force of the explosion. In its 
 upward passage the weight of the piston only was overcome, 
 but on its downward stroke the piston rod, made in the 
 form of a rack, and geared to a pinion on the flywheel axle, 
 caused the latter to rotate. A special ratchet gear was 
 fitted to the pinion, so that it could revolve freely upon the 
 flywheel axle during the upstroke of the rack. A few 
 engines of this type are working at the present time, though 
 they, are regarded as curiosities amongst gas motors. The 
 consumption of gas was only about 40 cubic feet per brake 
 horse power per hour, but the noise and vibration were 
 sufficiently objectionable features to prevent its gaining 
 great popularity. 
 
 About this period (1861) it was again pointed out by 
 Gustave Schmidt and Million that economy would be 
 effected by compression before explosion. Million attempted 
 the practical illustration of the principle of compression 
 first upon lines similar to those of Barnard in 1838, but 
 afterwards he introduced the compressed gas into the 
 cylinder at the dead point, providing for that purpose an 
 equivalent to the clearance space of a gas engine cylinder 
 of the present day. 
 
 In 1862, a French engineer, Beau de Rochas stated 
 definitely the conditions necessary for maximum efficiency, 
 and formulated the well-known cycle, now termed the Otto 
 cycle because carried into practical effect by one of that 
 name. In the wording of the French patent taken out in 
 1862, Beau de Rochas manifests a clear and advanced know- 
 ledge of his subject. He says (translated) : There are four 
 conditions for perfectly utilising the force of expansion of 
 gas in an engine : 
 
 I. The largest possible cylinder volume contained by a 
 minimum of surface. 
 
 II. The highest possible speed of working. 
 
 III. Maximum expansion. 
 
 IV. Maximum pressure at the beginning of expansion. 
 
6 INTRODUCTION. 
 
 In connection with his first-stated condition, he goes on to 
 say that a cylinder of the largest possible diameter for a 
 given expenditure of gas should prove the most economical, 
 and he deduces from this that only one explosion cylinder 
 should be used in each engine. Following this significant 
 conclusion, he touches upon the effect of time of expansion 
 in causing the transmission of heat to the jacket water, and 
 states definitely that the more rapid the piston speed, and 
 consequently the expansion, the less will be the loss of heat 
 to the jacket water. Speaking of maximum expansion, he 
 points out that the lowest limit of pressure is soon attained, 
 because of the rapid condensation of the gases. To obtain 
 a maximum expansion, therefore, the gases should be 
 compressed when cool to raise the initial pressure of 
 explosion. In connection with this he also points out 
 that the obvious ' limit to compression must occur when 
 spontaneous combustion takes place. 
 
 Steam-engine makers, in taking up the manufacture of 
 gas engines, have a tendency to think that the best results 
 are only possible by a system of compounding gas engines. 
 It is, indeed, natural that such should be the case. It should 
 be remembered, however, that the best modern engines are 
 working upon the precise principles laid down in 1862 by 
 Beau de Rochas, and that their success testifies to the 
 soundness of his views. Economy cannot be obtained 
 by compounding, inasmuch as a great increase in com- 
 pression is limited by spontaneous combustion, and 
 expansion in two cylinders necessitates an increase in 
 surface. Beau de Rochas anticipated in his specification 
 still more, for he advocated combustion by compression, 
 instead of by the electric spark, and finally gives a hint as 
 to the value of a timing valve. 
 
 At this juncture, perhaps, the author may be permitted a 
 digression, to arrest the attention of those apt to scoff at 
 the value of scientific work in connection with the advance- 
 ment of practical engineering. The position of science in 
 relation to the development of the steam engine has been 
 
ARRANGEMENT OF ENGINE ROOM. 7 
 
 ably defended from the attacks of practical men by Prof. 
 Unwin in his "James Forrest" lecture, published in 
 ^volume cxxii. of the Proceedings of the Institution of Civil 
 Engineers. What stronger evidence than the history of 
 the gas engine could be adduced to prove that its ultimate 
 success depended upon the scientific principles laid down 
 by Beau de Rochas, who, although not concerning himself 
 with the practical details of a gas engine, so thoroughly 
 understood the scientific side of the question. Is it not 
 rather a reflection upon the so-called practical engineer that 
 it took fifteen ye&rs to bring about anything like a practical 
 development of the principles enumerated in 1862 1 
 
 The Otto and Langen engine was the most practical 
 form of gas motor in the market from 1863 to 1876. Many 
 attempts were made to improve it, but none were effectual 
 until Otto himself, in 1876, brought out an entirely new 
 design of the type now familiarly associated with his name. 
 This was the first successful engine designed by one of 
 high scientific attainments in which compression took 
 place in the explosion cylinder ; and its production marks 
 the period at which the gas motor became a serious 
 competitor against small steam engines. 
 
 CHAPTER II. 
 ARRANGEMENT OF ENGINE-ROOM. 
 
 THE reader being now acquainted with the main points in 
 connection with the historical side of the subject, we may 
 proceed at once to a description of the details of an Otto 
 gas engine of the type brought out in 1876, many of which 
 are working at the present time. Although Otto engines of 
 later design exhibit many important improvements in the 
 details which will be described on another page we shall 
 not get far astray in accepting the arrangement of gas 
 
8 ARRANGEMENT OF ENGINE-ROOM. 
 
 connections, jacket water circulation tanks, &c., as appli- 
 cable to any gas engine at present on the market. On 
 account of the unique position of the Otto engine as the 
 pioneer of all practical gas motors, and because nearly all 
 more modern engines work upon the Otto cycle, it is 
 necessary for the sake of completeness to treat of it here. 
 
 The general outside appearance of the engine is some- 
 what similar to that of a small horizontal steam engine 
 with overhanging cylinder, and a trunk piston, instead of 
 piston rod and guides. Fig. 1 is a sectional plan. The 
 piston B is shown at the back of its stroke ; the space C is 
 the clearance, provided in all gas engines, in which the 
 charge is compressed before ignition takes place. In this 
 engine the clearance volume is about 50 per cent of the 
 volume swept out by the piston. The passage I is for the 
 admission of air and gas, and also serves for the passage of 
 flame in igniting the charge. 
 
 The burnt gases are discharged through the valve D, 
 lifted by means of a lever worked by a cam on the lay shaft 
 P. The slide M works to and fro between the fixed part N 
 and the back of the cylinder, and effects both the admission 
 of air and gas and the igniting of the charge. The slide 
 M in moving upwards causes communication between the 
 air passage F, the gas port G, and the entrance to the 
 cylinder I. Air and gas are thus drawn into the cylinder 
 upon the outstroke of the piston. The spacing of the port 
 edges is so arranged as to allow the gas to enter the cylinder 
 just after the air port is shut. In this position the port L 
 is filled with gas at atmospheric pressure, which is ignited 
 by momentary contact with flame. The slide M then moves 
 rapidly downwards ; meanwhile the piston B is returning to 
 the back of the cylinder and compressing the charge to 
 about 40 Ib. per square inch ready for explosion. It is evident 
 that the burning gas entrapped in the port L would be 
 extinguished by the pressure of gas in C the moment L 
 communicates with I, and ignition would not occur. In 
 order to raise the pressure in the port L on its way to the 
 
ARRANGEMENT OF ENGINE-ROOM. 9 
 
 inlet I, there is a very small passage provided from the 
 clearance space communicating with the port L before it 
 reaches I. Through this passage the compressed gas finds 
 its way in very small quantities into L, raises the pressure 
 in L, and maintains the combustion by the additional supply 
 of air and gas. The passage through which this com- 
 munication is made is too small to permit the firing of the 
 charge in the cylinder, and this does not take place until 
 the port L carrying its burning gas now at the same 
 pressure as that in C reaches I and ignites the charge. 
 
 FIG. 1. Sectional plan of Otto gas engine. 
 
 The piston is now driven forward by the explosion. The 
 flywheels carry the piston back, driving the burnt gases out 
 through D. It will be observed that an impulse is only 
 given every two revolutions of the engine ; hence the 
 necessity for exceedingly heavy flywheels to maintain a 
 constant speed. The Otto cycle, upon which nearly all 
 modern engines work, may be summed up as follows : 
 
 , , . f Outstroke draws in air and gas. 
 
 First revolution... j Instro k e compresses charge. 
 
 , , . ( Outstroke caused by the explosion. 
 Second revolution | Instroke discharges burnt products. 
 
10 ARRANGEMENT OF ENGINE-ROOM. 
 
 In the description of various designs of gas engines, we 
 shall afterwards see that the mechanism for effecting this 
 cycle of operations has undergone many changes. For the 
 present we may turn our attention to the arrangement of 
 the engine-room best suited for the working of a gas engine, 
 and to those fittings external to the engine itself. When 
 possible, the room in which a large gas engine is placed 
 should be quite separated from any building in which 
 vibration is likely to cause annoyance. If, however, it is 
 not possible to provide a separate engine-house, two pre- 
 cautions may be taken to prevent unpleasant results. There 
 should be no communication between the main building and 
 the engine-room, excepting by means of well-fitting doors, 
 which should be made to close automatically. This caution 
 is not necessary in the case of small engines, but large 
 engines of from 40 to 60 I.H.P., running at a speed of from 
 150 to 200 revolutions per minute, cause a rapid pulsation of 
 air owing to the displacement of the large trunk piston ; 
 this pulsation may be transmitted through the passages and 
 corridor to almost every room in the building, causing even 
 the windows to vibrate to every revolution of the engine. 
 For the same reason the engine-room should not be ventilated 
 by connection with the air shafts of the building. Artificial 
 ventilation is absolutely necessary, as the engine-room 
 otherwise becomes unbearable. A small air propeller driven 
 from the engine will answer this purpose admirably. 
 
 Another fruitful source of annoyance is the vibration of 
 the engine transmitted through its foundations to the walls 
 of the building. In some instances it has been found that 
 an increase in the speed has checked the vibration of the 
 foundations and walls. It is quite probable that every 
 building has a period of vibration peculiar to itself, and if, 
 therefore, a machine is running so as to have the same period 
 of vibration, the results may be very noticeable. Increase 
 the speed of the machine, alter its period of vibration until 
 it no longer synchronises with that of the building, and the 
 latter will not respond ; thus the effect will be minimised to 
 an inappreciable extent. 
 
ENGINE-ROOM. 11 
 
 Another method by which vibration of the engine itself 
 may be prevented from transmission to the building is to 
 bolt the engine bed down upon a somewhat soft and springy 
 material. Let some timber baulks be placed longitudinally 
 with the engine bed. Over these put a layer of iron plate, 
 say in. thick, then put over the plate a layer of cocoanut 
 matting. It is well to have these mats specially made. 
 They should be manufactured after the manner of a thick 
 pile cocoanut door mat ; two should be made of the required 
 area ; the pile faces should be put together, and the two 
 bound at the edges into one thick mat. By putting them 
 together in this way a thickness of 4 in. or 5 in. may be 
 obtained of springy texture, and presenting an exterior of 
 comparatively tough material. Another plate over the mat 
 serves to distribute the weight of the engine over the entire 
 area, and then the whole may be loosely held together by 
 means of the foundation bolts. If this method is adopted, 
 the engine will no doubt rock considerably, and extra large 
 bolts may be necessary for holding down. The rock of the 
 engine will cause considerable oscillation of the belt, and 
 for this reason a rather short drive is to be preferred. 
 These objections are, however, entirely outweighed by the 
 satisfactory prevention of the transmission of vibration 
 throughout the building. 
 
 The next point to be considered in the arrangement of 
 the engine-room is the position and size of the gas meter 
 required. All engines should be provided with a separate 
 gas meter, so that the consumption of the engine may be 
 checked independently. The meter should be placed outside 
 the engine-room in an atmosphere of normal temperature, 
 for it must not be forgotten that an increase of temperature 
 results in an increase in the volume of gas. If, therefore, 
 the meter is kept at a high temperature, the cubic feet 
 registered by it will be greater for the same weight of gas 
 than if kept at normal atmospheric temperature. No 
 injustice is done to the producers of the gas by taking this 
 precaution, because the temperature of the supply is itself 
 normal. 
 
12 ARRANGEMENT OF ENGINE-ROOM. 
 
 The size of a gas meter is usually gauged by the maximum 
 number of lights for which it is designed. Thus, small 
 houses are fitted with a 5-light meter, others a 10-light meter, 
 according to the requirements, and this indication of the 
 size is usually printed upon the index of the meter. A 
 useful rule for determining the size of a dry meter required 
 for a gas engine, the brake horse power of which is known, 
 is as follows : 3 4 x brake horse power + 5 = size of meter 
 required (measured as above stated). 
 
 Thus, suppose an engine of 30 brake horse power is being 
 erected, the meter required will be 3'4 x 30 + 5 = 107-light 
 meter. In this instance a 100-light meter would suffice. 
 
 Gas and oil engine makers are now quoting in their 
 catalogues the brake or effective horse power, instead of 
 indicated and nominal horse powers. The relation between 
 these will be dealt with upon following pages, devoted 
 to gas-engine testing ; but it may here be said that this 
 innovation is a great improvement, inasmuch as the pur- 
 chaser knows exactly the capacity of an engine for doing 
 useful work. 
 
 Having determined the size of meter suitable for measuring 
 the gas supplied to an engine, we must next consider the 
 size of pipe from the meter to the engine. The following 
 rule gives the correct size of pipe for a given size of meter : 
 
 Meter size (measured in lights) x O'OOS + 075 = bore of 
 pipe in inches. In the example previously worked it was 
 calculated that a 100-light meter is suitable for a 30 brake 
 horse power engine. The size of gas pipe for this meter 
 will be equal to 100 x 0008 + 0'75 = 0'8 + 075 = T55 in. 
 Putting in a size of pipe to the nearest in., we should, 
 in this instance, couple the gas supply to the engine by 
 means of a 1^ in. diameter pipe. 
 
 Another rule may be used for obtaining the diameter of 
 gas pipe when only the brake horse power is known. If 
 D = inside diameter of pipe in inches, then we have 
 
 D = 027 x brake horse power + 075. 
 
ARRANGEMENT OF ENGINE-ROOM. 13 
 
 Working out the diameter suitable for a 30 brake horse 
 power engine, we have 
 
 D = G'027 x 30 + 075 
 = 0'81 + 75 
 = 1*56 in. ; 
 
 and putting in a size to the nearest fa i n - we obtain 1| in. 
 diameter, as before calculated from the other data. 
 
 Between the gas meter and the engine there should be 
 fitted a flexible bag through which the gas passes. The bag 
 may be made by clamping two sheets of indiarubber about 
 $ in. thick to each side of a cast-iron ring, the diameter of 
 which may be from 1 ft. 6 in. to 2 ft. 6 in., or more, according 
 to circumstances. By this arrangement rapid fluctuation of 
 pressure in the gas mains is prevented, thereby maintaining 
 a steady light given by the burners in proximity to the 
 engine. To contribute to this end, two bags are sometimes 
 fitted in the following way : The bags, each composed of a 
 cast-iron dish to form one side, and sheet rubber the other, 
 are connected together, so that the gas enters and fills one, 
 then passes on to the next. Between the entrance to the 
 first gas bag and its pipe flange is fitted a cast-iron box con- 
 taining a plain slide valve.' This valve is moved by a lever 
 attached to the centre of the rubber side of the bag in such 
 a way as to close the entrance to the bag when the latter is 
 full of gas. A similar valve is fitted between the two con- 
 secutive bags, and the whole forms a very efficient means of 
 governing the pressure in the mains. In putting up the gas 
 piping and bags, the rubber sides should be placed next the 
 wall, and the bags should be as near the engine as possible. 
 It is advisable to leave sufficient room between the wall and 
 rubber to enable the attendant at any time to place his 
 hand upon the rubber to flatten it. It is often necessary to 
 do this in starting an engine, to get rid of an accumulation 
 of poor gas in the bags when the machinery has been 
 standing several days. It is also a useful precaution to have 
 a small gas cock fitted to the mains near the meter, to which 
 a U glass tube may be attached by an indiarubber pipe 
 
14 ARRANGEMENT OF ENGINE-ROOM. 
 
 By partially filling the tube with water and opening the 
 cock the fluctuation of pressure in the mains may be 
 observed, and the efficiency of the gas bags tested when the 
 engine is running. The rubber side of the bags must be 
 kept perfectly clean and free from oil, or its corrosive effect 
 will soon be evident, and, further, the rubber must not be 
 subject to excessive temperatures. 
 
 The diameter of the engine exhaust pipe may be found 
 from the following rule for any engine larger than 5 brake 
 horse power : If D e diameter of exhaust pipe, 
 
 then we have D e = 528 horse power ' 57 
 Taking the brake horse power at 30, as in the previous 
 example, and applying the rule given to find the exhaust 
 diameter, we have 
 
 G'528 x 30^ = D K - 
 log 30 = 1'477 ;* 
 and 1-477 x 0'57 - 0'838 (nearly) = the log of (300-57). 
 
 Finding the number corresponding to 0"838, we obtain 
 6'89 and 0'528 x 6'89 = 3"64 diameter of exhaust pipe in 
 inches, 
 
 We should here adopt a 3 in. or 3| in. diameter pipe. 
 Engines of from 1 to 5 brake horse power may have the 
 exhaust pipes from 1 in. diameter to If in. 
 
 It is very important that the exhaust pipe should be free 
 from bends, but where necessary they should be of large 
 radius, say 6 in. Owing to the high temperature of the 
 exhaust gases, the pipe should be isolated from inflammable 
 material. When the mouth of the pipe discharges into the 
 open air it should point downwards, to prevent the 
 collection of rain water. When it is desirable to deaden 
 the noise of the exhaust, it may be discharged into a cast- 
 iron chamber partially filled with flint pebbles, or into the 
 bottom of a trench filled with pebbles. Whatever form of 
 
 * All calculations will be worked out upon a 10 in. slide rule, and the results 
 will be accurate to about ona-half per cent. 
 
ARRANGEMENT OF ENGINE-ROOM. 15 
 
 exhaust chamber is made use of, it should always be placed 
 in an accessible place, and provision made for efficiently 
 draining water which may collect. 
 
 The next subject for consideration in the arrangement of 
 gas-engine plant is the delivery of water to the jacket 
 surrounding the working cylinder of the engine. The 
 jacket water circulates round the cylinder in an annular 
 space, formed by the outer wall of the explosion cylinder 
 and the inner wall of a concentric casing. This casing is 
 cast with the cylinder, and the space allowed for the water 
 is from f in. to 2 in. A flange is provided at the under side 
 of the jacket for the inlet of the water, and another at the 
 top and front end for the outlet. For small engines the 
 inlet and outlet pipes may be the same size, but when 
 circulation tanks are used, and the engine is over 20 horse 
 power, it is better to have the outlet pipe from in. to f in. 
 larger in diameter. 
 
 The function of the jacket water is to carry away the excess 
 of heat due to combustion in the cylinder, which, under the 
 present conditions, cannot be converted into work. It is 
 found in practice that the quantity of heat accounted for 
 by the jacket water amounts to from 30 to 50 per cent of the 
 total heat derived from the combustion of the gases in the 
 cylinder, and that under these conditions lubrication is 
 possible. In estimating, therefore, the quantity of jacket 
 water required, it is necessary to bear in mind that it must 
 absorb not less than 30 per cent of the heat of combustion, 
 and that its maximum temperature should not exceed 
 150 deg. Fah. for continuous running. When the cooling 
 water runs to waste, the maximum temperature may be 
 somewhat higher, but when circulating tanks are used it is 
 better not to exceed 150 deg. Fah. If we assume an 
 average gas consumption of 22 cubic feet per indicated 
 horse power per hour, we may estimate the heat units 
 generated in the cylinder for each H.P. to be 22 x 620 
 = 13,640 units per hour, because 1 cubic foot of coal gas 
 develops on an average 620 British thermal units. At our 
 
16 ARRANGEMENT OF ENGINE-ROOM. 
 
 lowest estimate, 30 per cent of this must be taken up by the 
 jacket water ; this amounts to 4,092 units. We have said 
 that the jacket water should not exceed 150 deg. Its inlet 
 temperature may be taken at 60 deg. Fah., thus allowing a 
 rise of temperature of 150 60 = 90 deg. Fah. per pound 
 
 of water, the pounds per I.H.P. required = -*j^L= 45'5 
 
 (nearly) per hour. This is equivalent to 4| gallons of 
 water per I.H.P. per hour, and it approaches the minimum 
 quantity allowable for continuous working. Thus we find 
 that if the water supply is continuous, and passes through 
 the jacket, afterwards running to waste, at the rate of 
 4 gallons per I.H.P. per hour, efficient lubrication may be 
 maintained. With such an arrangement, the size of the 
 jacket inlet and outlet pipes will depend entirely upon the 
 head of water in the source of supply, but as the arrange- 
 ment described is only suitable for experimental engines 
 and other special cases, the jacket pipes are put in from 1 in. 
 to 2 in. diameter up to engines of 20 I.H.P. Above 20 H.P. 
 the diameters range from 2 in. to 3 in. for the inlet, and 
 from 2^ in. to 3| in. for the outlet. 
 
 The most usual method of supplying the jacket is to place 
 it in communication with a set of tanks, through which 
 there is a free circulation of water. The diagram (fig. 2) 
 shows an arrangement of circulation tanks. The tank A is 
 coupled to the jacket inlet, and receives the feed necessary 
 to make up the loss from evaporation. The outlet pipe is 
 taken to the normal water level of the tanks. This pipe 
 should have a minimum rise of 2 in. per foot where it leads 
 into the tank, in order to secure an easy circulation. The 
 tank B communicates with A through a pipe starting from 
 the bottom of B, passing out at the normal water level 
 across to A. A drain pipe, slightly above the normal water 
 level, serves to take away any excess of feed water. For 
 clearness in the drawing the engine is shown between the 
 tanks, but it is usual to have the tanks side by side, outside 
 the engine-room, in as cool a place as possible. They should, 
 
ARRANGEMENT OF ENGINE-ROOM. 
 
 17 
 
 however, not be placed in the open air, as, in the event of 
 severe frost, much inconvenience may be occasioned. In 
 this connection it may be well to mention, that when an 
 engine is subject to frost, and is not running continuously, 
 it is a wise precaution to have two valves fitted to the inlet 
 jacket pipe, the one near the tank A to shut off the supply 
 from A ; the other at the bend of the pipe as it enters the 
 jacket, and communicating with the atmosphere. Thus the 
 jacket only may be drained, and be secured from bursting if 
 the frost gets to the engine. The circulation tanks should 
 have a capacity of from 20 gallons to 30 gallons per I.H.P., 
 
 <?= 
 
 r 
 
 -V |_j 
 
 
 
 
 
 
 
 n 
 
 -i 
 
 ;i' 
 
 WATER 
 LCVEL 
 
 \ 
 
 >| 
 
 
 
 
 ^ 
 
 
 
 L 
 
 i 
 
 j: 
 
 
 
 
 
 
 
 i 
 
 
 
 
 1 
 
 :!;':| 
 
 
 
 I! 
 
 
 
 
 e 
 
 
 
 
 
 1 * 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 B 
 
 
 
 
 
 
 
 A 
 
 
 
 
 i 
 
 i 
 
 
 
 >; 
 
 L* 
 
 : 1 7 : 
 
 
 
 
 
 
 
 |, 
 
 
 
 ^ 
 
 W 
 
 
 
 i 
 
 
 ! 
 
 
 
 
 
 ->-- 
 
 >f 
 
 fT 
 
 - 
 
 
 ^,' ; 
 
 
 !i 
 
 
 Jjjj. 
 
 
 
 UJ 
 
 J 
 
 s 
 
 
 
 | 
 
 
 
 PRACTICAL ENGINEER.. 
 
 FIG. 2. Arrangement of water tanks. 
 
 so as to allow a sufficient time for cooling. The form of 
 tanks is usually cylindrical, though need not be. It is, 
 however, necessary that the effective height should be from 
 two and a half to three times the diameter or breadth, 
 otherwise a good circulation will not be induced. 
 
 The circulation is induced by the expansion of the column 
 of water in the pipe e leading from the jacket to the tank. 
 When the engine is started the jacket temperature may 
 reach about 160 deg. Fah., and the volume of the water will 
 increase about *Uth. Thus, a column of water 8ft. high 
 will increase when heated nearly to 8ft. 3| in. There will 
 SG 
 
18 ARRANGEMENT OF ENGINE-ROOM. 
 
 then be an equivalent head of water in the cold tank A of 
 3 in., which, neglecting friction in the pipes, will cause the 
 cold water to enter the jacket at the rate of 4 '4 ft. per 
 second. Thus, immediately upon starting the engine, a very 
 brisk circulation of the water is induced. After some time 
 has elapsed, the water in A rises nearly to the same tem- 
 perature as at the bottom of B, so that the equivalent head, 
 which afterwards maintains the circulation, may be calcu- 
 lated from the difference in temperature between a stratum 
 of water at the bottom and that at the top of the tank B. 
 
 We have now dealt with the chief external fittings of a gas 
 engine worked by town's gas. The economy to be gained 
 by putting down plant for the production of Dowson gas 
 depends upon many conditions which may or may not exist 
 in any particular instance. Undoubtedly, for the constant 
 running of large engines, producer gas is by far the most 
 economical. This important subject will be more fully 
 discussed later. 
 
 Before concluding these remarks upon the arrangement of 
 an engine-room the following points may be noted : When 
 possible, arrange for the tight side of the belt to run from 
 the bottom of the driving-wheel. This is not always con- 
 venient when driving a dynamo, for the dynamo pulley 
 must run in the same direction as the hands of a watch 
 when viewed from a position facing the pulley end of the 
 dynamo. When cramped for room, it is necessary that the 
 gas-engine cylinder should lie between the dynamo and the 
 engine crank shaft ; thus, with an engine receiving its 
 impulse on the top throw of the crank, the tight side will be 
 at the top. With a fair width of belt no serious amount of 
 slip will take place in transmitting, say, 40 horse power, 
 even though the driving pulley be 7 ft. diameter, the driven 
 pulley 18 in., and the shaft centres only 15 ft. apart. Gas 
 engines have in some instances been coupled directly to the 
 dynamo shaft, thus effecting a great saving of space. At 
 Sunderland a gas-engine plant has recently been put down 
 driving the centrifugal pumps at the docks without the use 
 of either belts or gearing. 
 
UNIVERSITY 
 
 TYPES OF GAS ENGINES, 19 
 
 CHAPTER III. 
 
 TYPES OF GAS ENGINES. 
 ATKINSON'S "DIFFEKENTIAL" AND "CYCLE" ENGINES. 
 
 SINCE the Otto patent expired, the commercial value of 
 other engines has much depreciated, and in many cases 
 makers have abandoned their own patents in favour of 
 the Otto principle. In 1885, Mr. J. Atkinson, M.I.M.E., 
 patented a motor known as the Differential engine. This 
 engine was constructed upon different lines to any other 
 gas engine that has been made, and from a theoretical point 
 of view is the most perfect engine yet invented. Regarded 
 however as a mechanical contrivance, Mr. Atkinson him- 
 self says it is not good, though it may be considered an 
 ingenious combination of link motion. Although not now 
 manufactured, no work upon the gas engine would be 
 complete without a description of this machine. Figs. 3, 
 4, 5, and 6 show the engine diagrammatically in four 
 positions of the crank pin. The cylinder is open at both 
 ends, and water jacketed on the top only. Two pistons 
 work in the cylinder, and are each connected with the crank 
 pin P by means of a freely jointed connecting rod c, 
 attached to the bent lever L. This lever oscillates upon the 
 fixed pin D, which is attached to the framing of the engine. 
 The upper end of the lever L drives the crank pin by means 
 of the short connecting rod R. Fig. 3 shows the engine 
 with the charge compressed between the two pistons. The 
 left-hand piston is just uncovering the entrance to the 
 ignition tube I. It will be observed that as the crank pin 
 follows the direction indicated by the arrow, the rod R x 
 merely revolves about the upper end of the lever L^ until 
 P arrives at position shown in fig. 5. The left-hand 
 piston has therefore remained almost stationary, whilst 
 the right-hand piston has moved very rapidly to the right 
 
20 
 
 TYPES OF GAS ENGINES. 
 
 ATKINSON'S DIFFERENTIAL GAS ENGINR. 
 FIG. 3. Volume before explosion. FIG. 4. Volume when expanded. 
 
 ATKINSON'S DIFFERENTIAL GAS ENGINE. 
 
 FIG. 5. Showing total expulsion of FIG. G Volume of charge before 
 burnt gases. compiession. 
 
TYPES OF GAS ENGINES. 
 
 21 
 
 ATKINSON'S CYCLE GAS ENOINF. 
 
 FIG. 7. Showing vc^ume before 
 explosion, 
 
 Fio. 8. Showing volume when 
 expanded. 
 
 ATKINSON'S CYCLE GAS ENGINE. 
 
 FIG. 9. Showing total expulsion of FIG. 10. Showing volume of charge 
 burnt products. . before compression. 
 
22 TYPES OF GAS ENGINES. 
 
 and the charge between the pistons is expanded, thus 
 giving an impulse to the crank pin in the direction of its 
 motion. The effect of the pressure upon the left-hand 
 piston during the time the crank pin moves between the 
 positions shown in figs. 3 and 4 merely produces a down- 
 ward thrust upon the crank-shaft journals, and only a small 
 resolved force acts against the rotation of the crank pin. 
 On the other hand, the rod E, attached to the right-hand 
 lever L, exerts an almost tangential effort on the crank pin 
 during this quadrant. The exhaust port is now uncovered 
 by the piston c, and remains so during the next quarter 
 revolution. In the mean time the left-hand piston C L moves 
 rapidly to the right, and expels the whole of the burnt 
 gases. The piston c now moves slowly to the left, covering 
 the exhaust port, while C L moves rapidly to the left, 
 drawing in the fresh charge. Fig. 6 shows the position of 
 the pistons when the charge is complete, and it should be 
 observed that the volume enclosed by the pistons is here 
 less than in fig. 4, j ust when the exhaust opens. It there- 
 fore follows that the charge is expanded beyond its original 
 volume at atmospheric pressure, thus carrying out the prin- 
 ciple of maximum expansion laid down by Beau de Rochas. 
 
 The four characteristic features of this engine are an 
 ignition per revolution, the expulsion of the burnt gases, 
 expansion to greater than the original volume, also perfectly 
 automatic valves and means of timing the ignition. Not- 
 withstanding the mechanical defects, this engine was the 
 most economical of its time, for it consumed only 26 cubic 
 feet of gas per brake horse power in small engines, and this 
 was reduced to 24 cubic feet in large engines. This engine 
 was first exhibited at the Inventions Exhibition held in 
 London in 1885, and was there awarded a gold medal. 
 
 Its mechanical defects were so obvious that its manufacture 
 was abandoned in favour of another type, patented by Mr. 
 Atkinson, and known as the Cycle engine. Here the 
 patentee sought to combine the advantages of the Differen- 
 tial engine with more durable mechanism, and he was able, 
 
TYPES OF GAS ENGINES. 
 
 23 
 
 by the arrangement shown in figs. 7 to 10, to reduce the 
 gas consumption to 22 cubic feet per brake horse power per 
 hour. Fig. 7 shows the position of the piston when ignition 
 takes place. The fixed centre F being rather below the 
 horizontal centre line of the cylinder causes the rod R to 
 rise through a greater angle above than below F. By this 
 means the explosion stroke was ll'l in. Fig. 8 gives the 
 position of the links at the end of the explosion stroke. 
 
 FIG. 11. ATKINSON'S CY8LK GAS ENGINE. 
 
 During the exhaust stroke it will be noticed (fig. 9) that 
 the joints at C are inclined instead of horizontal, as in fig. 
 7, thus increasing the length of the backward stroke to 
 12 '4 in., and bringing the piston close to the back of the 
 cylinder. The whole of the burnt gases being driven out, 
 the piston makes another forward stroke, but, as before 
 mentioned, owing to the smaller angle made by the link R 
 below the horizontal, this stroke only measures 6'3 in. Fig. 
 10 shows the position of the crank pin when the charge is 
 
24 TYPES OF GAS ENGINES. 
 
 completed. Fig. 11 gives a more complete view of the Cycle 
 engine, many of which are at the present time working. 
 
 The rod a operates the gas valve by means of a cam on the 
 crank shaft. A similar rod in front of a operates the exhaust 
 valve. The ignition tube in this engine is screwed direct 
 into the back of the cylinder, the time of ignition being 
 determined by the length of the tube. The hot ignition tube 
 communicates with the combustion chamber by means of the 
 small hole, through which passage the fresh charge is com- 
 pressed into the tube. After the explosion, expansion, and 
 exhaust have occurred, the tube remains full of products of 
 combustion, and it is evident that, before a fresh charge can 
 be ignited, these products must be so far driven to the end 
 of the ignition tube as to allow the fresh mixture to come in 
 contact with the red-hot portion of the tube. Hence the 
 length of an ignition tube plays an important part in 
 determining the amount of compression necessary, and 
 consequently the time of an explosion. In the Cycle engine 
 the firing of the charge is regular, and little trouble is 
 experienced with it. This is probably, in a great measure, 
 due to the pure charge put into the cylinder. There is 
 much diversity of opinion upon the value of a timing valve. 
 
 To show the result of the application by Mr. Atkinson of 
 the principles laid down by Beau de Rochas, the following 
 figures may be quoted. A 6 horse power Cycle engine was 
 put down by the Hampton Wick Local Board to drive a 
 single-acting air compressor in connection with the Shone 
 pneumatic sewage system. This was tested against a 
 duplicate set of pumps driven by a Crossley Otto engine 
 under precisely similar conditions. It was found that the 
 Cycle engine consumed 18'4 per cent less gas than the 
 Crossley engine. Notwithstanding the high efficiency of 
 the Cycle engine with regard to gas consumption, it has 
 been found to be costly in up-keep, requiring more lubrica- 
 tion and more frequent repairing, consequently, now that 
 the Otto patents have become public property, the Cycle 
 engine is not manufactured. It is nevertheless worthy of 
 
TYPES OF GAS ENGINES. 25 
 
 careful study as an example of a very successful attempt to 
 put into practice the true theory of an economical gas engine. 
 It has been stated previously that the majority of gas 
 engines now being manufactured work on the Otto cycle 
 but owing to the irregularity of the impulse given to the 
 piston it is certainly desirable to aim at a more equal 
 distribution of the turning effort exerted upon the crank 
 pin. This object may be to some extent attained by 
 building an engine with two Otto cycle cylinders instead of 
 one, and a still greater regularity in the impulses has been 
 attained by the construction of double-acting cylinders, 
 made possible by improvements in details relating to 
 lubrication and glands. 
 
 COMPARATIVE CYCLES OF GAS ENGINES. 
 
 The following diagram, fig. 12, shows the steadiness of 
 running with the various cycles now used. The shaded 
 squares represent indicator diagrams taken from the back 
 or front of the piston when an explosion takes place. It is 
 seen that in the Otto cycle single-cylinder engine, in six 
 revolutions, only three impulses are given to the piston. 
 The Griffin engine, about which more will be written, has a 
 double-acting cylinder. The series of operations on one side 
 of the piston only is as follows : 
 
 1. Out/stroke : Explosion and expansion. 
 
 2. Instroke : Products of combustion expelled. 
 
 3. Scavenger stroke outstroke : Draws in pure air only. 
 
 4. Scavenger stroke instroke : Clears out the cylinder. 
 
 5. Outstroke : Draws in gas and air. 
 
 6. Instroke : Compresses ready for explosion. 
 
 It is therefore evident that the back of the piston receives 
 one impulse every three revolutions. Similarly, the front 
 of the piston receives one impulse every three revolutions ; 
 consequently during the six revolutions two impulses are 
 received by one side of the piston, and two by the other, 
 making in all four. The object of the third outstroke for 
 drawing into the cylinder a charge of pure air will be fully 
 
26 
 
 TYPES OF GAS ENGINES. 
 
 02 
 O 
 
 02 
 
 I 
 
 Q 
 
 - 
 
 s 
 
 o 
 O 
 
 I 
 
TYPES OF GAS ENGINES. 27 
 
 treated later ; but here it may be stated that if all the 
 products of combustion are entirely swept out of the 
 cylinder by this means, it is generally believed that the 
 efficiency of the next charge of gas is increased. This is no 
 doubt true when Dowson gas is used, but experiments 
 carried out by the author upon mixtures of coal gas and 
 air point to the conclusion that the deleterious effect of 
 burnt products is much overrated, excepting when the gas 
 is more than *l\ per cent of the air by volume. 
 
 The Otto cycle, when applied to an engine having two 
 cylinders, gives an impulse every revolution of the fly- 
 wheels in the order shown upon the diagram. The Griffin 
 type, having but one double-acting cylinder, and working 
 without a scavenger stroke, gives an average of one 
 impulse every revolution, though the sequence of the 
 working stroke is perhaps somewhat less conducive to steady 
 running. This is, however, improved by the use of two 
 cylinders, each with its scavenging strokes, giving eight 
 impulses in six revolutions. And, lastly, the Griffin engine, 
 with two double-acting cylinders and no scavenger stroke, 
 gives twelve impulses during six revolutions. Regarding 
 i only the number and sequence of motor strokes per 
 revolution, the above-mentioned cases afford an example of 
 what may be done by the various combinations of cylinders ; 
 but it will be understood that there are other engines, 
 which, whilst fulfilling one set of conditions indicated by 
 the diagram, are distinct in design, and possess advantages 
 peculiar to themselves. 
 
 THE CROSSLEY ENGINE. 
 
 The modern Otto, as built by Messrs. Crossley Brothers 
 Limited, Manchester, is very different from that illustrated 
 in fig. 1. The old method of flame ignition is now entirely 
 superseded by allowing the explosive mixture in the cylinder 
 to enter a red-hot tube attached to the cylinder. The slide 
 valve controlling the admission of the air and gas has been 
 replaced by two simple mushroom valves, somewhat similar 
 
28 TYPES OF GAS ENGINES. 
 
 to the exhaust valve on the original Otto engine. There are 
 many objections to the employment of a slide valve on a 
 gas engine, and it is probable that a change would have 
 been made much earlier had it not been generally believed 
 that the slide valve effected a better distribution of the gas 
 in the cylinder than was otherwise possible. In describing 
 the original Otto, it was pointed out that the spacing of the 
 port edges, admitting the air and gas, was such that air 
 entered the cylinder first, then air and gas, and lastly pure 
 gas only. It was believed that when the full charge had 
 entered the cylinder it existed in three distinct layers the 
 purer gas assisted the ignition, the more dilute mixture 
 spread the flame, and, lastly, the pure air deadened to some 
 extent the percussive effect upon the piston. A deceptive 
 experiment in support of this theory may be made by 
 fitting a simple piston in a glass tube. Let a cork be fitted 
 tightly to the end of the tube beneath the piston, and into 
 a hole in the cork pass a cigarette. Push the piston to the 
 bottom of the tube, and then withdraw it sloioly to about 
 half the length of the tube. Upon lighting the cigarette, 
 and completing the slow movement of the piston to the end 
 of the tube, smoke will follow in a distinct stratum, and 
 will remain unmixed with the air for some time. This slow 
 movement, however, does not exist in a gas engine, and, 
 moreover, the gas more readily follows the piston than in 
 our experiment, and consequently a practically homogeneous 
 mixture is contained in the cylinder before ignition takes 
 place. That there is no stratum of pure air next the piston 
 of a gas engine has been proved by igniting the charge at 
 the forward end of the cylinder, instead of at the back of 
 the clearance space. 
 
 As the theory of stratification was proved to be a fallacy, 
 there appeared no reason why the mechanical difficulties of 
 the slide valve should not be overcome by the use of separate 
 gas and air valves. By the use of mushroom valves leakage 
 is entirely prevented, rendering it possible to compress up 
 to about 80 Ib. or more per square inch before ignition, 
 
TYPES OF GAS ENGINES. 
 
 29 
 
30 TYPES OF GAS ENGINES. 
 
 whereas in the old slide engines a pressure of only 40 Ib. was 
 obtained. These valves are invariably held upon their seats 
 by a spiral spring, and lifted by levers actuated by cams on 
 the lay shaft. 
 
 According to their latest designs, Messrs. Crossley are 
 building engines with one cylinder indicating up to 100 
 horse power, fitted with one or two flywheels, according to 
 the requirements. Engines indicating 250 horse power are 
 built with two cylinders placed opposite each other, their 
 connecting rods working on the same crank pin. One 
 specially heavy flywheel, supported by an extra bearing, is 
 fitted to these engines for electric lighting purposes. Figs. 
 13 and 14 show the general external appearance of these 
 engines. Small vertical engines, indicating up to 8 horse 
 power, are built for various purposes. Those fitted with 
 hoisting drums are geared to the latter by friction wheels, 
 and an engine indicating 6 horse power may be calculated 
 to lift 1,110 Ib. at the rate of 60ft. per minute, whereas a 2 
 horse power engine will lift 280 Ib. at 80 ft. per minute. 
 These figures depend to some extent upon the gearing used, 
 and it is probable that under favourable circumstances both 
 the weights and speeds might be increased, for it will be 
 found from the above-quoted figures that the mechanical 
 efficiency is from 33 per cent to 53 per cent, a somewhat low 
 estimated efficiency for a direct-driving motor. 
 
 The ignition tube now used on the Crossley engines is a 
 modified form of fig. 15. The central tube T is maintained 
 at a bright red heat by means of a bunsen flame. A 
 cylindrical cover R of non-conducting material serves to 
 concentrate the heat. The action when working is as follows : 
 During the compression stroke the small valve E closes upon 
 the upper orifice by the upward movement of the lever L. 
 The compressed mixture enters the cavity G. The moment 
 ignition is required, the valve E descends by the pressure 
 exerted by the spring S, when the lever L drops. The 
 compressed gas then rushing into the tube T is ignited, and 
 the flame strikes back into the cylinder. The valve E now 
 
TYPES OF GAS ENGINES. 
 
32 
 
 TYPES OF GAS ENGINES. 
 
 rises, and the burnt products in the tube T escape through 
 the lower orifice of the valve E. In this design a rather 
 long ignition tube is required, to ensure the fresh mixture 
 reaching the hottest part of the tube T, by forcing to its 
 upper closed end any of the burnt products remaining in the 
 tube. 
 
 FIG. 15. Otto Tube Igniter. 
 
 In 1893 Messrs. Crossley patented a self -starting arrange- 
 ment for their gas engines, which consists essentially of a 
 hand pump for forcing a charge into the cylinder when the 
 crank is in a proper position for receiving an impulse. 
 The exhaust cam on the lay shaft is so constructed as to 
 prevent excessive compression when starting. When a 
 charge has been pumped by hand into the cylinder, a timing 
 valve is opened, and the charge fired by contact with the 
 
TYPES OF GAS ENGINES. 
 
 33 
 
 ignition tube. When the engine has acquired a consider- 
 able momentum, the exhaust valve is made to close early, 
 and so give the required compression for continuous 
 running. 
 
 The latest improvement in the Crossley engines is said to 
 be due to the introduction of a scavenging arrangement, by 
 which the exhaust gases usually remaining in the clearance 
 space are drawn away at the end of the stroke. The 
 innovation is fully described in a paper read by Mr. J. 
 Atkinson, M.I.Mech.E., before the Manchester Association 
 of Engineers, from which the following description and 
 illustrations have been derived. 
 
 Referring to fig. 16, which shows a diagrammatic view of 
 the piston, cylinder, and path of the crank, we may take the 
 
 PRACTICAL ENGINEER 
 
 FIG. 16. Diagram of Atkinson and Crossley 's scavenging arrangement. 
 
 point A as the position of the crank pin when the exhaust 
 valve opens for the discharge of the products of combustion. 
 In the ordinary Otto engine this valve would close at the 
 point C, and at the same time the air admission valve 
 would open. But in the engines now made with Messrs. 
 Atkinson and Crossley's scavenging arrangement, the ex- 
 haust valve remains open until the crank pin is at B, whilst 
 the air admission valve opens at D. Thus, it will be noticed, 
 both the air and exhaust valves are open during one-quarter 
 of a revolution. When the exhaust valve first opens at A 
 there is a pressure in the cylinder of about 35 Ib. This is 
 discharged into an abnormally long exhaust pipe, about 
 65 ft., through which it is driven by the returning piston. 
 When the velocity of the piston approaches zero at the end 
 of the instroke, this long column of exhaust gas, by virtue 
 
34 TYPES OF GAS ENGINES. 
 
 of its inertia, causes a slight vacuum to be formed in the 
 combustion chamber. At this time the air valve is opened 
 (see position marked D), and a draught of fresh air is drawn 
 through the combustion chamber into the exhaust pipe, 
 sweeping out all the remaining burnt products and replacing 
 them with fresh air. The exhaust valve closes at B, and 
 the gas for the next charge may then enter through the 
 gas valve. 
 
 The variation of pressure in the cylinder is clearly shown 
 by the accompanying indicator diagram, fig. 17, taken with 
 a in. spring, in order to show only the effect of the long 
 exhaust pipe. The top line of the diagram has no signi- 
 ficance, for the indicator piston is merely pressing upon a 
 stop provided in order to prevent damage to the weak 
 spring during the compression and explosion strokes. The 
 
 FIG. 17. Diagram taken with Jth spring. 
 
 pencil of the indicator leaves this horizontal line on the 
 exhaust stroke of the engine, and the effect of the first 
 puff of the exhaust is to set in motion a long column of gas, 
 the inertia of which causes the slight vacuum shown at V. 
 After this point the engine piston, travelling at its maximum 
 velocity, causes a slight increase in the pressure, and the 
 consequent acceleration of the gases in the exhaust pipe, 
 which finally produces about 3| Ib. vacuum, when the air 
 valve admits fresh air. The suction stroke is completed at 
 from 1 Ib. to 2 Ib. below the atmospheric line. 
 
 In order to assist the draught of air through the cylinder, 
 the piston is specially shaped as shown in figs. 18 and 19. 
 This drawing, together with fig. 20, also illustrates the usual 
 style of mushroom valves used in nearly all modern gas 
 engines. It will be noticed that no stuffing boxes are used, 
 
TYPES OF GAS ENGINES. 
 
 35 
 
 PRACTICAL ENGINEER 
 
 Sectional plan. 
 
 FIGS. 18 and 19. Arrangement of air passages in Atkinson and Crossley's 
 scavenging arrangement. 
 
36 
 
 TYPES OF GAS ENGINES. 
 
 but that specially long, well-fitted spindles, working in long 
 sleeves, are sufficient to prevent escapes. The valve spindles 
 are always on the face subjected to the least pressure. Mr. 
 Atkinson states that silencers and quietening chambers do 
 not prevent the action of the exhaust, provided they are 
 
 ' I J-' 1 f 
 
 j i n ', ' 
 p .' ^ 
 
 ULJ 
 
 TiTU 
 
 \\ N 
 
 V ^x 
 
 
 FIG. 20. Mushroom valves used in gas engines. 
 
 placed at the end of a 65ft. length of exhaust pipe of 
 uniform diameter. No sudden enlargements are permissible 
 within about 65ft. of the engine, although drain pockets 
 may be inserted for the collection of condensed moisture. 
 
 THE "STOCKPORT" GAS ENGINES. 
 
 The Stockport gas engine is manufactured by Messrs. J. 
 E. H. Andrew and Co. Limited, Reddish, and works upon 
 the Otto cycle. This firm commenced the manufacture of 
 the "Bisschop" engine a small motor not much used at 
 the present in 1878, and from that time have built over 
 7,000 motors of the Bisschop and Stockport design. The 
 latest Stockport is a horizontal engine, the general features 
 of which may be seen from the engraving (fig. 21). 
 
TYPES OF GAS ENGINES. 
 
 The following table, supplied by the makers, is of use in 
 the arrangement of details external to the engine itself : 
 
 Effective 
 horse 
 power. 
 
 Revs, 
 per 
 
 minute. 
 
 Size of 
 flywheels. 
 
 Standard 
 size of 
 pulley. 
 
 Overall 
 dimensions, 
 engine only. 
 
 Approxi- 
 mate 
 weight. 
 
 Vertical : 
 U 
 
 220 
 
 Diam. 
 One: 
 Ft. In. 
 3 1J 
 
 Width 
 One: 
 In. 
 
 4 
 
 Diam. 
 
 In. 
 10 
 
 Width 
 
 In. 
 
 6 
 
 Length. 
 
 Ft. In. 
 3 4 
 
 Breadth. 
 
 Ft. In. 
 
 3 
 
 Cwt. Qr. 
 13 2 
 
 5 
 
 200 
 
 4 
 
 5 
 
 18 
 
 7 
 
 3 6 
 
 3 3 
 
 22 2 
 
 Hori- 
 zontal : 
 
 1 
 
 240 
 
 Two: 
 2 6 
 
 Two: 
 3* 
 
 9 
 
 5 
 
 4 9 
 
 2 6 
 
 9 2 
 
 2 
 
 220 
 
 2 11 
 
 21 
 
 10 
 
 6 
 
 5 3 
 
 3 
 
 13 2 
 
 3 
 
 220 
 
 3 2 
 
 3 
 
 14 
 
 6 
 
 5 9 
 
 3 9 
 
 18 
 
 5 
 
 220 
 
 3 6 
 
 4 
 
 16 
 
 7 
 
 6 9 
 
 4 
 
 27 3 
 
 7 
 
 220 
 
 3 11 
 
 4 
 
 18 
 
 7 
 
 7 
 
 4 3 
 
 32 2 
 
 9 
 
 200 
 
 4 3 
 
 5 
 
 21 
 
 8 
 
 8 
 
 4 6 
 
 40 
 
 11 
 
 200 
 
 4 3 
 
 5 
 
 21 
 
 8 
 
 8 3 
 
 5 
 
 45 
 
 13 
 
 190 
 
 4 8 
 
 5i 
 
 23 
 
 9 
 
 8 6 
 
 5 3 
 
 54 2 
 
 ; 15 
 
 190 
 
 4 9 
 
 5J 
 
 23 
 
 10 
 
 8 9 
 
 5 6 
 
 58 2 
 
 17 
 
 185 
 
 5 2 
 
 61 
 
 27 
 
 12 
 
 8 10 
 
 5 9 
 
 66 
 
 19 
 
 185 
 
 5 3 
 
 61 
 
 27 
 
 12 
 
 9 
 
 6 
 
 70 
 
 22 
 
 185 
 
 5 3 
 
 6J 
 
 32 
 
 12 
 
 9 3 
 
 6 3 
 
 75 3 
 
 26 
 
 180 
 
 5 3 
 
 6 
 
 36 
 
 12 
 
 9 9 
 
 6 6 
 
 90 
 
 30 
 
 180 
 
 5 4 
 
 6 
 
 42 
 
 14 
 
 10 
 
 6 9 
 
 100 
 
 35 
 
 170 
 
 5 5 
 
 6 
 
 48 
 
 15 
 
 10 6 
 
 7 
 
 112 2 
 
 42 
 
 160 
 
 6 2 
 
 9 
 
 54 
 
 19 
 
 11 10 
 
 7 3 
 
 145 
 
 55 
 
 160 
 
 6 3 
 
 9 
 
 54 
 
 23 
 
 12 
 
 8 3 
 
 190 
 
 67 
 80 
 100 
 
 155 
 150 
 
 150 
 
 6 10 
 7 
 
 7 2 
 
 10 
 10 
 10 
 
 ) Special ( 
 V pulleys ( 
 I to suit. I 
 
 13 
 13 3 
 17 
 
 8 6 
 8 9 
 10 
 
 210 
 245 
 
 360 
 
38 
 
 TYPES OF GAS ENGINES. 
 
 The construction of the Stockport gas engine now built, 
 may be followed by an examination of the accompanying 
 
 FIG. 21. "Stockport" Gas Engine. 
 
 FIG. 22. "Stockport" Gas Engine Sectional Elevation. 
 
 drawings. Fig. 22 shows a section through the back of the 
 cylinder, with the gas and air valves. It will be noticed 
 
TYPES OF GAS ENGINES. 
 
 39 
 
 that the back of the cylinder is cast separately from the 
 remainder of the cylinder and engine framing ; and further 
 (see fig. 25) that the metal forming the back of the water 
 jacket is specially thin. By this arrangement the casting 
 and boring of the cylinder is much facilitated, and in the 
 event of frost getting to the jacket, when full of water, 
 
 FIG. 23. "Stockport" Gas Engine Section of Timing Valve Bracket 
 and Self -starter combined. 
 
 fracture due to the formation of ice will occur only at the 
 thin wall of the jacket. This is easily and cheaply replaced, 
 whereas if fracture occurred in the forward end of the 
 cylinder, a large and expensive casting would be ruined. 
 
 In fig. 22 the supply of gas is coupled to the mouth of 
 the pipe shown downwards. In this pipe is a wing throttle 
 valve actuated by a lever from the governor. The gas 
 
40 
 
 TYPES OF GAS ENGINES. 
 
 passes through the mushroom valve, when the latter is 
 lifted by a cam, and passes into a chamber beneath the air 
 valve. Here it mixes with the air, as both are drawn into 
 the cylinder through the upper mushroom valve. This 
 valve is lifted by a cam shown in fig. 24, and closes by its 
 own weight, assisted by the spiral spring upon the spindle. 
 When the mixture is compressed by the return stroke of 
 the piston, it is forced up to the face of the timing valve 
 shown in the end sectional view, also in detail in fig. 23. 
 At the end of the stroke the timing valve F is pushed from 
 
 FIG. 24.-"Stockport" Gas Engine Sectional End Elevation. 
 
 its seat by the lever D, and the gas entering the red-hot 
 tube G is exploded. The timing valve may be adjusted by 
 the set screw on the end of the lever D. If the explosion 
 occurs too early, the set screw should be slightly withdrawn, 
 if too late it should be slightly screwed forward and locked 
 by the nut for that purpose. 
 
 <The ignition tube G has an internal tube of smaller 
 diameter entering its lower end. This affords an annular 
 space communicating with the valve A. This valve, when 
 down upon its seat, permits the escape of the gas in the tube 
 after explosion, through a groove cut in the seating. This 
 
TYPES OF GAS ENGINES. 41 
 
 renders the tube more certain in its action, for it is obvious 
 that if the burnt products were pent up in the tube at a high 
 pressure, the fresh mixture to be exploded would not enter, 
 and would therefore not ignite. Messrs. Andrews make 
 their ignition tubes of a special alloy of silver, which they 
 maintain will resist the corrosive action of the explosive 
 gases at high temperatures. The valve A, fig. 23, is of special 
 use to assist the self starting of the engine. This is done in 
 the following way. The engine is barred round until the 
 connecting rod is on the top throw and about at right angles 
 to the crank. The best position for starting is governed 
 to a great extent, by the nature and quality of gas used, but 
 
 Fio. 25. "Stockport" Gas Engine Section of Cylinder End. 
 
 a few trials will enable the attendant to secure the best 
 position for a given case. The exhaust valve lever is geared 
 to the starting cam to prevent excessive compression at 
 starting. After taking the precaution to admit gas gradu- 
 ally to the gas bags and to have the ignition tube at a bright 
 red heat, the valve C, fig. 23, may be opened by raising the 
 weighted lever B into a vertical position. The timing 
 valve F will be open with the crank in the position for 
 starting ; consequently if gas is admitted to the cylinder by 
 means of a special starter pipe, it will find its way, together 
 with displaced air, through the valve F, up into the ignitiori 
 tube, downwards as indicated by the arrows, and out at A. 
 
42 
 
 TYPES OF GAS ENGINES. 
 
 This stream of air and gas continues until the mixture 
 becomes rich enough in gas to explode. The engine will 
 then start, and receive its next charge of gas through the 
 main gas valve supplying the engine. The exhaust lever 
 bowl is put to run upon the ordinary working cam ; the 
 valve A is closed by the action of the weighted lever B 
 pressing the inclined plane at 0. It will be noticed that if, 
 
 FIG. 26. "Stockport" Gas Engine Governing Arrangement. 
 
 in attempting to start, the gas enters the cylinder very 
 rapidly, it may displace all the air first and not be sufficiently 
 diffused to effect combustion when it passes into the hot 
 tube. For this reason the gas is not turned fully on to the 
 gas bags, so that it may have time to diffuse somewhat with 
 the air in the cylinder before passing into the ignition tube. 
 Immediately after the first impulse the gas should be fully 
 turned on. 
 
TYPES OF GAS ENGINES. 
 
 43 
 
 The engine is governed by first throttling the gas in the 
 inlet pipe when the governor rises. Should the speed 
 increase after throttling has taken place, the governor, still 
 rising, brings the throttle-valve lever, fig. 22, in contact 
 with the bell-crank lever actuating the gas-valve spindle, 
 and so cuts off entirely the gas supply. The most recently 
 
 FIG. 27. "Stockporf vaas Engine Governing Arrangement. 
 
 designed governor used upon the Stockport engine is of the 
 horizontal type, shown in figs. 26-27. A cam on the lay 
 shaft drives the gas-valve lever E. At the end of E is a 
 bell-crank lever, held in the position shown by the spiral 
 spring. The rod N, actuated by the governor, terminates 
 in a fork, which when moved sideways (see fig. 27) engages 
 the end of the bell-crank lever, overcomes the tension of the 
 
44 
 
 TYPES OF GAS ENGINES. 
 
 spiral spring, so that when the tripper blade rises it misses 
 the notch under A on the gas-valve spindle, thus cutting 
 out entirely the gas supply. The range of action of the 
 throttle valve depends to some extent upon the width of 
 the fork on the lever N, and when its limit is reached 
 the gas is entirely cut off. The speed of the engine is raised 
 
 FIG. 28. "Stockport" Gas Engine Arrangement of Governor for Small Engines. 
 
 by compressing the spring on the governor spindle by 
 means of two adjusting lock nuts. 
 
 The peculiar and sudden action of cam-driven levers 
 renders it possible to entirely cut out the gas supply, when 
 the speed increases, by a simpler mechanism than that just 
 described. Fig. 28 shows a form of simple vibrating 
 governor, which is fitted to the smaller engines of Stockport 
 
TYPES OF GAS ENGINES. 
 
 45 
 
 design. The valve lever carries a vertical spindle S. Upon 
 this spindle a cast-iron weight slides freely, but is supported 
 by a spiral spring, the height of which may be adjusted by 
 the two lock nuts shown. If the speed of the engine 
 increases beyond the limit, the inertia of the weight upon 
 
 FIG. 29. Moscrop Recorder Diagram. 
 
 the spindle S overcomes and compresses the spring support- 
 ing it. This in effect rotates the bell-crank lever slightly 
 in a clock-wise direction, causing the tripper blade to miss 
 the notch on the gas-valve spindle. 
 
46 
 
 TYPES OF GAS ENGINES. 
 
 CHAPTER IV. 
 THE "GRIFFIN" ENGINE. 
 
 This engine has already been mentioned, and the cycle has 
 been explained in the text relating to fig. 12. The Griffin 
 engine is manufactured by Messrs. Dick, Kerr, and Co., of 
 Kilmarnock, and since its introduction has been much 
 simplified, and has had a considerable sale in England. 
 Amongst other special features of the engine, the steadiness 
 of running under varying loads is noteworthy. Fig. 29 is 
 a Moscrop recorder diagram taken from a double-acting 
 engine with single cylinder, developing 70 horse power on 
 
 COMPARATIVE CYCLES or CAS ENGINES 
 
 OTTO ENCINI ONE CYUf,Dt* 
 
 A 
 
 ~ 
 
 
 
 . 
 
 - 
 
 
 
 A 
 
 - 
 
 
 
 A 
 
 
 
 A 
 
 
 
 
 i 
 
 
 
 
 i 
 
 KILMARNOCK tNc.Nt (WC^^om 
 
 
 - 
 
 
 - 
 
 
 
 - 
 
 * 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Kl LMARNOCKtNUNC Two Of UNOCM 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 OTTO ENC.NL O,CYUNOCR 
 
 ^ 
 
 
 
 
 
 
 
 
 A 
 
 
 
 
 ^ 
 
 
 
 A 
 
 
 
 
 
 
 
 
 
 
 GRIFFIN EMCINC. ONE.CYLINCCH 
 
 fc 
 
 
 
 
 
 
 ^ 
 
 
 
 ^ 
 
 
 
 ^ 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 ^ 
 
 KILMARNOCKtNUMt ONI C-rupotR 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 OTTO ENC.NI. ... TwoCTLiNOLR* 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 H 
 
 AL 
 
 F 
 
 Po 
 
 WE 
 
 K 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 OTTO ENGINE Jut CYUNOIK 
 
 n 
 
 
 
 
 
 
 
 
 ^ | 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 A 
 
 
 L^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 A 
 
 
 
 
 
 _ 
 
 
 
 
 
 
 ^ 
 
 KILMARNOCKtNCiNt ONCCYUNOIR 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 O-'o ENC.NC TwoCrLiNDOi 
 
 ^ 
 
 
 
 
 
 
 
 
 rt 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 FIG. 30. 
 
 A. Single-acting engines. B. Double-acting engines. 
 
 the brake, and fitted with a special electric lighting 
 governor. The distance between the horizontal lines 
 represents a variation in speed of 5 per cent, and the 
 distance between the vertical lines is traversed by the paper 
 in five minutes. The trial lasted If hours, and the loads 
 were varied between full load and one-third load, with a 
 variation in speed of about 2| per cent. The following 
 d agram, fig. 30, supplied to the author by the makers, 
 
UNIVERSITY 
 
 .CALIPOV^! 
 
 TYPES OF GAS ENGINES. 
 
 47 
 
 shows the impulses given to the pistons of various types of 
 engines running at full, three-quarter, and half power. It 
 will be noticed here that the Kilmarnock engine, running 
 at half load, has twice as many impulses as the ordinary 
 Otto engine running at full load, and 3| times as many at 
 half load. All the engines constructed by Messrs. Dick, 
 Kerr, and Co. larger than 12 brake horse power are double 
 acting. The double-acting six-cycle engines, having two 
 
 FIG. 31. 
 
 6AS SOPPL' 
 
 Fia. 32. 
 Sectional Elevation and Plan of Griffin Double-acting Gas Engine. 
 
 idle strokes, for scavenging purposes, are termed the 
 "Griffin" motor, and the double-acting engines, dispensing 
 with the idle strokes, are called the " Kilmarnock." In the 
 latter type the turning effort exerted upon the crank pin is 
 more regular, though, according to accepted theory, some 
 loss in efficiency may be occasioned by the presence of 
 burnt gases in the cylinder of this latter type. 
 
48 TYPES OF GAS ENGINES. 
 
 The details of a double-acting engine will be understood 
 on reference to fig. 31. This is an illustration of a recent 
 design of Griffin engine. The engine is of the horizontal 
 pattern, with balance weights on the crank webs. The side 
 shaft is driven by a 3 to 1 worm gearing on the crank shaft, 
 shown in fig. 32. The cylinder is closed at both ends, and 
 is consequently kept free from the grinding action of dust, 
 which in some factories may have a very serious efiect upon 
 an open-ended cylinder. The cylinder is water jacketed, 
 together with the piston-rod gland. From the one side 
 shaft all the valves and governor are driven. The two 
 exhaust ports are shown at E E, and are operated by cams, 
 acting on levers passing under the cylinder, lifting plain 
 mushroom valves. On the opposite side of the cylinder are 
 placed the governor gear, distribution valves, and igniting 
 slides. At / in fig. 32 the exterior of the gas- valve boxes 
 are shown. After passing through the valves at/, the gas 
 meets air drawn up through the bed, as indicated by the 
 arrows in fig. 31. Mixing with the air, it now passes 
 through the inlet valve R into the cylinder, the valve R 
 closes, and the charge is compressed in the combustion 
 chamber and the space above the valve R. 
 
 The ignition slide I is now opened by the eccentric A, 
 and makes communication between the flame in C and the 
 compressed mixture. The governor controls the gas supply 
 by throttling its passage through the valves in/ and/ until 
 the gas is entirely cut off. Referring only to the back end 
 of the cylinder, the sequence of operations is as follows : 
 The gas valve / is opened by its cam, and gas, mixing with 
 air, passes through R into the cylinder, R closes, and the 
 charge is compressed on the return stroke of the piston. 
 Ignition takes place, driving the piston forward. The 
 exhaust opens, and the products of combustion are expelled. 
 The exhaust valve E closes and R is opened, but as the gas 
 valve f remains closed, air only is drawn into the cylinder. 
 This in turn is expelled through the exhaust port E, and 
 the cycle being completed, a fresh charge is again drawn 
 
TYPES OF GAS ENGINES. 
 
 49 
 
 5G 
 
50 TYPES OF GAS ENGINES. 
 
 into the cylinder. Precisely the same action takes place in 
 the front end of the cylinder, thus giving two impulses to 
 the piston every six strokes. 
 
 Fig. 33 shows an outside view of the Kilmarnock 
 engine, the details of the valves and igniting arrangements 
 being somewhat revised. A timing valve, of the mush 
 room pattern, is substituted for the slide, and a hot 
 tube is substituted for the flame shown in the previous 
 illustrations. 
 
 Fig. 34 is an illustration of a recent design of engines as 
 supplied to the Corporation of Belfast for the generation of 
 electricity for lighting purposes. Four of these engines, 
 each indicating 120 horse power, are driving four 60 
 kilowatt dynamos, and two smaller engines, having single 
 instead of tandem cylinders, are driving two 26 kilowatt 
 dynamos. The efficiency of the engines and dynamos 
 combined proved to be 76 per cent, and the gas consumption 
 per electrical horse power was less than 24 cubic feet. This 
 value might be somewhat reduced if rich coal gas were 
 used, instead of, as in the present instance, the gas being 
 composed of coal gas and enriched water gas. The tandem 
 design has been adopted by Messrs. Dick, Kerr, and Co. 
 because of its narrow width, and consequent economy of 
 space. The increase in length, consequent upon the 
 position of the tandem cylinder, is utilised by placing the 
 cylinders between the dynamo and the engine crank shaft 
 an arrangement advocated upon a previous page. Each 
 dynamo is driven by eight J in. ropes, and for this reason it 
 is essential to run the engine contra-clockwise, in order 
 that the tight side of ropes shall be at the bottom. This 
 precaution, as has been already pointed out, is not essential 
 with a belt drive, but when ropes are used there is a danger 
 of them leaving the grooves if not kept tight at the bottom. 
 Each engine is fitted with a flywheel on the opposite end 
 of the shaft occupied by the rope-driving pulley. 
 
 In this design of engine there is an explosion every 
 stroke, but no scavenging strokes are made. In the front 
 
TYPES OF GAS ENGINES. 
 
 51 
 
 > 
 
 f 
 
 f UNIVERSITY 
 
52 TYPES OF OAS ENGINES. 
 
 and back of each cylinder a complete Otto cycle is carried 
 out once in two revolutions in the following order : First) 
 the charge in the back end of the back cylinder is fired ; 
 second, that in the front end of the forward cylinder ; third 
 that in the back end of the forward cylinder ; and lastly, 
 the charge is fired in the front end of the back cylinder. 
 The governing is carried out by a special electric light 
 governor, reducing the quantity of gas supplied to each of 
 the cylinders. This method is found preferable to one in 
 which the gas supply is entirely cut out. At Belfast the 
 starting of the engine is effected by utilising electric 
 energy stored in the batteries, by sending a current through 
 the dynamos, converting them for the time being into 
 motors, and so driving the gas engines, until a charge is 
 drawn into the cylinder and ignited. 
 
 The exhaust pipe of each engine communicates with a 
 main exhaust pipe, to which a silencer is fitted, but it is 
 found that there is less probability of noise when two or 
 more engines are working than when only one is ex- 
 hausting, for the flow of exhaust gas is then more 
 regular, and does not cause that coughing noise often so 
 objectionable. 
 
 Besides a large circulating tank for supplying the water 
 jackets, provision is made for admitting water from the town 
 mains direct to the jacket ; and as a still further precaution, 
 each engine is provided with a circulating pump for driving 
 water through the jacket. 
 
 At Coatbridge,* near Glasgow, Megars. Dick, Kerr, and Co. 
 have supplied engines for driving the electric lipht, worked 
 by producer gas. In this installation each engine has two 
 cylinders placed side by side instead of tandem. The 
 starting of the engines is effected by utilising the pressure 
 in the steam boiler used in connection with the gas- 
 producing plant, about which more will be written. A 
 special valve is fitted to the back of one cylinder of the 
 engine, and is operated entirely by a hand lever. A steam 
 pipe is connected to each of these valves, and steam enters 
 
 Hinco tho publication of tho firm edition of ilnu bonk, these engines h*v 
 been removed and Ntoum substituted. 
 
TYPES OF GAS ENGINES. 63 
 
 the cylinder when the operator opens the starting valve. 
 The operator, performing the work of the eccentric on a 
 steam engine, opens and closes the starting valve until the 
 engine has acquired a momentum sufficiently great to draw 
 into the cylinder, and compress, a charge. Ignition then 
 takes place. The largest gas engine made by this firm 
 indicates over 700 horse power, and is specially interesting, 
 inasmuch as it is constructed upon a compound principle. 
 The engine somewhat resembles a large triple-expansion 
 horizontal steam engine;, and has two 10 ft. diameter by 
 14 in. face flywheels. Two 21 in. diameter by 30 in. stroke 
 cylinders are placed one on each side a 32 in. diameter 
 cylinder having a 36 in. stroke. One of the high-pressure 
 cylinders only exhausts into the central cylinder, the other 
 being kept separate for starting the engine by the applica- 
 tion of steam pressure, as above described. 
 
 CHAPTER V. 
 
 THE TANGYE ENGINE 
 
 MESHKH. TANGYK'H gas engines work upon the Otto cycle, 
 but improvements have been made in the igniting and 
 governing gear, and in a self-starting arrangement. Fig. 35 
 bhows an outside view of an engine which will give a 
 maximum indicated horse power of 115. Messrs. Tangye 
 construct single-cylinder engines, indicating HP. to 
 150 H.K, all of which work with either town or other gases 
 made by their own producer plant. The combustion 
 chamber is specially formed to prevent shock to the 
 working parts during explosion, and also assists in making 
 the charge more homogeneous. Messrs. Tangye claim that 
 these improvements render their engines specially useful 
 for electric lighting, a large number having already been 
 made for that purpose. It is stated that, working with gas 
 
54 TYPES OF GAS ENGINES. 
 
 costing 2s. 6d. per 1,(00 cubic feet, a 30 H.P. engine will 
 produce 1 kilowatt for each penny expended, which works 
 out to about 2 cubic feet of gas per 10 candle power lamp. 
 The weight of anthracite coal used in the producers for 
 running an engine of about 70 brake horse power is said to 
 be less than 1 Ib. per B.H.P. per hour. 
 
 The large engines are now started with a new and very 
 powerful starter, which is capable of starting the engine 
 against half its working load. A charge of gas and air is 
 pumped by hand into a detached receiver, there being an 
 exc< ss of gas present, thus rendering the charge inexplosive 
 in the receiver. The engine is barred round into the 
 position for the commencement of a forward stroke. The 
 charge in the receiver is pumped up to a pressure of about 
 40 Ib. per square inch. At this pressure communication is 
 made with the working cylinder, and the charge enters the 
 combustion chamber, and by mixing with the air present in 
 the combustion chamber forms a strong explosive mixture. 
 This, however, is at first prevented from igniting by the 
 closed ignition valve being tightly held upon its seat by a 
 lever operated by a cam upon the side shaft. Under the 
 initial pressure of the gas entering the combustion chamber 
 the piston moves forward ; the cam then releases the 
 ignition valve, and explosion takes place. 
 
 By setting the engine in motion before igniting the 
 charge, excessive stress upon the working parts is greatly 
 avoided. Fig. 36 is a drawing of the self-starter patented 
 by Mr. C. W. Pinkney, to whom many other improve- 
 ments in Messrs. Tangye's gas engines are due. In the 
 drawing, the ignition tube I and starting pump are shown 
 together ; this arrangement is, of course, subjected to 
 modifications necessary in special cases. The valves a and 6 
 are the suction and delivery valves respectively of the 
 starting pump P. The valve a, shown separately, is con- 
 tained in the same valve box as 6, and is in communication 
 with the passage c. A gas pipe is coupled to a, and by the 
 small port shown the gas mixes with the air drawn in 
 
TYPES OF GAS ENGINES. 
 
 55 
 
56 
 
 TYPES OF GAS ENGINES. 
 
 through the valve a. The gas port terminates in an annular 
 space, to facilitate the mixing with the air. The proportion 
 of these valves is such as to admit a mixture of about four 
 volumes of air to one of gas. This may be delivered to the 
 receiver, as previously mentioned, or may be pumped 
 directly into the combustion chamber. In the latter case 
 
 FIG. 36. Pinkney's Self-starter for G*s Engines. 
 
 the handwheel li controls admission to the ignition tube, 
 and the engine is arranged to fire only when the hand 
 wheel is turned. The valve e is held up to its seating by a 
 spiral spring beneath it, and the spring is adjusted so that 
 the valve e may open when the desired pressure in the 
 pump and combustion chamber, or receiver, has been 
 
TYPES OF GAS ENGINES. 57 
 
 acquired, thus giving an indication that the mixture is 
 ready for firing. In connection with this starter, Mr. 
 Pinkney has patented an arrangement for preventing the 
 movement of the engine piston when the pressure of the 
 hand pump is first felt in the combustion chamber ; for it is 
 obvious that, when the hand pump delivers directly into 
 the combustion chamber of a small engine, the crank will 
 slowly move towards the forward dead centre, and be in 
 that position when the charge is ready for firing. To 
 obviate this, a stud upon the crank web engages with a 
 spring catch of sufficient strength to hold the crank in its 
 starting position until the greater pressure of the explosion 
 causes its release. The bar, with its catch, then falls clear 
 of the crank as it revolves. This gear is only suitable for 
 
 
 FIG. 37. Starting and Working Diagram taken from Tangyes' Gas Engine. 
 
 small engines fitted with self-starters, larger engines being 
 fitted with a special receiver to contain the charge until it 
 is admitted to the combustion chamber by means of a valve 
 operated by hand. 
 
 An indicator diagram taken from an engine of this latter 
 type is shown in fig. 37. It is seen from this diagram that 
 
58 TYPES OF GAS ENGINES. 
 
 the engine piston is standing in a position equal to one- 
 ninth of its forward stroke before communication is made 
 between the combustion chamber and the receiver. The 
 pressure immediately rises when the communicating valve 
 is opened to about 40 Ib. per square inch. The diagram 
 also shows that two-ninths of the forward stroke are 
 completed before ignition takes place. The remainder of 
 the full-line diagram shows the explosion and expansion 
 during the first revolution. The dotted diagram has been 
 taken after the engine has acquired its proper working 
 speed. The latter shows a maximum pressure of 200 Ib. per 
 square inch, and a compression pressure of 72 Ib. per square 
 inch. 
 
 Messrs. Tangye's engines have a specially hard metal 
 liner fitted to the cylinder, which can quickly be replaced. 
 Lubrication is effected by motion derived from the side 
 shaft. A ratchet wheel, shown on the lubricator in the 
 external view of these engines, slowly revolves, and thereby 
 raises a plunger against the resistance of a strong spiral 
 spring. Oil follows into the plunger chamber, and is 
 injected into the cylinder by the release of the plunger from 
 the top of its stroke. The worm wheels driving the side 
 shaft are cut from solid metal, and run in an oil bath, the 
 casing of which can easily be removed for examination of 
 the wheels. The governor is of the centrifugal high-speed 
 type, and acts directly upon the gas supply without 
 throttling. 
 
 THE FIELDING GAS ENGINE. 
 
 This engine is made by Messrs. Fielding and Platt, of the 
 Atlas Works, Gloucester, and is constructed upon the Otto 
 cycle. Fig. 38 shows an outside view of the engine. The 
 valves, which are of the plain mitre type, are driven in the 
 usual way from the crank shaft, by means of cams on a side 
 shaft. The governor is of the ordinary high-speed type, and 
 is driven from the side shaft by bevel wheels. Another 
 type of governor used on Messrs. Fielding and Platt's earlier 
 
TYPES OF GAS ENGINES. 
 
60 TYPES OF GAS ENGINES. 
 
 engines was actuated by a cam on the side shaft. In this 
 case the governor consisted of a simple dashpot and piston, 
 the rod of which was connected to a lever working the gas 
 
 FIG. 39. Indicator Diagram from Fielding and Platt's Engine, running with 
 no load. 
 
 valve. If the speed of the engine increased beyond that to 
 which the dashpot was adjusted, then the cam left the 
 piston at the top of its stroke, coming round into its lifting 
 
 FIG. 40. Indicator Diagram from Fielding and Platt's Engine, running at 
 about one-third load. 
 
 position before the piston had time to descend ; consequently 
 the gas valve was not opened. 
 
 A feature in the governing of this firm's electric lighting 
 
 FIG. 41. Indicator Diagram from Fielding and Platt's Engine, running at 
 nearly full load. 
 
 engine is illustrated by the accompanying indicator dia- 
 grams, figs. 39, 40, and 41. The governor is of the ball type, 
 and regulates the gas and air supply, so that even when 
 
TYPES OF GAS ENGINES. 61 
 
 running light an explosion is never missed. This arrange- 
 ment, though possibly slightly increasing the consumption 
 of gas, undoubtedly conduces to steady running under a 
 fluctuating load. It is stated by Messrs. Fielding and Platt 
 that the variation in speed between light and full load is 
 about 3^ per cent. The gas consumption in an engine of 
 this type, of 100 B.H.P., is claimed to be under 20 cubic feet 
 per hour per B.H.P. 
 
 Messrs. Fielding have recently patented an efficient system 
 of starting their larger gas engines by the use of compressed 
 air. When the engine is about to be stopped after its first 
 run the gas is turned off, and the piston and engine cylinder, 
 by a slight alteration in valves, is converted into an air 
 compressor, which discharges into a reservoir until a pressure 
 of about 60 Ib. is obtained. In this way the energy stored 
 in the flywheels is utilised until the engine comes to rest. 
 The charge of air thus obtained is sufficient to start from 
 six to twelve times. The starting is effected in the following 
 way : The crank is barred just over the back dead centre, 
 and gas is admitted by a hand gas tap to the combustion 
 chamber. The displaced air is allowed to escape through an 
 open cock until gas which will burn with a clear blue flame 
 begins to issue from the open cock. These cocks are now 
 closed, and compressed air from the reservoir is admitted to 
 the combustion chamber, forming a rich mixture. The 
 igniting valve is now opened, and a powerful impulse is 
 given to the piston. It is said that the engine will start by 
 this means even against half the full load. 
 
 THE "PREMIER" GAS ENGINE. 
 
 The "Premier" gas engine, patented by Mr. J. H. Hamilton, 
 B.Sc., is manufactured by Messrs. Wells Brothers, Sandiacre, 
 near Nottingham. An external view is shown in fig. 42, 
 with modified valve gear. Its characteristic feature is the 
 scavenging arrangement. Fig. 43 is a sectional view of the 
 engine, and from it the working of the scavenging stroke 
 
62 
 
 TYPES OF GAS ENGINES. 
 
 
TYPES OF GAS ENGINES. 63 
 
 may easily be followed. There are two pistons, C and D, 
 cast together with a cylindrical connection. The piston C 
 receives the impulse due to the explosion of the charge, 
 whilst the piston D acts merely as an air pump. The 
 exhaust valve E is actuated by a lever driven by a cam on 
 the side shaft (see fig. 44). The gas and air valves are driven 
 by one lever, and are arranged, as shown, upon the same 
 gpindle. The gas valve is made an easy fit upon the air- 
 valve spindle, but is held up to its seat by a spiral spring. 
 Upon the outstroke of the pistons, air is drawn through 
 rubber or leather valves in the bed plate, shown at A. The 
 chamber marked A has a free communication with the 
 chamber marked B. When the air valve is depressed beyond 
 the position shown in the drawing, the gas valve is opened 
 by contact with the collar shown upon the spindle. The 
 exhaust valve E being closed during the outstroke of the 
 pistons, air, mixing with the gas, enters the combustion 
 chamber, as indicated by the arrows. When sufficient gas 
 to form a charge has entered, the air-valve spindle slightly 
 lifts, closes the gas supply, but still admits air to complete 
 the charge. Upon the instroke of the pistons, the charge 
 is compressed in the combustion chamber to about 60 Ib. per 
 square inch ; at the same time the air entrapped between the 
 two pistons becomes compressed to about 9 Ib. per square 
 inch. Thus the air valve has upon its lower face a pressure 
 of 60 Ib. per square inch, and upon its upper face 9 Ib. per 
 square inch ; the difference, 51 Ib. per square inch, tends to 
 keep the valve tight upon its seat. In order to control 
 the action of the air valve, and to make its movement 
 independent of the difference of pressure upon its upper 
 and lower faces, a strong spiral spring is attached to the 
 lever L, fig. 44, tending always to force the air valve on its 
 seat. The cam driving L gives the air valve a long and a 
 short stroke alternately, the former for the admission of the 
 charge, the latter for the admission of air only. 
 
 When the charge is compressed it is fired by a timing 
 valve, and both pistons travel outwards. During this out. 
 
64 
 
 TYPES OF GAS ENGINES. 
 
TYPES OF GAS ENGINES. 
 
 65 
 
 stroke the air already compressed by the larger piston D 
 expands, and gives back to the engine the energy stored in 
 it by compression. At the end of the outstroke the exhaust 
 valve opens, and releases the pressure in the combustion 
 chamber. It is evident that so long as the pressure in 
 the combustion chamber during the exhaust stroke is 
 nearly atmospheric, the piston D will displace a volume of 
 
 FIG. 44.- End View of Premier Gas Engine. 
 
 air and drive it through the combustion chamber into the 
 exhaust pipe. This effectually clears the combustion 
 chamber of its burnt products, and this done, the exhaust 
 closes, and the cycle is repeated. The indicator diagram 
 shown in fig. 45 has been taken from the working cylinder, 
 whilst the diagram shown in fig. 46 has been taken from 
 the pumping cylinder. Following the direction of the 
 arrows on fig. 46, it will be seen that little or no work is 
 lost in compressing the entrapped air in the chamber B, for 
 
 6G 
 
66 
 
 TYPES OF GAS ENGINES. 
 
 the air-expansion curve nearly coincides with the compres- 
 sion curve. At the commencement of the exhaust stroke the 
 pumping piston raises the air pressure, the air valve opens, 
 and air is forced into the combustion chamber against the 
 
 FIG. 45. Working Diagram from Premier Gas Engine. 
 
 pressure of the exhaust. This point is shown on the diagram 
 at P ; the area P A S E of this diagram gives the work done 
 by the pump in retarding the motion of the engine ; and the 
 net indicated horse power is found by subtracting from the 
 
 FIG. 46. Pumping Diagram from Premier Gas Engine. 
 
 I.H.P. of the working cylinder the I.H.P. of the pumping 
 cylinder. It must be remembered that the area P A S E of 
 the pumping diagram occurs only at every alternate revolu- 
 tion; consequently, in obtaining the H.P. the revolutions 
 
TYPES OF GAS ENGINES. 67 
 
 per minute must be divided by 2. It must also be noted 
 that the effective area of the piston D, fig. 43, is not its 
 whole area, but only that which remains after subtracting 
 from it the area of the piston C. 
 
 In engines of this type it is extremely important, from 
 the makers' point of view, to have a free exhaust pipe, for 
 cases are not unknown in which the undue throttling of the 
 exhaust pipes has so raised the back pressure as to cause 
 the burnt products to enter the chamber B as soon as the 
 air valve opens, thus entirely preventing the scavenging 
 action. It is, for another reason, highly important to keep 
 the exhaust pressure low, for the area of the pumping 
 diagram depends upon it, and represents lost work. 
 
 This engine is started by means of a hand pump in the 
 following way : The bowl on the exhaust lever is made to 
 engage with a projection on its cam, so that the compression 
 of the charge may be relieved. The engine is then barred 
 round until the crank has turned through about 45 deg. 
 of its forward working stroke. It is very important that 
 the engine should start from such a position that all the 
 valves are ready for the impulse, for it is obvious that by 
 careless handling the engine might be started from the 
 position corresponding to the suction stroke. In the latter 
 case the exhaust valve would not open at the end of the 
 stroke, and the burnt products would enter the chamber B, 
 and probably damage the valves at X. When the engine is 
 in the correct position, the timing valve is closed by a short 
 lever inserted by hand. Gas is then turned to the hand- 
 pump suction valve, and two or three strokes taken with 
 the lever H. The gas supply to the pump is then turned 
 oft, and the main gas cock to the engine opened. The hand- 
 pump lever is again worked, and the pump delivers air to 
 the combustion chamber. As soon as the engine piston 
 commences to move forward under the pressure of the 
 charge, the timing valve is released by hand, the charge 
 enters the ignition tube, and a strong forward impulse is 
 given. When the engine has acquired some momentum, 
 
68 TYPES OF GAS ENGINES. 
 
 the exhaust bowl is put into its working position, and gives 
 the correct compression to the charges. The relative number 
 of strokes of the hand pump when drawing gas and air 
 respectively must be determined for each case by trial, as 
 the quality of gas varies considerably. 
 
 The governing of the engine is effected by the movement 
 of the rod M, fig. 44, actuated by an ordinary high-speed 
 centrifugal governor. When M is moved to the right its 
 knife edge acts upon the gas lever L at a greater distance 
 from its fulcrum ; consequently the gas valve is not fully 
 
 FIG. 47. End of View of " Forward" Otto Gas Engine. 
 
 opened. It is merely a matter of arrangement whether the 
 gas supply be entirely cut off or gradually decreased. In 
 fig. 44, the lever M would first throttle the gas supply, and 
 after a further increase in speed cut it off entirely. 
 
 The ignition tubes supplied by this firm are made in 
 nickel, and will stand continuous use for several months. 
 
 The " Forward " gas engine is manufactured by Messra 
 T. B. Barker and Co., of Birmingham. Figs. 47 show a 
 sectional elevation and end view of this engine. The 
 cylinder is 10 in. diameter by 20 in. stroke, and running 
 at a speed of 180 revolutions per minute the engine 
 will develop about 25 I.H.P., and about 20 on the brake. 
 
f UNIVERSITY 1 
 \o^ ^J 
 
 TYPES OF GAS ENGINES. 
 
 69 
 
70 TYPES OF GAS ENGINES. 
 
 An important feature in this design is the reduction 
 of port surface to a minimum. It will be noticed that the 
 air and exhaust valves A and E, fig. 47 A, - are placed as near the 
 combustion chamber as possible. They are fitted to separate 
 seatings, as shown at A and E. The gas is admitted by the 
 valve G, fig. 47, and mingles with the air on its way to the 
 valve A. The spring S is connected to a bar, one end of 
 which rests upon the crossbar C, and the other upon the 
 levers driving the valves. By this arrangement one spring 
 suffices. Messrs. Barker and Co. do not use an ignition 
 valve, but rely entirely upon the compression of the charge 
 into the ignition tube for the correct timing of the 
 explosion. The cylinder is formed by a liner of special 
 cylinder metal, having an asbestos joint where it is attached 
 to the combustion chamber. This to some extent pre- 
 vents the conduction of heat from the combustion chamber 
 to the fore end of the cylinder. The joint at the front end 
 of the cylinder is made with indiarubber, and the liner is 
 free to expand in the direction of its length as it becomes 
 heated. As the function of this joint is merely to retain 
 the jacket water, there is no necessity for an exceedingly 
 tight fit. The crank shaft is of forged Siemens-Martin 
 steel, and is fitted with phosphor-bronze bearings, the 
 mean pressure upon the bearings being 100 Ib. per square 
 inch during the working stroke. The crank-pin pressure 
 allowed is about 370 Ib. per square inch, whilst the piston 
 pin carries a mean pressure of 530 Ib. per square inch. 
 
 These figures are worked out on the mean effective 
 pressure given by the indicator diagram during a working 
 stroke. The piston is packed with three cast-iron rings 
 in. wide, having a baffle groove J in. wide cut in each 
 ring, dividing the bearing surfaces into two bands, each 
 in. wide. The pressure exerted upon the walls of the 
 cylinder amounts to about 5 Ib. per square inch of surface. 
 The piston speed of this engine running at 180 revolutions 
 is nearly 600ft. per minute, and the velocity of the inlet 
 and outlet through the valves is about 100 ft. per second. 
 
SELF-STARTERS. 71 
 
 The following figures are quoted from a trial of a " Forward " 
 engine of 30 B.H.P., conducted by Messrs. Morrison and 
 Ldinchester for the City of Birmingham Gas Dept. in 1894. 
 When running with no load, the ratio of air to gas by volume 
 was 14*5 to 1 ; when running with full load, the ratio was 9 '9 
 air to 1 volume of gas. In the latter case the products of 
 combustion were not swept out as the engine fired every 
 cycle ; consequently the volume of air plus products to that 
 of the gas was 11 to 1. The gas consumption in cubic feet 
 per hour per I.H.P. was 1775, and per B.H.P. 21'05 cubic 
 feet. To this should be added a total of 5 '25 cubic feet per 
 hour used by the burner for heating the ignition tube. The 
 mechanical efficiency worked out to 84 per cent, and the 
 thermal efficiency calculated on the I.H.P. was 22'2 per cent, 
 and 18'6 when calculated on the B.H.P. The engine was 
 governed when running light by cutting out the gas 
 supply. 
 
 CHAPTER VI. 
 
 SELF-STARTERS. 
 
 THE Forward engine is fitted with Lanchester's patent self- 
 starting gear, the application of which is shown in the 
 following drawing. Fig. 48 shows a general view of the 
 arrangement fitted to the engine. Gas enters the combus- 
 tion chamber from a | in. branch through the valve A. The 
 stream of gas and air passes out of the cylinder through 
 the cock C, and ignites, when rich enough in gas, at the 
 naked flame shown at F. The mixture burns at the orifice, 
 at first with a pale blue flame, which becomes deeper in 
 colour as it becomes richer. When the flame burns with a 
 roaring noise the valve A should be closed. The flame at 
 the orifice will now strike back into the cylinder and ignite 
 the entire volume of the mixture in the combustion chamber. 
 The action of this starter depends entirely upon the size of 
 
72 
 
 SELF-STARTERS. 
 
 the outlet through the cock, for if this outlet were large 
 in diameter the velocity of the gas would be less than 
 the velocity of flame propagation ; consequently, imme- 
 diately the gas ignited at the orifice, the flame would strike 
 back and ignite the mixture in the cylinder before it was 
 rich enough to give a starting impulse to the piston. This 
 
 FIG. 48. Lanchester's Self-starter for Gas Engines. 
 
 is prevented by the contracted area of the outlet nozzle, 
 causing the gaseous mixture to flow rapidly from the orifice. 
 This is, of course, the principle upon which all atmospheric 
 burners depend, and the slight report which is always heard 
 when turning out an atmospheric gas stove, is due to the 
 return of the flame along the pipes towards the stop cock 
 at the moment the latter is turned off. 
 
SELF-STARTEKS. 73 
 
 To prevent the escape of the pressure through the cock 
 from the engine cylinder, when the first ignition takes 
 place, an automatic valve is placed inside the chamber in 
 the cock C. This valve is shown at V, and rests upon its 
 lower face, excepting when its weight is overcome by the 
 
 FIG. 49 Green's Self-starter for Gas Engines. 
 
 rush of an explosion. Grooves cut in this valve allow the 
 passage of gas through the open cock so long as the valve 
 remains in its lowest position. Immediately the starting 
 impulse is given this valve is projected upwards, and as 
 the grooves are not deep, the centre of the valve entirely 
 closes the orifice, thus preventing the escape of the pressure. 
 The cock C may now be turned off. In using this starter 
 the engine must always be barred round to the commence- 
 ment of its working stroke, otherwise serious damage may 
 be done to the gas bags and other fittings. 
 
74 
 
 SELF- STARTERS. 
 
 A self-starter similar in principle to that already 
 described is shown in figs. 49 and 50. This is known as 
 Green's self-starter. It differs from that already described 
 in the construction of the automatic valve. Referring to 
 fig. 50, the passage marked 17 communicates directly with 
 the combustion chamber. The valve B is cylindrical, and 
 hollowed out to the point C, but solid at the lower end. 
 The holes 4 1} 4, when the valve is in position for starting, 
 form a passage for the gas and air, which rise to the orifice 
 6, and ignite by the naked name issuing from the gas 
 burner E. When the mixture issuing from the orifice 6 
 
 FIG. 50. Green's Self-starter for Gas Engines. 
 
 burns with a dark blue flame, the gas supply to the 
 combustion chamber is shut off, the flame strikes back 
 through the passages 4, 18, 17, and ignites the charge in the 
 combustion chamber. The pressure thus generated forces 
 up the valve B, thereby closing the passages 4 1} 4. The 
 hand lever shown in fig. 49 is made of spring steel, and 
 when in the position shown presses downwards upon the 
 lever engaging with the automatic valve, thus effectually 
 keeping it in its highest position, and preventing the escape 
 of pressure from the combustion chamber when the engine 
 is working. When the starter is in use, the hand lever is 
 
SELF-STARTEES. 
 
 75 
 
 drawn forward, and holds the automatic valve in the position 
 shown in fig. 50. The advantages of this form of con- 
 struction are that the automatic valve serves to throw the 
 starter out of use, and being at the same time accessible 
 from the outside, there is no inconvenience occasioned by 
 the valve sticking. 
 
 The self-starters we have described ignite the mixture at 
 atmospheric pressure ; hence the impulse of the starting 
 stroke is limited, and the maximum pressure in the cylinder 
 will not usually exceed 80 Ib. per square inch. In order to 
 increase this pressure, Mr. Dugald Clerk was the first to 
 apply the apparatus shown diagrammatically in fig. 51, 
 which is known as the Clerk pressure starter. A chamber 
 C is connected to the engine cylinder by the pipe P and 
 a check valve V. The passage H leads into C from a hand 
 
 Fio. 51. The Clerk Pressure Starter. 
 
 pump, which is used to charge C and the engine cylinder 
 with an explosive mixture slightly above atmospheric 
 pressure. I is an igniting valve, through which the mixture 
 passes when open, and comes in contact with the flame F, 
 which causes the ignition. The action of the starter is as 
 follows : The cylinder and chamber C being filled at 
 atmospheric pressure with an explosive mixture, the igniting 
 
76 SELF-STARTERS. 
 
 valve is opened, the mixture fires at the orifice, and the 
 flame strikes back into the chamber C. As the gas burns 
 downwards it expands, driving a large portion of the 
 mixture yet unburnt through the pipe P into the cylinder. 
 In this way the pressure of the mixture in the cylinder is 
 considerably raised before the flame reaches it. When, 
 therefore, the mixture in the cylinder ignites, it explodes 
 
 FIG. 52. Clerk- Lunchester High-pressure Self-starter. 
 
 violently, developing a maximum pressure of 190 Ib. per 
 square inch, instead of 80 Ib., as with the low-pressure 
 starters of Lanchester and Green. 
 
 The self-starter shown in fig. 52 is known as the Clerk- 
 Lanchester. It is, as will be seen, a combination of the 
 Clerk pressure starter and the Lanchester automatic valve. 
 It is otherwise more convenient than the Clerk starter, as 
 no pump is required to fill the cylinder and chamber C with 
 
THE DAY GAS ENGINE. 77 
 
 an explosive mixture. The action of this starter is as 
 follows : The engine cylinder and chamber C are filled with 
 air at atmospheric pressure. To ensure pure air occupying 
 these spaces, it is usual to open the valve V before the 
 engine is stopped, thus drawing pure air through C into 
 the cylinder. When ready for starting, the engine is barred 
 round to the commencement of its working stroke, the 
 valve V opened, and gas turned on ; the gas jet G is lighted, 
 and serves to ignite the mixture as it issues from the orifice. 
 Immediately the flame strikes back into the chamber C the 
 valve L lifts and closes the orifice. The action is now 
 precisely the same as already described in the Clerk pressure 
 starter. 
 
 All makers supply with their larger engines some self- 
 starter, which, though differing in detail from those already 
 described, may nevertheless be recognised as dependent 
 upon one or other of the principles explained.* 
 
 CHAPTER VII. 
 TWO-CYCLE AND OTHER ENGINES. 
 
 THE large number of gas-engine makers who are now 
 competing has given rise to a variety of distinguishing 
 names, which the author ventures to think are far in excess 
 of the corresponding differences which those names are 
 supposed to characterise. Having now fully described a 
 variety of engines which may be considered typical, but 
 little will be gained by dwelling at length upon minute 
 details. It will be advisable, before leaving the descriptive 
 side of the subject, to touch briefly upon points of interest 
 in connection with engines not already described. 
 
 The Day gas engine, made by Messrs. Day and Co., Bath, 
 is illustrated in fig. 53. This engine is noticeable for its 
 simplicity and fewness of parts. Though not suitable for 
 
 * Other methods of starting gas engines are described on jages 41, 52, 5j, 56. 
 61, and 67 
 
78 
 
 THE DAY GAS ENGINE. 
 
 large powers, it is a moderately efficient engine where only 
 light work is done. It will be noticed that the crank and 
 connecting rod run in a closed space, into which is fitted one 
 mushroom valve. Through this valve both gas and air are 
 
 FIG. 63. The Uay Gas Eugme. 
 
 sucked during the up stroke of the piston. On the down 
 stroke this mixture is slightly compressed. When the 
 piston descends, the port P is uncovered, and the mixture 
 Hows from the lower chamber into the working cylinder. 
 
THE PALATINE GAS ENGINE. 79 
 
 On the return stroke the port entrance to the cylinder is 
 covered by the piston, and the charge being compressed 
 into the ignition tube I, is fired without the use of a timing 
 valve. The piston is moved downwards by the expanding 
 gases, which escape through the port E when uncovered by 
 the piston. It will be observed that the exhaust port is 
 uncovered before the inlet, and therefore the pressure of the 
 expanding gases is released before the end of the stroke. 
 The inlet and exhaust are open together for a short period 
 of time, and the first rush of fresh mixture is deflected up- 
 wards into the cylinder ; in this way the greater portion of 
 the burnt products are displaced. As the working of the 
 valve is entirely automatic, the engine may be run in either 
 direction without any alteration whatever. This engine 
 gives an impulse every revolution. 
 
 The Palatine gas engine is constructed somewhat upon 
 this principle, the crank shaft running in a closed chamber. 
 The object of this design is to supply a scavenger stroke, 
 and the impulse only occurs every two revolutions. The 
 objection to these designs is the inaccessibility of the 
 connecting rod. 
 
 Messrs. Alfred Dougill and Co., Leeds, make an Otto 
 cycle engine, which is noticeable for its patent governor 
 gear. This gear is shown in figs. 54 and 55. The object of 
 the arrangement is to keep the ratio of air to gas uniform 
 throughout all variations of load. Instead of merely 
 intercepting the gas supply, the governor acts on both the 
 air and gas valves, and when the speed rises the total 
 volume of the charge admitted to the cylinder is decreased, 
 but the ratio of air to gas remains unchanged. The 
 essential parts of the gear are the stepped cam O and the 
 sleeve L. The lever P lifts both the gas and air valve 
 spindles N and M, and is driven by the stepped cam 
 operating upon the movable sleeve L. In fig. 55, L is 
 shown in line with the highest step ; consequently the 
 admission of the charge is a maximum. When the governor 
 balls F rise the sleeve is drawn towards the lower steps, and 
 
80 
 
 DOUGILL'S GOVERNOK. 
 
 oe 
 
ACME GAS ENGINE. 81 
 
 the lift of the valves is diminished. It will be observed that 
 since an increase in speed must take place on the explosion 
 stroke, the roller of the sleeve L is at the time of the 
 increase running upon the cylindrical portion of the cam O. 
 For this reason it is perfectly free to move sideways when 
 the speed varies. Diagrams obtained from this engine are 
 shown in fig. 56. From these it will be seen that the 
 compression is reduced when running light, and the ex- 
 plosion is nearly as rapid as at full load. This governor is 
 said to prevent a variation in speed of more than 2 per 
 cent on account of the continuity of the impulses even at 
 
 FIG. 56. Diagram taken from Dougill's Otto Gas Engine. 
 
 light loads, and it is therefore highly suitable for electric 
 lighting purposes. 
 
 The Acme engine, made by Messrs. Alexander Burt, 
 Glasgow, was designed with the object of obtaining maxi- 
 mum expansion. This engine comprised two cylinders, 
 working with two crank shafts. The shafts were geared 
 together with two to one gearing, so that one piston made 
 twice the number of strokes of the other. The gases expan- 
 ded from the higher speed cylinder to the lower one, and the 
 exhaust was said to be much cooler than usual. The results 
 obtained showed the engine to be economical ; but the noise 
 of the gearing, and the extra wear of these parts, probably 
 operate against the more general adoption of the engine. 
 
82 TWO-CYCLE ENGINES. 
 
 During the last few years there has been a growing 
 tendency for all makers to adopt the Otto cycle in prefer- 
 ence to their own particular makes. We have already 
 mentioned that this cycle leaves much to be desired in 
 regard to steadiness of running. It is not surprising, there- 
 fore, that a reaction is taking place, and not a few are 
 trying to solve the problem of constructing a satisfactory 
 engine having an impulse every revolution. The Griffin 
 engine, as already described, gives an average of one 
 impulse every revolution, but utilises both ends of the 
 cylinder for the combustion of the charges. For more 
 reasons than one it is desirable to utilise only the space 
 at the back of the piston as an explosion chamber, the 
 forward end merely acting as a pump. 
 
 The first engine built, in 1878, to give an impulse 
 every revolution was designed by Mr. Dugald Clerk, 
 Assoc.M.Inst.C.E. It will be well to describe briefly the 
 working of this and other two-cycle engines, not because 
 they are regarded as modern gas engines, but because the 
 author believes that the gas engine of the future will 
 probably be a combination of past and present attempts to 
 construct an engine having an impulse every revolution. 
 
 Mr. Dugald Clerk's first engine consisted of two cylinders, 
 one of which was used as a motor cylinder, the other as a 
 pump to deliver the mixture into a reservoir at a pressure 
 of about 70 Ib. per square inch. The motor piston travelled 
 nearly to the end of the cylinder. The usual combustion 
 chamber was dispensed with. When the motor piston had 
 moved forward about 2 in. on its working stroke, a slide 
 valve admitted the mixture from the reservoir. The supply 
 being cut off, ignition took place. In this engine the in- 
 9 an tor sought to combine the advantages of compression 
 before ignition with the complete expulsion of the exhaust 
 gases by minimising the clearance at the back of the cylinder. 
 The difficulties met with in this engine were chiefly due to 
 back ignition in the compression reservoir, and to excessive 
 shock in the motor cylinder. The former of these diffi- 
 
TWO-CYCLE ENGINES. 83 
 
 culties was never overcome, though the latter was minimised 
 by modifying the shape of the combustion chamber. 
 
 In 1880 Mr. Clerk designed another engine, in which the 
 pump delivered its charge directly into the motor cylinder 
 slightly above atmospheric pressure. The exhaust ports were 
 uncovered by the working piston, and just before being closed 
 by the return of the piston the displacer pump delivered 
 its charge into the working cylinder. This was compressed 
 on the return stroke and ignited at the commencement of 
 the forward stroke. Many engines of this pattern were 
 made, and in the larger sizes worked quite as economically 
 as the best engines of their time, though there were 
 difficulties in the governing. It is also obvious that there 
 was great risk of delivering the unburnt mixture into the 
 
 FIG. 57. Tbe Clerk Two-cycle Engine. 
 
 exhaust ports, and Mr. Clerk states that in the smaller 
 engines this could hardly be avoided. 
 
 ! In 1890 Mr. Clerk built another engine, which is shown 
 diagrammatically in fig. 57. Here M is the motor piston, 
 P the charging and expelling piston. The link work is so 
 arranged as to give a maximum speed of motion to one 
 piston whilst the other remains almost stationary. Thus, 
 when M uncovers the exhaust E, P travels rapidly towards 
 it and expels the burnt gases ; at the same time P is 
 drawing a fresh charge into the back of the cylinder. The 
 two pistons then travel back together, and the charge 
 passes from the back of piston P to the space between 
 the pistons. Upon ignition, M moves forward, whilst P 
 remains practically stationary. The inventor believes that 
 this engine is well adapted for both moderate and large 
 
84 THE EOBEY GAS ENGINE. 
 
 powers, and that it will give exceptionally economical 
 results when given a thorough trial. It seems hardly 
 probable that its mechanical efficiency will be high after 
 a short period of use, though the engine is no doubt very 
 effective in design from a thermal, if not from a mechanical 
 point of view. 
 
 Messrs. Robey and Co., of Lincoln, have for some years 
 manufactured an engine upon the Otto cycle, with improve- 
 ments introduced by Messrs. Richardson and Norris. An 
 important feature of the engine is the ease with which it 
 can be reversed. The cams are made to run in either 
 direction, so that the alteration can be effected in a few 
 moments. The speed is regulated by Richardson's high- 
 speed governor, similar to that used on other Robey 
 engines, fully described and illustrated in the Engineer of 
 October 14th, 1892. The governor is loaded partly by dead 
 weight and partly by springs, which latter can be regulated 
 whilst running to give the required speed. The governor 
 acts upon the gas valve, cutting off the supply of gas when 
 the speed increases. Many of the Robey engines are specially 
 designed for electric lighting, and for this purpose Mr. 
 Richardson has introduced an electrical governor. This, in 
 principle, consists of a solenoid wound as a shunt from the 
 current generated by the engine and dynamo. A soft iron 
 core is sucked into the solenoid as the voltage rises. Its 
 movement is connected by a bell-crank lever to the pecker 
 operating the gas valve, causing it to hit or miss the knife 
 edge on the gas-valve spindle. By this means the gas supply 
 may be entirely cut out. 
 
 Mr. Norris is now perfecting a two-cycle engine, with 
 the object of increasing the steadiness of running and 
 increasing the power derived from a given weight of engine. 
 The arrangement is briefly as follows : The front end of the 
 cylinder is closed, and acts as a pump for drawing in the 
 charge and expelling it to the combustion chamber at the 
 back of the cylinder. The piston rod carries a slight 
 enlargement upon it, which traverses a cylinder attached 
 
THE NORRIS GAS ENGINE. 85 
 
 to the front cover. This small cylinder is used to pump in 
 the gas. 
 
 The cycle of this engine is carried out in the following 
 order : The return stroke of the piston draws air into the 
 front end of the cylinder through an automatic valve, 
 whilst gas is drawn into the small cylinder attached to the 
 inside of the front cover. The next outstroke compresses 
 the air into an intermediate chamber at about 51b. per 
 square inch above atmospheric pressure, until the piston 
 has completed nine-tenths of its outward stroke. At this 
 point a port in the engine piston connects the combustion 
 chamber with the receiver. The exhaust valve is opened 
 a little before nine-tenths of the outward stroke, and a 
 column of exhaust gases is therefore already in motion by 
 the time air enters the combustion chamber from the 
 receiver. Air continues to flow from the receiver through 
 the combustion chamber until one-tenth of the inward 
 stroke is complete, when the exhaust valve closes. At this 
 time pure air only occupies nine- tenths of the cylinder 
 volume. After the air and exhaust valves are closed, gas 
 from the small pump is admitted to form an explosive 
 mixture, which is compressed during the remainder of the 
 instroke. The compression, in the opinion of Mr. Norris, 
 should not exceed 95 Ib. per square inch. 
 
 In larger sizes of this engine a differential piston is used, 
 in order to increase the volume of air pumped into the 
 receiver. 
 
FRENCH ENGINES. 
 
 CHAPTER VIII. 
 
 FRENCH ENGINES. 
 SIMPLEX OAS ENGINE. 
 
 THE Simplex gas engine was patented by Messrs. Edward 
 Delamare and Malandin, in 1884. Whilst resembling in 
 many respects the Otto engine, this motor has several 
 distinctive features. It works upon the usual Otto cycle, 
 but the charge is fired just after the commencement of the 
 outstroke of the piston, instead of, as usual, on the dead 
 centre. It is stated that this modification greatly reduces 
 shook upon the working parts. The ignition of the charge 
 is etlected by means of an electric spark produced by a 
 battery and induction coil. The ilifliculty usually met with 
 in electric igniters is found to be duo to the inefficiency of 
 the wake and brake mechanism. This is obviated in the 
 Simplex engine by the generation of a continuous spark 
 within a specially-formed chamber. The entrance to this 
 chamber is controlled by a slide valve, and the charge in 
 the cylinder can only come in contact with the continuous 
 spark when the port through the slide opens com- 
 munication between the cylinder and sparking chamber. 
 Many of the continental engines fire by electricity, but this 
 method has been little used in this country. There are 
 points in favour of electric ignition ; but the maintenance 
 of the battery, the possibility of the sparking points 
 becoming coated with deposit, and the difficulty of 
 maintaining the insulation are serious objections, and it is 
 probable that as the hot tube is made more durable it will 
 entirely supersede electric ignition. 
 
 The governing of this engine is effected by the use of an 
 air cylinder and piston. The piston is stationary, and the 
 air cylinder moves backwards and forwards with the 
 igniting slide spindle. Air enters the cylinder through an 
 orifice regulated by a micrometer screw. If the speed of 
 
FRENCH ENGINES. 87 
 
 tho engine increases, the air in the cylinder is unduly 
 compressed, and so forces outwards a second piston, which is 
 in communication with the air cylinder. This latter piston 
 carries the knife edge for actuating the gas valve. The 
 micrometer screw can be so regulated as to cause the out- 
 ward motion of tho second piston at any desired maximum 
 speed. In this way the knife edge is moved out of line 
 with that on the gas-valve spindle, and the entire supply of 
 gas is cut off. On larger engines of the Simplex pattern a 
 form of pendulum governor has been adopted. 
 
 Tho self-starter fitted to the Simplex engine consists of 
 a three-way cock, suitably proportioned to admit a charge 
 of gas and air to the cylinder. The cock is opened when 
 the piston is at the commencement of its forward 
 expansion stroke, and when charged the current is turned 
 on. The high temperature of the electric spark renders 
 the ignition of a weak mixture certain. 
 
 LENOIK ENGINE. 
 
 This is another engine in which electric ignition is used 
 to fire the charge. The chief object of the designer was to 
 obtain a very high compression and temperature of the 
 charge before firing. This was effected by jacketing the 
 working part of the cylinder only. The compression 
 chamber was bolted to the cylinder casting, with an asbestos 
 joint to prevent the transmission of heat from the com- 
 pression chamber to the jacketed cylinder. The compression 
 chamber was cast with a series of deep ribs outside, to 
 distribute the heat by convection of air instead of by the 
 use of a water jacket. By this means the combustion 
 chamber was kept at a much higher temperature than the 
 working cylinder, and a weak mixture was easily ignited 
 by the electric spark. 
 
 CHARON ENGINE. 
 
 An attempt is made in this design of engine to expand 
 the charge beyond its volume at atmospheric pressure when 
 drawn into the cylinder. To effect this, the air admission 
 
88 GERMAN ENGINES. 
 
 valve remains open during part of the compression stroke, 
 and through it a portion of the charge is driven into a long 
 spiral tube open to the atmosphere, and of sufficient length 
 to retain the mixture forced into it. The valve then closes, 
 and the charge is compressed. After expansion and exhaust 
 the gases in the spiral pipe are drawn into the cylinder 
 with the new charge, and the cycle is repeated. The engine 
 is governed by reducing the gas supply and by allowing 
 more of the charge to enter the spiral pipe, thus reducing 
 the quality and volume of the gas ignited. 
 
 NIEL ENGINE. 
 
 In this engine the charge is admitted through a conical 
 revolving plug, which is kept tight during the compression 
 and explosion strokes by the pressure generated in the 
 cylinder, acting upon the back of the plug so as to force 
 it into its seating. 
 
 LALBIN ENGINE. 
 
 This engine is designed with the object of equalising the 
 turning effort upon the crank shaft. It consists of three 
 cylinders, one placed vertically with the crank shaft beneath 
 it, and one at each side inclined at an angle of 60 deg. with 
 the horizontal. Each cylinder works upon the Otto cycle ; 
 hence in two revolutions the crank shaft receives three 
 impulses, instead of only one. The mechanism is so arranged 
 as to permit of the engine being reversed. 
 
 GERMAN ENGINES. 
 
 BENZ ENGINE. 
 
 Of the German engines perhaps the Benz is one of the 
 most interesting from a constructional point of view, inas- 
 much as the cylinder is arranged to give an impulse every 
 revolution, without the aid of a separate large pump, such 
 as used upon the early gas engines of the Clerk type. The 
 engine is horizontal, and has electric ignition. The cylinder 
 is closed at both ends ; the back of the piston receives the 
 
GERMAN ENGINES. 89 
 
 impulse from the ignition of the mixture, and the front 
 of the piston drives, at each forward stroke, a volume of 
 fresh air through a suitable valve into an air reservoir 
 formed in the bed plate of the engine. A small plunger, 
 worked from the crosshead, acts as a gas pump, and the 
 plunger cylinder becomes filled with gas during every 
 forward stroke of the engine. It is possible with this 
 arrangement to produce an explosion at the commencement 
 of each outstroke of the piston. Imagine the gases to have 
 ignited and driven the piston forward. By this stroke the 
 air reservoir is replenished and the gas pump charged. On 
 the return stroke the exhaust valve opens, and just before 
 the end of the stroke the air from the reservoir is admitted 
 to the combustion chamber. This effectually clears out the 
 burnt gases and replaces them with fresh air, which at the 
 same time tends to cool the cylinder. The air and exhaust 
 valves then close, and the gas pump forces its charge into 
 the combustion chamber. The air and gas mingle sufficiently 
 to permit of ignition by means of the electric spark. The 
 consumption of gas in these engines is said to be about 23 
 cubic feet per indicated horse power. 
 
 KOERTING-LIECKFELDT ENGINE. 
 
 The first of these engines was designed with vertical 
 cylinders, and is specially interesting for the method of 
 ignition of the charge. By a special apparatus, shown in 
 fig. 58, the external flame F ignites the charge. The chamber 
 C is in direct communication with the motor cylinder during 
 the compression stroke and at the time of ignition. The 
 cylindrical piece B is free to move upwards under the 
 pressure of the mixture. The rod R is worked up and down 
 by a lever, thus opening and closing the port P opposite the 
 naked flame. When compression commences the rod R is 
 raised, so that the port P is open to the flame. The mixture 
 enters the chamber C, and raises the plug B. It is then 
 forced through the small entrance of the cone, and, expand- 
 ing, arrives at P at nearly atmospheric pressure, where it is 
 
GERMAN ENGINES. 
 
 ignited by the flame F. The flame then travels a little way 
 down towards the small end of the cone, but cannot get 
 very far, because the velocity of the outcoming gas is greater 
 at the lower end of the cone than the rate of flame propaga- 
 tion. The rod R descends at the end of the instroke of the 
 piston, and, covering the port P, effectually stops the flow of 
 gas ; the pressure within the cone becomes equalised, the 
 
 FIG. 58. Koercing-Lieckfeldt Igniter. 
 
 piece B drops, and the flame travels downwards towards the 
 chamber C, and so ignites the entire charge. 
 
 The governing is effected by a somewhat novel arrange- 
 ment of the centrifugal type, which cuts out the gas and air 
 supply, and at the same time keeps the exhaust open, so 
 that the products of combustion are drawn back into the 
 cylinder. This method is advantageous in oil motors, inas- 
 much as it helps to maintain the temperature of the 
 combustion chamber ; but it is a doubtful practice in connec- 
 tion with gas engines. 
 
DAIMLER GAS ENGINE. 91 
 
 THE DAIMLER GAS ENGINE. 
 
 This engine was first exhibited in 1889, and is specially 
 noticeable for its high speed and simplicity of parts. The 
 later designs have two cylinders, placed nearly vertically 
 over one crank shaft. The connecting rods work upon one 
 crank pin, set between heavy balanced crank webs. The 
 connecting rods and crank are entirely encased in an air- 
 tight covering, in a similar way to the engine shown in 
 fig. 53. An automatic air valve admits air only to this 
 casing as the pistons rise. The gas and exhaust valves are 
 placed at the upper end of each cylinder, the gas valve 
 being- automatic, whilst the exhaust is lifted by the equiva- 
 lent of a cam on the disc crank web. Instead of the air 
 port P, shown in fig. o3, a valve is placed in the centre of 
 each piston, for passing the air from the crank chamber to 
 the combustion chamber. The action of the engine is as 
 follows : Upon the down stroke one cylinder expands its 
 burning charge, whilst the other draws in gas, and at the 
 same time passes air from the crank casing into the work- 
 ing cylinder through the valve in the piston. On the 
 return stroke the one cylinder exhausts, whilst the other 
 compresses. In this way an impulse is received every 
 revolution, although each individual cylinder works upon 
 the Otto cycle. The governing is effected by holding open 
 the exhaust valve every revolution when the speed rises, 
 thus preventing the automatic gas valve from being opened 
 by the formation of a slight vacuum in the cylinder. This 
 engine is worked entirely without jacket water. This is no 
 doubt possible on account of the high speed at which it is 
 run (from 500 to 700 revolutions per minute), coupled with 
 the fact that only very small powers are attempted. It is 
 claimed that the engine will consume about 33 cubic feet 
 of gas per hour per I.H.P., and when it is remembered that 
 the engines only range from 1 to 2 horse power, this figure 
 is not very excessive. 
 
92 TESTING GAS ENGINES. 
 
 THE CAPITAINE GAS ENGINE. 
 
 The Capitaine gas engine is designed to give a high speed 
 of revolution with a nominal piston speed. This is done 
 by increasing the diameter of the cylinder, whilst decreasing 
 the length of stroke. The combustible mixture is drawn 
 into the cylinder through an enlarging entrance, so that its 
 velocity is decreased, the object being to prevent the 
 mingling of the new charge with the exhaust gases. How 
 far this arrangement is effectual it is difficult to estimate, 
 though it is certain that the time of combustion is 
 lengthened by the presence of burnt products in a rich 
 mixture of, say, eight volumes of air to one of gas. The 
 ignition tube in the Capitaine engine is of porcelain, and is 
 claimed to be very durable and efficient. 
 
 CHAPTER IX. 
 
 TESTING GAS ENGINES. 
 
 THE testing of gas engines specially fitted up for experi- 
 mental work is a comparatively easy task, and involves but 
 little forethought from the experimenter ; when, however, 
 complete tests are to be made by the use of improptu 
 apparatus, there is ample scope for the ready exercise of 
 one's wits in arranging simple and reliable means of 
 measuring the quantities to be dealt with. In what follows 
 we shall confine our attention to testing a gas engine alone 
 that is, not in conjunction with the machinery it may be 
 driving. There may be cases in which the work done by 
 the driven machine is easily ascertained. The horse power 
 obtained from the terminals of a dynamo driven by a gas 
 engine affords a measure of the useful work done, and, 
 making allowance for the efficiency of the dynamo, gives the 
 work transmitted from the engine flywheel to the dynamo 
 pulley. The work done in pumping a certain weight of 
 
TESTING GAS ENGINES. 
 
 93 
 
 water against a known head, in a given time, is easily 
 converted into useful horse power. But in all such cases 
 the efficiency of the machines has to be allowed for, and as 
 this may be a rather uncertain quantity, the only reliable 
 method of testing is to run the engine separately, and to 
 measure the power derived from the flywheels by suitable 
 means. 
 
 The simplest and most reliable form of absorption dynamo- 
 meter is illustrated in fig. 59. It consists of two or more 
 
 FIG. 59 -Friction Brake. 
 
 lengths of good hemp rope placed over the upper semi- 
 circumference of the flywheel, and guided by wood blocks. 
 The ends of the ropes are spliced together so that a spring 
 balance may be attached at one side of the wheel, whilst a 
 hook for carrying weights may be attached to the other 
 side. The spring balance is anchored to the ground by 
 
94 TESTING GAS ENGINES. 
 
 means of a hook and adjustable screw. The latter is of 
 great convenience, for without means of adjustment the 
 ropes often stretch and allow the weights to touch the 
 ground. The weights for loading the brake should be cast 
 with slots in, so that they may be placed upon the hook. 
 When putting on the weights it is important that the slots 
 should not coincide, for if they do the vibration will 
 invariably cause the weights to fall over. 
 
 As a precaution against accident, the bar shown at B, fig 
 59, should be fixed to a suitable place and pass between the* 
 brake ropes. This prevents the undue lifting of the weights 
 should the friction increase, and has often been the means of 
 preventing a serious accident. 
 
 When absorbing large powers the rim of the wheel will 
 become much heated, and its undue expansion may lead 
 to fracture. It is therefore better to cast the rim in the 
 form of a trough (I I), and arrange for the delivery of cold 
 water to the rim. When the wheel is revolving the water 
 will follow it in the direction of rotation and completely 
 line the trough ; centrifugal force, acting upon the water, 
 will prevent its displacement from the trough. In order to 
 change the water in the rim the pipe shown at P, fig. 59, is 
 fixed so as to skim the surface and drain away the water 
 so caught. In this way a constant circulation may be 
 maintained. 
 
 It is important that the brake weights be removed when 
 the engine is being stopped, for if not the spring balance 
 will be broken by their sudden fall. Another arrangement 
 of the brake ropes is shown in fig. 60. This is more suitable 
 for large powers, but whether one form or other is adopted 
 is largely decided by convenience of arrangement, though 
 for very small powers this latter form of brake is unsuitable. 
 
 When an economy trial is to be made the brake should be 
 loaded, with the gas cock fully open, until the engine runs 
 at its nominal speed, and fires every cycle with a full charge 
 of gas. To adjust the brake and other measuring gear, a 
 preliminary trial is invariably necessary. 
 
TESTING GAS ENGINES. 
 
 95 
 
 In calculating the load upon the brake, the weight of the 
 hook supporting the load should, of course, be included, and 
 allowance should be made for any unbalanced part of the 
 brake rope. The spring balance should always be tested 
 before use, as the spring often gets a slight permanent set, 
 and the index does not usually read zero when the balance 
 is unloaded. It is of the highest importance that the 
 weights hang freely, other wise serious errors may arise. 
 Although these remarks deal with details apparently trivial, 
 we cannot too strongly urge the necessity of their strict 
 
 FIQ. 60. Friction Brake. 
 
 observance. A successful trial depends entirely upon the 
 immediate observation of any disturbance of the measuring 
 apparatus, and this can only be detected by constant 
 watching. 
 
96 TESTING GAS ENGINES. 
 
 During the trial the reading of the spring balance should 
 be taken every five or ten minutes, according to the duration 
 of the test, and from these an average may be calculated. 
 
 Let S = the mean spring balance reading in pounds ; 
 W = the total weight hanging upon the brake in 
 
 pounds ; 
 II = effective radius of brake wheel in feet, 
 
 = radius of brake wheel + radius of brake rope ; 
 N = revolutions of wheel per minute ; 
 
 then the brake horse power 
 
 For the same brake gear the factors 
 
 2?r R 
 
 33000 
 
 remain constant ; hence they may be worked out once for all, 
 and the result be booked as the brake constant = C. The 
 formula for the B.H.P. may then be written 
 
 B.HP. = (W - S)NC. 
 
 INDICATING. 
 
 Indicator diagrams obtained from gas engines cannot 
 be entirely relied upon, though they may, with careful 
 manipulation and a suitable instrument, be considered 
 accurate enough for practical purposes. Before any impor- 
 tant trial is to be made the indicator springs should 
 themselves be tested under the conditions of temperature 
 likely to occur. The average difference in the deflections 
 of indicator springs when hot and cold amounts to as much 
 as 5 per cent, and more in some cases ; a cold test is there- 
 fore no guarantee. In the case of a gas engine, a simple 
 method of testing the springs is as follows : Construct a 
 stirrup of brass, as shown in fig. 61, having a hole in it 
 sufficiently large to allow the indicator piston rod to pass 
 
TESTING GAS ENGINES. 
 
 97 
 
 through it. This stirrup is placed in position by disconnect- 
 ing the links attached to the pencil lever Next obtain a 
 100 Ib. spring balance, and carefully test it, with the stirrup 
 attached, against standard weights, noting the plus or 
 minus errors, if any. Then pass the cord attached to the 
 
 FIG. 61. Arrangement of Indicator for Testing Gas Engines. 
 
 upper end of the balance over a suitable pulley, as shown. 
 Weights may then be placed upon a scale pan at the 
 end of the cord, and adjusted until the balance reads various 
 definite loads. The use of an accurate balance can hardly be 
 dispensed with, as otherwise the friction of the pulley will 
 give an error of probably 5 per cent. By the arrangement 
 
 8G 
 
98 TESTING GAS ENGINES. 
 
 shown this is eliminated. Suppose a spring is being tested 
 which is supposed to register 100 Ib. per square inch for 
 every one inch vertical movement of the pencil on the drum. 
 Put the indicator on the engine, and arrange the tackle 
 above it as shown in fig. 61. Start the engine, and whilst 
 running slightly turn on the indicator cock so as to heat up 
 the spring. When the indicator is at its working tempera- 
 ture, close the indicator cock and mark off the atmospheric 
 line on the drum, revolving the latter by hand. Now load 
 the balance until it reads 10 Ib., and mark off the corres- 
 ponding height of the indicator pencil. Proceed in this 
 way by increments of 10 Ib. at a time until the limit of the 
 instrument is reached. Carefully measure the diameter of 
 the indicator piston and obtain the area. In our example, 
 suppose the area of the piston to be 0'5 square inch, and 
 suppose when the balance reads 50 Ib. the movement of 
 the pencil above the atmospheric line measures 1*03 in. 
 The pressure per square inch at 1'03 in. rise 
 
 = 100 x 1'03 = 103 Ib. per square inch. 
 
 But the actual pressure upon an area of 05 square inch 
 is 50 Ib., which is equivalent to 
 
 ^ = 100 Ib. per square inch. 
 O'o 
 
 We see, therefore, that the indicator is recording a pressure 
 3 per cent in excess of the real pressure acting upon the 
 piston. 
 
 When testing springs by this method, care must be taken 
 not to overload the instrument, for it must be remembered 
 that the whole stress is transmitted through the small 
 piston rod of the indicator. 
 
INDICATORS FOR GAS ENGINES. 
 
 99 
 
 CHAPTER X. 
 
 INDICATORS FOR GAS ENGINES. 
 CROSBY INDICATOR. 
 
 THIS instrument, shown in figs. 62 and 63, has a well-earned 
 reputation of being both durable and accurate at high 
 speeds. The heavy stresses brought to bear upon the 
 working parts of the instrument when used for gas-engine 
 indicating has led to modifications in design, rendering it 
 in every way more adaptable to gas-engine work. The 
 most obvious alteration is the adoption of the straight link 
 in the parallel motion, in place of a longer and curved link 
 attached to the extreme end of the pencil bar. The indicator 
 shown in fig. 63 is specially designed for the gas engine, but 
 
 Fio. 62. Crosby Indicator. 
 
 it can be used with equal efficiency upon a steam-engine 
 cylinder. It will be observed that the barrel of the 
 indicator is bored out to two diameters, the larger part 
 
100 INDICATORS FOR GAS ENGINES. 
 
 having an area equal to twice that of the smaller. Two 
 pistons are supplied, to either of which the same spring 
 may be fitted. In this way a spring marked 100 will 
 register 100 Ib. per square inch for each inch of vertical 
 movement of the pencil when the spring is fitted on the 
 larger piston. If, however, this spring is fitted to the 
 smaller piston, as recommended for gas-engine work, then 
 
 FIG. 63. Crosby Indicator. 
 I 
 
 lin. vertical movement of the pencil corresponds to 200 Ib. 
 pressure per square inch. With this instrument a variety 
 of pressures may be dealt with by a small assortment of 
 springs, a feature which readily recommends itself to the 
 economist. In both designs the working cylinder is jacketed, 
 to enable it to expand readily with the piston, and so 
 obviate undue friction. All the piston springs are double 
 spiral, to prevent the slight rotation and tilting of the 
 piston which is inevitable with a single spiral. A very 
 light pencil bar is used upon these indicators for steam- 
 engine work, but for gas-engine work a bar of L section ia 
 necessary to withstand the sudden jar due to the explosion. 
 
INDICATORS FOR GAS ENGINES 
 
 101 
 
 FIG. 64. Section of Tabor Indicator. 
 
102 INDICATORS FOR GAS ENGINES. 
 
 The general construction of the indicators will be gathered 
 from the illustrations. 
 
 THE TABOR INDICATOR. 
 
 The Tabor indicator is shown in fig. 64. Its three most 
 characteristic features are the straight-line motion, the 
 working cylinder, and the involute spring used to actuate 
 the drum. The vertical movement of the marking point is 
 secured by the curved slot, in which runs a small roller 
 attached to the pencil bar. The cylinder in which the 
 indicator piston works consists of a thin tube att ohed to 
 the body of the indicator at its lower end only. The air 
 jacket around the tube admits of free expansion of the tube 
 when in use, and the thin walls of the latter assist in 
 equalising the temperature. These indicators are some- 
 what larger than the Crosby, and will give diagrams 5 in. 
 long at slow speeds, though when working at 200 revolutions 
 per minute the makers recommend a 4 in. diagram, reducing 
 it to 2 in. at 500 revolutions in order to minimise the effects 
 of inertia of the drum. 
 
 The indicator illustrated in fig. 65 requires no reducing 
 motion as usually fitted. It will be seen that the worm R 
 engages with a worm wheel upon the drum B. The cord 
 from the pulley O is attached to the crosshead of the 
 engine, and travels through its entire stroke. Various 
 lengths of diagrams are obtained by altering the size of the 
 pulley O. The cage d contains the spring which gives 
 the return motion to the pulley. The backlash of the worm 
 gear is taken up by a spring within the drum B. By a 
 simple movement of the milled head, shown at u, the pulley 
 O disengages with the worm, and revolves freely upon its 
 spindle. The drum B then remains stationary. This gear 
 is very convenient for indicating at high speeds ; but it is 
 obvious that the speed of the engine must be considerably 
 reduced to enable the hook from the indicator pulley to be 
 attached to the crosshead. 
 
 The only automatic gear for attaching and detaching 
 
INDICATORS FOR GAS ENGINES. 103 
 
 the reducing motion from the engine is that now being 
 introduced by Messrs. Crosby and Co. This gear, which 
 we shall describe, enables the operator to attach a complete 
 
 FIG. 65. Tabor Indicator with Reducing Gear. 
 
 indicator gear to any high-speed engine whilst running at 
 full speed Another important feature is that the reducing 
 motion is thrown out of action immediately after the 
 diagram has been taken. 
 
 THE WAYNE INDICATOR. 
 
 The design of this instrument, which is made by Messrs. 
 Elliott Brothers, is a complete departure from the usual 
 lines. Two views are shown in figs. 66 and 67, from which 
 
104 
 
 INDICATORS FOR GAS ENGINES. 
 
 fche general appearance may be gathered. It will be seen 
 fchat the paper upon which the diagram is traced is fixed 
 by two spring clips to an aluminium guide, and forms a 
 
 FIG 66. Wayne Indicator. 
 
 concave surface. The latter moves horizontally by means 
 of a cord passing over a spring pulley and attached to any 
 convenient reducing gear. The marking point traverses 
 
INDICATORS FOR GAS ENGINES. 
 
 105 
 
 the arc of a circle, its movement being resisted by the 
 torsion of a spiral spring shown on the right of the 
 illustrations. This spring may be changed to si 
 
 OF TRT- 
 
 UNIVERSITY 
 
 FIG. 67. Wayne Indicator. 
 
 pressures dealt with, and determines the scale of the 
 pressure ordinates. The circular movement of the pencil 
 is produced by the pressure of the steam upOD two steel 
 tongues a, fig. 68, projecting from a horizontal rod passing 
 
106 
 
 INDICATORS FOR GAS ENGINES. 
 
 through the cylinder. This rod carries the pencil at one 
 end and the spiral spring at the other. The diagram- 
 matic sketch, fig. C8, will show how the pressure acts upon 
 the steel tongues, and how the steam which passes the 
 tongues escapes through the holes e, e. The chief features 
 in this indicator are : The absence of joints or parallel 
 motion, small movement of working parts, and accessibility 
 
 a 
 
 FIG. 68. Sectional Elevation of the Wayne Indicator. 
 
 of the pressure spring. For indicating engines at very 
 high speeds, such as 600 revolutions per minute, an 
 ingenious fitting is added to the instrument, termed by the 
 makers "the lining attachment." By this gear, shown in 
 fig. 67, the indicator diagram is built up line by line, the 
 movement of the pencil for each part added being about 
 one-twentieth of an inch. By turning the handle shown in 
 the front of fig. 67 a worm is rotated. This gears with a 
 segment of a worm wheel, which in turn rotates in a con- 
 centric circle with the cylinder. A radial arm, visible in 
 the illustration, fits into a hole in the piston rod, and is 
 attached to the worm-wheel segment by means of a screw 
 passing through a slotted hole. This slotted hole enables 
 
INDICATORS FOR GAS ENGINES. 
 
 107 
 
 the pencil to move a distance of | in. when the worm wheel 
 is stationary. The method of taking a " line" diagram is 
 as follows : First turn the handle operating the worm and 
 worm segment until the slot in the radial arm stands quite 
 free of the set-screw passing through it. This enables the 
 atmospheric line to be accurately taken. Then turn on 
 the indicator cock, and commence to rotate the handle 
 before mentioned. The distance between the horizontal 
 lines of the diagram will depend entirely upon the loss of 
 time occasioned by the slot in the radial arm, and conse- 
 quently upon the speed of rotation of the handle operating 
 the worm-wheel segment. For any given case a few trials 
 will enable the operator to accommodate the movement of 
 the handle to produce a distance of about one-twentieth 
 of an inch between the horizontal lines, as recommended by 
 the makers. A line diagram is shown in fig. 89. It is 
 
 FIG. 69. Diagram taken with Wayne Indicator. 
 
 noticeable that a weaker spring may be used with the 
 line attachment ; at the same time a much more accurate 
 diagram is produced. The author has found the lining 
 attachment extremely efficient and simple to work. It may 
 be attached or removed from the instrument in a few 
 
108 
 
 INDICATORS FOR GAS ENGINES 
 
 seconds, and certainly produces a most perfect diagram at 
 all speeds. The accessibility of the pressure spring is a 
 great convenience, and its uniform temperature should 
 
 FIG. 70. Simplex Indicator. 
 
 increase the degree of accuracy obtained. For convenience 
 in adjusting the instrument it is attached to the cock with 
 a swivel joint. 
 
REDUCING GEAR FOR GAS ENGINES. 109 
 
 THE SIMPLEX INDICATOR. 
 
 A photograph of this instrument is shown in fig. 70. The 
 instrument is made by Messrs. Elliott Brothers, and is 
 exceedingly strong. The chief departure from the usual 
 construction is in the arrangement and form of spring used 
 to record the pressures. This spring is visible at S. The 
 upper limb is held in the upright standard, which is rigidly 
 attached to the body of the instrument. The lower limb is 
 held in the top of the piston rod of the indicator. When 
 pressure is exerted upon the piston, its movement is resisted 
 by the spring. The spring is quite easily removed by a 
 slight pressure upon its side. Various strengths of springs 
 are supplied. 
 
 The pencil levers, which are somewhat heavy, pass 
 through a wide slot cut in the upper part of the instrument, 
 and are attached to the piston rod by a collar, which is free 
 to rotate upon the piston rod. The straight line motion is 
 obtained by four parallel bars, three of which are visible in 
 the photograph. 
 
 CHAPTER XL 
 
 REDUCING GEAR FOR G-AS ENGINES. 
 
 IT is invariably necessary to attach to an engine some 
 mechanism to cause the drum of the indicator to reproduce 
 the motion of the piston to a small scale. Such mechanisms 
 are numerous, and for a complete description of the various 
 types the reader is referred to text-books on indicating. 
 The essential requirements of a reducing gear are simplicity, 
 durability, and accuracy. The great fault of nearly . all 
 reducing gears arises from the necessity of stopping the 
 engine to attach the gear. This also necessitates the gear 
 running until the engine may be stopped. The wear of the 
 
110 
 
 REDUCING GEAR FOR GAS ENGINES. 
 
 parts tkus becomes abnormal, causing considerable shake 
 and inconvenience, especially in high-speed engines. 
 
 Motion for the gear is usually derived from the crosshead 
 of a steam engine or trunk piston of a gas engine. A con- 
 venient attachment for the latter is illustrated in fig. 71, 
 at X ; the pipe shown is welded to the adjusting screw of 
 the piston pin, and carries the upright bar Y. The lock 
 nuts shown serve to adjust the position of the bar. The 
 motion of Y may be reduced to suit the stroke of the 
 indicator drum in a variety of ways, and a practical 
 
 Fio. 71. General view of Grover's Indicator Gear. 
 
 engineer will often be able to devise means suited to 
 individual cases. It is, however, almost essential, to ensure 
 economical working, to ha ye a gas engine tested periodically 
 in order to detect defective valves, and in view of this it is 
 convenient to be able to fit up an entire indicator gear 
 without stopping the engine. In order to secure this 
 condition the author has recently designed and provision- 
 ally protected the arrangement illustrated in figs. 71, 72, 73. 
 A number of these gears are being made by Messrs. Crosby 
 and Co. , for use in connection with their roll reducing gear 
 and improved indicator. 
 
 Referring to fig. 72, it will be seen that the cord from the 
 reducing wheel W passes through a tube, to which a handle 
 
GROVER'S INDICATOR GEAR. 
 
 Ill 
 
112 GROVER'S INDICATOR GEAR. 
 
 is attached. The tube slides in a boss upon the rocking 
 lever B, and is pressed outwards by the spiral spring shown. 
 When the spring is compressed, the bolt fitted in the handle 
 rides up the incline, and, dropping over the edge, prevents 
 the release of the spring. By pushing the handle over, as 
 indicated by the arrow in fig. 73, the spring is released and 
 the bolt in the handle runs up an opposite incline, dropping 
 over its edge, as before. The handle must be pulled back 
 again before the spring can be compressed. Upon the front 
 of the tube a distance piece D is pivoted. This part drops 
 downward by its own weight, but may occupy the horizontal 
 position shown dotted. 
 
 A coiled spring inside the reducing wheel W causes the 
 cord R to be in tension ; hence the hook H, resting against 
 the front of the handle, tilts the lever B against the stop Si. 
 When the hook is drawn forward, B rests against the 
 stop S 2 . 
 
 An important part of the arrangement is the swinging 
 link L attached to the bar Y, figs. 71 and 72. This link, by 
 reason of its inertia, swings from an angle of about 45 deg. 
 into a horizontal position at the end of each instroke. By 
 this means it is made to engage with the hook H when the 
 latter is in position for coupling up. 
 
 The only permanent attachment to the engine is the bar Y, 
 with its link L, and in the hands of a skilful operator the 
 whole of the preliminary adjustments of the gear may be 
 made even when running at full speed, though it is advisable 
 to keep a record of the length of hook required for each 
 engine, and to make such observations as will assist in 
 quickly replacing the gear when once adjusted. 
 
 We will suppose that our cord from the indicator drum is 
 attached to the reducing wheel, and that all is ready for 
 taking a diagram. The first thing to be done is to compress 
 the spiral spring. If this is omitted the gear may couple up, 
 but it will immediately uncouple itself on the return stroke. 
 Next, draw forwards and upwards the hook H, at the same 
 time lifting the distance piece D until it is in the dotted 
 
GROVER'S INDICATOR GEAR. 
 
 113 
 
 position, fig. 72. Now allow the hook H to return until it 
 presses against the distance piece D. It will now be found 
 that the distance piece and the hook remain in line, but the 
 hook is tilted upwards because the tension of the cord pulls 
 B against the stop S lt The gear is. now ready for coupling, 
 and this is effected by bringing B gently against the stop S 2 . 
 Immediately the hook leaves the distance piece, the latter 
 
 Fia. 73. End view of parts A, 13, and C of Grover's Indicator Gear. 
 
 drops down by its own weight, and when next the hock H 
 returns there is a quarter of an inch clearance between it 
 and the face of the tube. The hook therefore remains coupled 
 to the link until the diagram is taken. 
 
 To uncouple the gear it is merely necessary to rotate the 
 handle in the direction of the arrow on fig. 73. The spiral 
 spring is then released, the handle flies out, and occupies the 
 position shown in fig. 72. It is obvious that the hook 
 H will now collide with the handle, and its further inward 
 movement will be arrested. Hence the tension of the cord 
 
 9G 
 
114 GAS-ENGINE TRIALS. 
 
 R will again tend to bring B against the stop S^ In con- 
 sequence, the hook H tends to lift from the link L, and 
 when L completes its inward stroke the clearance is so 
 arranged that H is free to tilt upwards, and thus auto- 
 matically uncouple itself. 
 
 CHAPTER XII. 
 
 GAS-ENGINE TRIALS. 
 
 WHEN taking indicator diagrams from a gas engine, it is 
 advisable to keep the pencil upon the diagram paper for 
 about half a minute, in order that a series of diagrams may 
 be taken. If the engine is running at full power, the pencil 
 will traverse the same path during each cycle and trace one 
 distinct diagram. If, however, the engine is running lightly 
 loaded, it is possible that it may not receive a full charge of 
 gas, owing to the action of the governor ; hence the pencil 
 will trace out a fresh diagram for each cycle. Methods of 
 working out the cards will be given later ; at present we 
 shall treat only of the practical details of a test ; suffice it 
 to say here that the object of the experimenter should be to 
 obtain a set of diagrams at each operation of the indicator, 
 ranging from a maximum to a minimum area. 
 
 The springs used for indicating gas engines are necessarily 
 heavy (from TTO to ^) ; hence only the compression, 
 explosion, and expansion curves are clearly shown upon the 
 diagram. It is nevertheless important that the small 
 variations in pressure during the exhaust and suction 
 strokes should be ascertained as a check against bad valve 
 setting and undue throttling in the pipes. This is usually 
 done by using a light spring in the indicator (say a th). 
 In order to prevent the excessive compression of the spring 
 during the explosion it is usual to pass a in. length of A in. 
 or in. brass pipe over the indicator piston rod before 
 
GAS-ENGINE TRIALS. 115 
 
 putting the instrument together for use. This acts as a 
 distance piece between the piston and top cover of the 
 indicator cylinder, and thus prevents the spring from being 
 crushed. The length of the brass pipe will depend upon 
 circumstances. In the Crosby indicator a A in. length suffices. 
 A diagram taken in this way has been given in fig. 17, and 
 is explained in the text relating thereto. The excessive 
 throttling of a silencing chamber, exhaust pipe, or valves is 
 at once detected by a light spring diagram, and no engine 
 should escape this test after it is completely fitted up. 
 
 Measurements of the Jacket Water In a gas-engine trial 
 there are three observations to be made with respect to the 
 jacket water: 
 
 1. The quantity (measured in pounds). 
 
 2. The temperature of the inlet (measured in deg. Fah.). 
 
 3. The temperature of the outlet (measured in deg. Fah ). 
 The Quantity. The most convenient method of measuring 
 
 the weight of jacket water depends so much upon local 
 conditions that no specific advice may be given. The author 
 has found the following method convenient, and of wide 
 application : Most engines not specially fitted for testing 
 will be arranged as in fig. 2. This being so, empty the 
 tank B by syphon or otherwise, and plug the orifice of the 
 upper circulating pipe leading into tank A. No circulation 
 through the jacket is now possible without opening the 
 feed cock to A. The feed water delivered to A will pass 
 through the jacket and find its way into tank B. This 
 tank may be calibrated to read pounds of water, either by 
 means of a float or by the insertion of a glass tube passing 
 up the outside of the tank. In some instances it has been 
 less trouble to uncouple a union on the pipe e and conduct 
 the water to a tank resting upon a weighing machine. In 
 either case the quantity should be regulated by the feed 
 pipe to A, until the temperature of the outlet water is 
 about 150 deg. Fah. Observations of the weight of water 
 may be taken every ten minutes, though with a steady flow 
 
116 GAS-ENGINE TKIALS. 
 
 this is not necessary. When short handed only the total 
 weight at the end of the trial need be observed. 
 
 Jacket Temperature^ Inlet. The inlet temperatures should 
 be taken on the inlet pipe near the tank A. If convenient, 
 insert a _L piece in the inlet pipe, and through a cork in the 
 _L place a Fahrenheit thermometer vertically, so that the 
 mercury bulb is directly in the stream of water. Many 
 errors arise through inattention to this. If the vertical 
 part of the J_ is long, a small volume of stagnant water 
 remains in it and gathers heat from the engine. The ther- 
 mometer readings may be affected by this, and become 
 altogether unreliable. In most cases the inlet temperatures 
 may be observed from the water in the tank A, thus 
 obviating any alteration to the pipes. 
 
 The Outlet Temperature. This should be taken as near 
 the exit from the jacket as possible. A convenient place is 
 in the bend at the top of the pipe e. The thermometer 
 should be placed vertically, especially if permanently fitted, 
 as, if not, the column of mercury tends to stick at the 
 maximum temperature recorded. It has been stated above 
 that the usual outlet temperature is about 150 deg. Fah. 
 The thermal efficiency of an engine may be increased by 
 raising the temperature of the jacket. There is therefore 
 a tendency to raise the temperature of the jacket beyond 
 its safe working limit when running short economy trials. 
 An increase in efficiency acquired in this way is sometimes 
 ignorantly ascribed to other causes, and fair comparisons 
 cannot be made without regard to the temperature of the 
 jacket outlet water. No doubt a time will come when the 
 jacket, as now usually fitted, will be dispensed with ; for, 
 although the gas engine is theoretically and practically 
 the most efficient heat engine, it is nevertheless a humilia- 
 ting admission that provision must be made for wasting 
 between 30 and 40 per cent of the heat generated. Where 
 hot water is required in factories there is no reason why 
 this heat should not be utilised by a convenient arrange- 
 ment of service pipes. The author has arranged a plant 
 
GAS-ENGINE TRIALS. 117 
 
 to work upon this system, which will effect a considerable 
 saving in fuel, and be otherwise extremely convenient. 
 
 Measurement of the Gas. The number of cubic feet of 
 gas used, also the temperature and pressure of the gas in the 
 meter, require to be known. The index of an ordinary gas 
 meter is usually provided with a set of three or four 
 pointers indicating respectively thousands, tens of thousands, 
 hundreds of thousands, &c., of cubic feet of gas. Besides 
 these, there is a smaller pointer placed above, which revolves 
 once for, say, 10, 20, or 30 cubic feet according to the size of 
 the meter. Besides these, another figure, which is frequently 
 misleading, will be found printed upon the dial. This 
 figure refers to the capacity of the measuring chamber 
 inside the meter, and this is frequently noted upon the dial 
 in the following way: "1'32 cub. ft. per rev." Mistakes 
 occur in supposing this figure, 1'32 cubic feet, to refer to 
 a revolution of the small pointer, instead of, as above 
 mentioned, to the capacity of the measuring chamber. It 
 should, therefore, always be remembered that this figure has 
 absolutely no reference to the pointers on the dial. 
 
 When the meter is situated in a room of practically 
 uniform temperature, the meter may be assumed to be of the 
 same temperature as the air in the room. It is, however, more 
 satisfactory to have a thermometer fitted to the exit pipe 
 from the meter, so that the temperature of the gas may be 
 more accurately ascertained. The pressure of the gas is best 
 ascertained by the application of a manometer, or U tube 
 filled with water, to some connection close to the meter. 
 The head of water sustained by the pressure of gas should 
 be noted a few times during the trial, to see that no 
 variation of pressure takes place. In connection with 
 the observation it is also necessary to take the height of 
 the barometer, so that all readings may afterwards be 
 reduced to one standard of atmospheric pressure. 
 
 Junker's Calorimeter. In very complete trials it is neces- 
 sary to know the average composition of the gas during the 
 trial, though in most tests the calorific value of the gas 
 
118 GAS-ENGINE TRIALS. 
 
 (which, for general purposes, is the only figure required) may 
 be ascertained by means of an instrument known as Junker's 
 calorimeter. This instrument, which may be obtained from 
 the sole agent, Herraon Kiihne, New Broad Street, E.G., is 
 illustrated in figs. 74 and 75. The principle of the instru- 
 ment is that the heat generated by a gas flame is absorbed 
 by a water jacket. The quantity of water, its inlet and 
 outlet temperatures, and the quantity of gas passing 
 through the instrument, afford data from which the calorific 
 value of the gas may be determined to within 0'5 per cent of 
 any other determination. Radiation is prevented by 
 surrounding the apparatus by an air jacket formed by a 
 nickel-plated cylinder. The flame shown at 28, fig. 74, is 
 surrounded by a water jacket, through which pass a number 
 of vertical copper tubes. The flame burns in the central 
 chamber, and the products of combustion pass down the 
 inside of the tubes to the outlet at 33. By this arrangement 
 the gases at the highest temperature meet the hottest 
 water, whilst as the gases cool they meet the colder inlet 
 water. This, of course, favours the transmission of heat, 
 and the products of combustion escape at the throttle at 
 atmospheric temperature, having parted with the heat of 
 combustion to the jacket water. In order to obtain 
 accurate results, it is necessary that the flow of water 
 through the instrument should be perfectly uniform. 
 This is secured by leading the water into a tank shown at 
 3, fig. 74. This tank is kej>t slightly overflowing, thus 
 producing a constant head. Two thermometers are fitted 
 for measuring the temperatures of the inlet and outlet, 
 and a graduated measuring glass is supplied for water 
 measurements. It is highly important that the gas 
 pressure remain constant, and for this reason a suitable 
 regulator is required to be used with the calorimeter. An 
 external view of the apparatus showing the gas meter and 
 regulator is given in fig. 75. Water formed by the com- 
 bustion in the central chamber is collected in a small 
 measuring glass at d, fig. 75. 
 
GAS ENGINE TRIALS. 
 
 119 
 
 FIG. 74. Section and Elevation of Junker's Calorimeter. 
 
120 GAS-ENGINE TRIALS. 
 
 The calibrations of the instrument have been made in the 
 metric system, and the unit of heat has been taken as the 
 quantity required to raise the temperature of 1 kilogramme 
 of water 1 deg. Cen. This unit is termed a calorie, and is 
 converted into British thermal units* by multiplication by 
 3-96. 
 
 In fixing the apparatus, the chief points to be observed 
 are as follow : 
 
 No draughts must be allowed to strike the outlet of the 
 exhaust gases. The quantity of gas passed through the 
 calorimeter should be at the rate of 
 
 4 to 8 cubic feet per hour of illuminating gas, 
 8 to 16 hydrogen, 
 
 16 to 32 Dowson gas. 
 
 Always light the burner outside the combustion chamber, 
 to avoid explosion due to accumulated gas. 
 
 Regulate the water passing through the jackets until its 
 rise of temperature is from 10 to 20 deg. Cen. 
 
 Having set up the apparatus and regulated the quantity 
 of gas and water, a test may be made in the following way : 
 Observe the reading of the gas meter at a convenient figure, 
 and at the same time remove the tube C, fig. 75, to discharge 
 into the graduated glass. Then take the readings of the inlet 
 and outlet thermometers. Take a few intermediate readings 
 of the outlet thermometer, and immediately the water 
 reaches, say, the 2 litre mark, turn off' the gas and read the 
 quantity of gas shown by the meter. The readings may be 
 booked as follow : 
 
 Cold water Hot water Water 
 
 Uusi ter. thermometer. thermometer. jacket. 
 
 4 cubic feet 
 4 268 cubic feet 
 
 7-98 
 
 5) 
 JJ 
 
 ... 27-54 ... 
 ... 27-52 ... 
 ... 27-53 ... 
 ... 27-55 ... 2 litres 
 
 Gas burnt 0'268 cubic feet 
 
 7'98 
 
 ... 27 '53 (mean) 2 litres 
 
 * The beat required to raise 1 lb. of water of maximum density 1 deg. Fah. is 
 equal to one British thermal unit. 
 
GAS ENGINE TKIALS. 
 
 Mean rise in temperature 
 
 = 2753 - 7'98 = 1955. 
 Calorific value of gas in calories per cubic foot 
 
 mean rise in temperature x litres of water 
 cubic feet gas burnt 
 
 121 
 
 = 145-5 x 3'9G = 577 B.T.U. 
 
 So far we have obtained a calorific value of the gas with- 
 out regard to any condensed water which may have collected 
 
 Fin. 75. External View of Junker's Calorimeter. 
 
122 GAS-ENGINE TRIALS. 
 
 in the small measuring glass marked D, in fig. 75. Now, this 
 moisture is the result of the combustion of hydrogen with 
 the oxygen of the air, and at the moment of combustion 
 steam is formed. In the calorimeter this steam is condensed, 
 and gives up its latent heat, to the calorimeter ; hence the 
 calorific value above obtained is the total heat evolved per 
 pound of gas when all the products of combustion are cooled 
 to atmospheric temperature. 
 
 In nearly all industrial processes the steam formed by the 
 combustion of hydrogen with oxygen passes away as steam ; 
 consequently some of the heat is not available unless the 
 steam is all condensed to water. The number of British 
 thermal units carried away per pound of hydrogen is, in 
 round numbers, 10,000. This is obtained in the following 
 way : 1 Ib. of hydrogen requires eight times its weight of 
 oxygen to completely burn it ; hence 9 Ib. of steam would be 
 produced. Suppose this to be cooled to water at 60 deg. 
 Fah., and at atmospheric pressure, the heat evolved by the 
 process would be equal to 9 x ] 966 + (212 - 60) f units per 
 pound of hydrogen, which amounts to 10,062 British 
 thermal units. It is evident, therefore, that two calorific 
 values may be obtained the one the total, the other the 
 available. In gas engines the heat of the exhaust does not 
 permit of condensation of the steam produced ; hence the 
 latent heat should be subtracted to give the available heat 
 for use in the gas engine. 
 
 For this determination it is advisable to continue the 
 operation of the calorimeter for a longer time, because of 
 the small quantity of condensed water produced. Suppose 
 the measuring glass D contains 60 cubic centimeters after 
 burning 3 cubic feet of gas ; then the correction in calories 
 
 _ cubic centimeters of water x 0'6 
 
 cubic feet gas burnt 
 60 x 0-6 
 
 3 
 = 12 calories. 
 
GAS-ENGINE TRIALS. 123 
 
 Then the net available heat 
 
 = 145'S - 12 = 133'8 calories 
 = 529 B.T.U. 
 
 It is sometimes argued that in estimating the efficiency of 
 a gas engine the total heat, and not the net available heat, 
 should be taken as the basis. Nearly all efficiency trials 
 have hitherto been worked upon the available calorific value 
 of the gas used, and for the sake of uniformity we shall 
 adopt this standard. After all, the question is merely one 
 of definition, and needs no further discussion here. 
 
 Method of Taking Sample of Gas. Should it be desired to 
 take a continuous sample of gas for future analysis, the 
 following arrangement will be found convenient : Fit an 
 indiarubber cork into a suitable glass jar, and through two 
 holes insert glass tubes. The one tube should be short, and 
 the other should reach to the bottom of the jar. Attach 
 the shorter tube to the gas supply by means of an india- 
 rubber pipe, and form a syphon of the other longer 
 tube by another piece of rubber pipe. Having completely 
 filled the apparatus with water, a pinch cock may be 
 adjusted on the syphon so that the latter draws the 
 water slowly from the bottle or jar. Gas then enters 
 the jar, and by suitably regulating the syphon small 
 quantities of gas may continuously enter the jar during the 
 whole period of the trial. In this way a fair representative 
 sample of gas is obtained for analysis. Mercury is, of course, 
 preferable to water, inasmuch as the constituents of coal 
 gas are more or less soluble in water. No serious errors 
 arise provided the gas is not left for a long period in contact 
 with a large surface of water. Precise instructions for 
 analysing gases will be given in another chapter. 
 
 Counting the Revolutions and Explosions. The revolutions 
 of the crank shaft are usually observed by means of some 
 form of speed counter. The following methods of actuating 
 the counter have been made use of by the author, and 
 are given here as suggestions. It often happens that 
 
124 GAS-ENGINE TRIALS. 
 
 incidental conditions adapt themselves to the requirements 
 of the experiments, and for this reason it is always well to 
 be on the alert rather than to attempt to laboriously carry 
 out a special method of fixing the measuring apparatus. 
 
 Suppose the counter to have a reciprocating lever which 
 requires to be moved through two short strokes at each 
 revolution of the crank shaft. When the reducing gear for 
 indicating is constantly working throughout the trial, 
 motion may be transmitted from some of its bars by a cord 
 attached to the lever of the counter. The cord may be held 
 in tension by attaching a strong indiarubber band to the 
 counter lever. Indiarubber bands are so extremely useful 
 in fixing up testing apparatus that they should be regarded 
 as part of the equipment for an engine trial. Should the 
 reducing gear be unsuitable for driving the counter, motion 
 may be derived from the valve levers. In this case it must 
 be borne in mind that the side shaft of an Otto cycle engine 
 is geared down to run at one-half the revolutions of the 
 crank shaft ; hence the counter readings must be doubled. 
 For the permanent attachment of a counter it is, of course, 
 better to employ a link or rod for operating the counter 
 lever, though the cord and indiarubber band is much more 
 easily fitted up, and is just as reliable for a three or four 
 hours' test. Small counters are procurable which may be 
 held against the centre of the crank shaft by hand. Some 
 are arranged to give the revolutions per minute on the dial, 
 others give the revolutions during the time the counter is 
 held in position. 
 
 The explosions are best counted from the exhaust pipe 
 outlet. If the engine is running at full power, and is 
 working on the Otto cycle, the explosions will, of course, 
 be equal to half the revolutions. The explosions should in 
 any case be counted occasionally, in order to verify the 
 assumption, as derangement of the governor gear and ignition 
 tube may falsify the results if not checked. 
 
 Temperature of the Exhaust. This measurement is diffi- 
 cult to obtain with any approach to accuracy. It is indeed 
 
GAS-ENGINE TRIALS. 125 
 
 often left to be calculated from the indicator card. This, 
 however, is not satisfactory, because of the alteration of the 
 specific heat of the gases forming the products of combustion. 
 It has been shown by Le Chatelier that the specific heat of 
 carbon-dioxide increases in the ratio of 1 to 1'6 between 
 temperatures of and 1,000 Cen. Until more is known of 
 the specific heats of gases at high temperatures, calculations 
 from the cards can only be regarded as approximating to 
 the results required. 
 
 Mr. Burstall has carried out experiments in this direction, 
 using a fine platinum wire suspended in the combustion 
 chamber. The electrical resistance was measured, and from 
 this the temperature deduced. This was found to range 
 from 1,045 deg. Cen. to 1,140 deg. Cen. At the present time 
 there is no simple and reliable instrument which can readily 
 be fitted up for measuring directly the exhaust tempera- 
 tures ; hence most experimenters have to adopt a method 
 of calculation from the indicator diagram. 
 
 We may now summarise the observations during a gas- 
 engine trial as follows : 
 
 Time of observations. 
 
 Brake readings { fP rin S balance. 
 I Load on brake. 
 
 Eevolutions of the engine by counter. 
 Explosions per minute. 
 
 {Quantity, cubic feet. 
 Temperature. 
 Pressure. 
 
 {Quantity, pounds. 
 Inlet temperature. 
 Outlet temperature. 
 Height of barometer. 
 Indicator diagrams. 
 Temperature of exhaust (when possible). 
 
126 THE PRACTICAL ANALYSIS OF COAL GAS. 
 
 CHAPTER XIII. 
 THE PRACTICAL ANALYSIS OF COAL GAS. 
 
 Analysing the Gas. Hempel's apparatus is most generally 
 used by engineers for the analysis of furnace and coal gases. 
 If carefully handled, the results may be relied upon as being 
 accurate to the half of 1 per cent. The principle upon 
 which the analysis depends is briefly as follows : From a 
 known volume of gas certain constituents are absorbed. 
 The remaining volume is measured after each absorption is 
 complete; hence the volumetric proportion of the con- 
 stituents may be determined. If there is a residue which 
 cannot be absorbed, as is the case with coal gas, the 
 remaining quantities may be determined by combustion in 
 a manner to be described. 
 
 We will suppose that a sample of coal gas has been 
 collected during the trial, as previously described. This 
 sample, we know, consists of carbon- dioxide (CO 2 ), olefines 
 (CeH*), oxygen (O), carbon-monoxide (CO), hydrogen (H), 
 and marsh gas (CEU). Our analysis will enable us to 
 determine the volumetric proportions of these constituents, 
 but it will be noticed that the method depends entirely 
 upon our knowledge of their presence. The constituents 
 must be absorbed in the following order : 
 
 1. Carbon-dioxide, absorbed by potassium hydrate. 
 
 2. Olefines, aborbed by strong sulphuric acid. 
 
 3. Oxygen, absorbed by phosphorus or pyrogallic acid. 
 
 4. Carbon-monoxide, absorbed by cuprous chloride dis- 
 solved in hydrochloric acid. 
 
 The apparatus for carrying out these processes is illus- 
 trated in figs. 76 and 77. A is a plain tube, called the 
 levelling tube ; B is a graduated tube, called the burette. 
 The burette is graduated up to 100 cubic centimetres. The 
 two tubes are connected together by means of a flexible 
 rubber pipe, which may be of comparatively thin section 
 
THE PRACTICAL ANALYSIS OF COAL GAS. 
 
 127 
 
 when water is used in the apparatus, but should be much 
 stouter tubing when mercury is used. 
 
 There is a cock at the top of the buretted and a pinch 
 cock upon the rubber connection. The level tube is first 
 
 PRACTICAL ENGINEER. 
 
 PIG. 76. Hempel's gas 
 burettes. 
 
 Fro. 77. Method of levelling when 
 measuring volume of gas. 
 
 filled with water (or mercury if procurable). The burette 
 is then lowered until the water gravitates into it and 
 
128 
 
 THE PRACTICAL ANALYSIS OF COAL GAS. 
 
 completely fills it up to the stop cock. Now connect the 
 syphon tube of the sample bottle to a water tap, and gently 
 turn on the water so as slightly to raise the pressure of the 
 sample gas. Allow a little gas to escape through the other 
 tube from the sample bottle, in order to get rid of any air 
 beyond the pinch cock. The sample may now be connected 
 to the burette, and the water run from the burette back 
 into the level tube. The sample gas will then be drawn 
 into the burette. It is better to draw in rather more than 
 
 FIG. 7 s . Hempel's simple absorption pipette. 
 
 100 c.c. before disconnecting the sample bottle. This done, 
 close the stop cock of the burette, and raise the level tube 
 until the surface of the water stands at zero. Close the 
 pinch cock on the flexible tube. The pressure in the burette 
 will now be slightly above that of the atmosphere ; hence 
 the stop cock of the burette should be opened momentarily 
 just to allow the excess of pressure to escape. This method 
 of filling contributes towards accuracy, for it is of the highest 
 importance that no measurements of volumes be taken 
 excepting when the gas is at atmospheric pressure. It is 
 
THE PRACTICAL ANALYSIS OF COAL GAS. 
 
 129 
 
 also of the highest importance that the temperature of the 
 burette remain constant, and for this reason the tubes 
 should never be touched ; all lifting should be done by 
 holding the wooden stands into which the tubes are fitted. 
 By drawing the hand once or twice down the burette, an 
 alteration of quite 5 per cent will be noticed in the 
 reading, thus showing the importance of uniformity of 
 temperature throughout the analysis. 
 
 Having measured off our 100 c.c. at atmospheric pressure, 
 we next proceed to absorb the carbon-dioxide. Each 
 absorption is carried out in a pipette, the forms of which 
 are shown in figs. 78, 79, and 80. Fig. 80 represents 
 two glass bulbs connected together by glass tubes as shown. 
 
 FIG. 79. Heinpel's double absorption pipette. 
 
 The larger bulb is filled with a strong solution of potassium 
 hydrate until the capillary U tube fills. It is advantageous 
 to place a number of & in. solid glass rods inside the long 
 bulb. The solution adheres to the surface of the rods, and 
 thus presents a very large surface for absorbing th CO 2 . 
 
 10G 
 
130 THE PRACTICAL ANALYSIS OF COAL GAS. 
 
 After making sure that the solution is well up the capillary 
 tube, couple the burette to the pipette, as in fig. 81, by a very 
 short piece of capillary tube. The admission of air in coupling 
 and uncoupling with the various pipettes used is a frequent 
 source of error, and the utmost care is required. A con- 
 
 Fio. 80. Hempel's pipette for solid substances. 
 
 venient way of making these joints is to slip a fin. length 
 of rubber tube over and beyond the end of one tube. Then 
 butt the two ends together, and by wetting the glass the 
 rubber can be slipped over the joint without enclosing a 
 large volume of air. When the joint is made, open the 
 stop cock, raise the level tube, and drive all the gas over 
 into the pipette. The absorption will be complete in about 
 half an hour. The process will be assisted by shaking the 
 solution or by drawing the gas back into the burette, so that 
 the glass rods may be again moistened. 
 
 When the process of absorption is completed, lower the 
 level tube and allow the water to flow from the burette 
 until the caustic potash in the pipette rises up the capillary 
 tube to the position it at first occupied. Now close the stop 
 
THE PRACTICAL ANALYSIS OF COAL GAS. 
 
 131 
 
 cock at the top of the burette, and hold the level tube against 
 the burette at such a height that the water in both tubes is 
 exactly at the same level. In this way the gases in the 
 burette are always brought to atmospheric pressure before 
 their volume is measured. In taking the readings great 
 
 PLAIN 
 
 TUBE 
 
 PRACTICAL ENGINEER 
 
 FIG. 81. Method of using Hempel's apparatus. 
 
 care should be exercised. When water is used its surface in 
 the tube will be concave ; with mercury the surface is 
 convex. In either case the line of sight when taking a 
 reading should be horizontal and tangential at the centre 
 of the concave or convex surfaces. Before taking a reading 
 
132 THE PRACTICAL ANALYSIS OF COAL GAS. 
 
 when water is used in the tubes, the burette should be 
 allowed to stand a few moments to allow drops to run down 
 the sides. It is only by careful attention to such details 
 that any approach to accuracy can be secured. An exactly 
 similar process is carried out in order to absorb the defines. 
 The pipette used is shown in fig. 78, and contains strong 
 sulphuric acid. 
 
 To absorb the oxygen, a pipette similar to fig. 80 is used. 
 This should be filled with water, and contain a number of 
 sticks of phosphorus of about % in. diameter. As phosphorus 
 is not readily procured in such thin sticks, it will be well to 
 describe how they may easily be produced from lumps of 
 phosphorus without danger to the operator. Procure a glass 
 tube, the bore of which is equal to the diameter of the sticks 
 required. At the end of this place a short length of rubber 
 pipe. Take a tumbler of water, at about 120 deg. Fah., and 
 quickly drop the lumps of phosphorus into the hot water. 
 The phosphorus will now melt at the bottom of the tumbler. 
 Squeeze the rubber pipe attached to the glass tube, and then 
 immerse the end of the glass tube in the melted phosphorus. 
 Owing to the head of water in the glass, it will be found 
 that the phosphorus will run up the glass tube when the 
 rubber at the other end is released. By suitably arranging 
 the head of water, sticks of phosphorus of the required 
 length may be cast. To get the stick of phosphorus from 
 the glass tube it is advisable to remove the rod into a vessel 
 of cold water, but before doing so take care to pinch tightly 
 the rubber connection, so that the phosphorus does not 
 drop when taken from the water. In a few seconds the 
 phosphorus will set, and will stick in the tube if not pre- 
 vented. Whilst setting the phosphorus should be kept 
 moving in the glass tube by pressing the rubber tube in one 
 or two places in addition to keeping its end tightly closed. 
 A little practice will enable the operator to cast phosphorus 
 s afely and quickly. It is extremely important in dealing 
 with such an inflammable material as phosphorus that it 
 
THE PRACTICAL ANALYSIS OP COAL GAS. 133 
 
 should be handled always under cold water, and never be 
 left exposed to the air for many seconds. 
 
 The absorption of oxygen is carried out exactly as 
 previously described. 
 
 The carbon-monoxide (CO) is absorbed in a double pipette, 
 as shown in fig. 79. The two bulbs on the right contain 
 water, whilst the two bulbs on the left attached to the 
 capillary tube contain the cuprous chloride and hydro- 
 chloric acid. In manipulating this pipette great care should 
 be taken not to pass the cuprous chloride over into the 
 water bulb If this be done, a white precipitate of cuprous 
 chloride will be formed in the water chamber, and if left 
 may block up the passage between the water bulbs. With 
 this additional caution, the work may be continued as 
 previously described. 
 
 We have now absorbed four of the constituents of the gas, 
 and we have a residue of hydrogen and marsh gas. To 
 evaluate these quantities an entirely different method is 
 employed. 
 
 To effect the explosion of the hydrogen and marsh gases, 
 a pipette of the form shown in fig. 78 is used, having two 
 platinum wires fused into the bulb B, terminating in 
 sparking points on the inside of the bulb. See that the 
 bulb B and the capillary tube C are filled with water. 
 Then drive from the burette 10 c.c. of the residue into the 
 pipette. Then, either by emptying the burette, or, pre- 
 ferably, by using a second burette, drive seven times the 
 volume of air that is, 70 c.c. of pure air into the pipette, 
 as a supply of oxygen for the explosion. It will be 
 found that this charge occupies about half the volume 
 of B. This should not be exceeded. By means of 
 a bichromate battery and induction coil send a spark 
 through the explosive gas, and violent combustion will 
 follow. On account of the sudden pressure, it is well to 
 bind the rubber connection to the pipette with wire. Some- 
 times a glass stop cock is provided between the chambers 
 A and B. If so, this cock must be open when the explosion 
 
134 THE PRACTICAL ANALYSIS OF COAL GAS. 
 
 takes place, or the bulb B will be broken by the pressure. 
 After the explosion measure the volume of the remaining 
 gas by drawing it back into the burette, observing the 
 cautions already given. It will be found that there is a 
 considerable diminution in volume, due to the free hydrogen 
 and the hydrogen in the marsh gas combining with oxygen 
 of the air to form steam, which ultimately is condensed on 
 the sides of the pipette to a negligible volume of water. 
 The gases remaining after the explosion consist of carbon- 
 dioxide the result of the combustion nitrogen, which was 
 drawn in with the air, and an excess of oxygen above that 
 required for the complete combustion of the hydrogen and 
 marsh gas. It will be observed that the diminution in 
 volume after the explosion is due to the disappearance of 
 gaseous hydrogen from two constituents, namely, the free 
 hydrogen and that contained in the marsh gas. It is known 
 that one volume of marsh gas when completely burnt with 
 oxygen forms an equal volume of carbon-dioxide. If, 
 therefore, we absorb the carbon-dioxide by means of the 
 caustic potash pipette, we shall obtain the volume of the 
 CO 2 formed by the explosion. This volume gives the 
 required volume of marsh gas. The remainder of the 
 diminution of volume after the explosion must therefore be 
 due to combination of free hydrogen with oxygen. And we 
 know that one volume of oxygen combines with two volumes 
 of hydrogen ; hence of the three volumes disappearing due 
 to hydrogen two-thirds of this contraction gives the 
 volume of hydrogen. The nitrogen in the coal gas may be 
 estimated " by difference." But it is preferable to absorb all 
 the oxygen from the residue, and so obtain a residue of 
 nitrogen only. The nitrogen contained in the air used for 
 the explosion may be calculated, and subtracting this from 
 the total remainder of nitrogen gives the volume due to the 
 coal gas. 
 
 As an example of the process, the following analysis of 
 Leeds coal gas, obtained by the author in 1895, is fully 
 worked out : 
 
THE PRACTICAL ANALYSIS OF COAL GAS. 135 
 
 90 c.c. of coal gas were taken into the burette. 
 
 1. Volume of water in burette before any constituents were 
 
 absorbed was 10 c.c. 
 
 Volume of water in burette after absorption of carbon- 
 dioxide (CO 2 ) 101 c.c. 
 
 Therefore volume of CO 2 in 90 c.s. of coal gas = O'l c.c. 
 
 Equivalent volume of C0 2 in 100 c.c. of gas = O'll c.c. 
 
 2. Volume of water in burette before absorption of olefines 
 
 (C 6 H 4 ) = lO'l c.c. 
 
 Volume of water in burette after absorption of olefines 
 
 (C 6 H 4 ) = 12-6 c.c. 
 
 Therefore volume of C 6 H 4 in 90 c.c. of gas = 2'5 c.c. 
 
 Equivalent volume of 6 H 4 in 100 c.c. of gas = 277 c.c. 
 
 3. No oxygen was detected. 
 
 4. Volume of water in burette before absorption of carbon- 
 
 monoxide (CO) = 12'6 c.c. 
 
 Volume of water in burette after absorption of carbon- 
 monoxide (CO) ..... = 20-0 c.c. 
 
 Therefore volume of CO in 90 c.c. of coal gas = 7'4 c. c. 
 
 Equivalent volume of CO in 100 c.c. of gas = 8 22 c.c. 
 
 10 c.c. of residue were taken into burette, and mixed with 
 80 c.c. of air ; hence 
 
 Volume of water in burette before explosion = 10 c.c. 
 
 Volume of water in burette after explosion = 26 c. c. 
 
 Therefore total contraction in volume was 16 c.c. 
 
 Volume of water in burette before absorbing CO 2 formed by 
 
 explosion = 26 c.c. 
 
 Volume of water in burette after absorbing C0 2 formed by 
 
 explosion = 30'7 c.c. 
 
 Therefore C0 2 = 4-7 that is, volume of CH 4 in 10 c.c. of 
 
 residue = 4 % 7 c.c. 
 
 But if 100 c.c. of coal gas had been taken into burette, then the total 
 residue of CH 4 and hydrogen would have been 
 
 100 - (O'll + 2-77 + 8-22) 
 = 100 - 11-10 
 = 88-9 residue of CH 4 and H. 
 
 Therefore the total CH 4 will be 
 
 ! 
 
 10 
 
 Now, we have seen that one volume of CH 4 combines 
 with two volumes of O 2 to form one volume of CO 2 and two 
 volumes of H 2 O. The two volumes of H 2 O condense to 
 water, and are quite negligible. Therefore we see that the 
 
136 THE PRACTICAL ANALYSIS OF COAL GAS. 
 
 diminution in volume due to CH 4 in the exploded gases is 
 equal to twice the volume of C0 2 formed by the combustion. 
 
 Volume of CO 2 absorbed from products of combustion 
 
 = 47 c.c. 
 
 Total contraction of volume after explosion = 16 c.c. 
 Therefore contraction due to free hydrogen 
 
 = total contraction - contraction due to CH 4 
 
 = 16 - 47 x 2 = 16 - 9'4 = 66 c.c. 
 
 Now, two volumes of hydrogen combine with one volume 
 of oxygen, and all three volumes disappear after the steam 
 formed by combustion is condensed. It is evident, therefore, 
 that out of the 6'6 c.c. of contracted volume only two-thirds 
 was free hydrogen. 
 
 The volume of hydrogen in 10 c.c. of residue 
 
 = 6-6 x | = 4'4. 
 
 Therefore the volume of hydrogen in total residue, 
 namely, 
 
 88-9 c.c. = 4'4 x - 39 1L 
 
 Out of 10 c.c. of residue we have obtained 4'4 hydrogen 
 and 47 CH 4 , which together make up 91 c.c. ; the remaining 
 0*9 c.c. should be the volume of nitrogen in the coal gas. 
 Therefore, by difference, the nitrogen amounts to 
 
 0-9 x ^ = 8'0 c.c. (nearly). 
 
 But our analysis ought to be further checked, and we will 
 proceed to estimate the nitrogen by direct measurement. 
 
 Our residue now consists of nitrogen due to coal gas + 
 nitrogen due to air taken in for the explosion + excess of 
 oxygen in the air not required to complete the combustion. 
 
 If we absorb the excess of oxygen, we shall at once obtain 
 the whole volume of nitrogen present. 
 
 .Referring to last reading of burette, we find 
 
THE PRACTICAL ANALYSIS OF COAL GAS. 137 
 
 Volume of water in burette before absorbing oxygen, 
 
 307 c.c. 
 Volume of water in burette after absorbing oxygen, 
 
 35 9 c.c. 
 
 Therefore excess of oxygen = 5'2 c.c. 
 Therefore the total nitrogen left in burette = 100 - 35*9 
 
 = 641 c.c. 
 
 The percentage of nitrogen in air is approximately 79 per 
 cent ; therefore out of 80 c.c. taken into burette the 
 nitrogen 
 
 = 80 x ^ = 63-2 c.c. 
 
 Therefore the volume of nitrogen due to the coal gas 
 = 641 - 63-2 = 0-9 c.c. (nearly). 
 
 This agrees with our determination by difference ; hence 
 we know our analysis to be very approximately correct. 
 Collecting our results, we get 
 
 Volume per cent of C0 2 = Oil 
 CH 4 = 277 
 
 ,. o= o-oo 
 
 C0= 8-22 
 H = 3911 
 CH 4 = 4178 
 N = 8 00 (by direct measurement) 
 
 99-99 error O'Ol * 
 
 The various producer gases may be analysed by Hempel's 
 apparatus in the manner described. The following tables 
 give the composition of coal and producer gases analysed in 
 
 *The author regards this near approximation to accuracy as the result of 
 rather larger compensating errors in the work. Nothing less than 01 c c. can be 
 read with certainty upon the burette scale ; hence an error of 0*0 1 on the 
 ai.alysis may give a false impression of the accuracy obtainable with this 
 apparatus. 
 
138 
 
 THE PKACTICAL ANALYSIS OF COAL GAS. 
 
 connection with gas-engine trials. It will be understood 
 that the figures given below refer to particular samples of 
 town gases, and that the daily variation in the composition 
 always necessitates an analysis when the constituents of the 
 gas are required with accuracy. 
 
 COMPOSITION OF COAL GAS IN VARIOUS CITIES. 
 
 Percentage by volume. 
 Constituents in gas. 
 
 Leeds, 
 1895 
 (Grover). 
 
 London 
 Society of 
 Arts trial. 
 
 Kilmarnock 
 ( Professor 
 Kennedy). 
 
 Paris 
 (Witz). 
 
 Marsh gas CH 4 
 
 41-78 
 
 37 '34 
 
 42 80 
 
 32'30 
 
 defines C 6 H 4 
 
 277 
 
 3-77 
 
 5'55 
 
 5-50 
 
 Hydrogen H 
 
 39 11 
 
 50-44 
 
 43-60 
 
 52-80 
 
 Carbon-monoxide 
 
 8-22 
 
 3-96 
 
 4-30 
 
 5-60 
 
 Nitrogen 
 
 8 '00 
 
 3'98 
 
 2 70 
 
 3-80 
 
 Ca bon-dioxide and oxygen . . 
 (Chiefly CO 2 with traces of O.) 
 
 0-12 
 
 0-51 
 
 5-35 
 
 o-oo 
 
 COMPOSITION OF POOR GASES.* 
 
 Name of gas. 
 
 g^ 
 
 s 
 
 B 
 |^ 
 
 -S^ 
 
 >,> 
 H 
 
 L 
 
 V' 
 
 Olefiant gas. 
 Vol. % 
 
 Carbon- 
 monoxide. 
 Vil. % 
 
 II s * 
 "S 'x ~* 
 
 3$ 
 
 I 5 * 
 
 
 
 Siemens producer 
 gas 
 
 
 8 60 
 
 2 '40 
 
 
 24'40 
 
 5-20 
 
 59-40 
 
 Water gas 
 
 o-io 
 
 50-50 
 
 0-60 
 
 
 44-40 
 
 1'60 
 
 
 Strong gas 
 Lowe gas 
 Dowson gas . . . . 
 
 0'03 
 
 53-00 
 30-00 
 18-73 
 
 0'31 
 
 0-31 
 
 35-00 
 28-00 
 25-07 
 
 4-00 
 34-00 
 6-57 
 
 8-00 
 
 s-oo 
 
 48-98 
 
 Dowson gas 
 
 0-23 
 
 24-36 
 
 1-16 
 
 0-15 
 
 17-55 
 
 6-07 
 
 50-48 
 
 Lenchanchez 
 
 0-50 
 
 20-00 
 
 
 4-0 
 
 21-00 
 
 5-00 
 
 49-50 
 
 ' From " Donkin on Gas, Oil, and Air Engines." 
 
AIE REQUIRED FOR COMBUSTION. 139 
 
 CHAPTER XIV. 
 
 CALCULATIONS REQUIRED IN WORKING OUT RESULTS OF 
 ENGINE TRIALS. 
 
 THE most important deductions to be made from the 
 analysis of the gases used are (1) The quantity of air 
 required for complete combustion ; (2) to obtain the calorific 
 value of the gas ; (3) to calculate the specific heat of the 
 waste gases. Of these the two first mentioned are of the 
 greatest importance from a commercial standpoint. The 
 loss of heat in the waste gases is difficult to determine with 
 accuracy, and where the temperature of the gases is only 
 approximately known it is an absurd refinement to calculate 
 the specific heat of the gases to three places of decimals. 
 
 For the complete explanation of the reactions which take 
 place between the constituents of coal gas and oxygen, the 
 reader is referred to any modern elementary text-book on 
 chemistry. A knowledge of the facts set forth in the 
 following table enables us to work out the theoretical 
 quantity of air required for combustion. 
 
 The nitrogen, carbon- dioxide, and oxygen are not com- 
 bustible. Strictly, the oxygen contained in the coal gas 
 should be deducted from that quantity given in column (h), 
 but the oxygen is usually in such small quantities as to be 
 quite negligible. 
 
 Air is composed of 23 per cent of oxygen and 77 per cent 
 of nitrogen by weight ; therefore the weight of air required 
 per pound of coal gas will be 
 
 2-89 x i?? = 12-5. 
 
 In other words, the proportion of air to gas by weight for 
 complete combustion is equal to 12 '5 to 1. One cubic foot 
 of air weighs 0'08082 Ib. Hence the proportion of air to gas 
 by volume equals 572 to 1. It must be remembered that 
 all these figures are worked out for a temperature of 32 deg. 
 
140 
 
 AIR REQUIRED FOR COMBUSTION. 
 
 Fah. and a pressure of 147 Ib. per square inch absolute. Any 
 increase of pressure causes an increase in density that 
 is, an increase in weight per unit volume. Conversely, 
 any increase in temperature causes a decrease in density 
 when the pressure remains constant that is, decrease in the 
 weight per unit volume. 
 
 
 (a) 
 
 GO 
 
 (c) 
 
 (d) 
 
 (<) 
 
 (/) 
 
 00 
 
 (A) 
 
 Constituents. 
 
 Volume per 
 cent. 
 
 Weight of one 
 cubic foot. 
 
 Calorific value 
 per pound. 
 
 Proportion of 
 weight of oxy- 
 gen required. 
 
 Weight, in one 
 cubic foot of 
 coal gas. 
 
 Proportion by 
 weight. 
 
 Calorific value 
 per cubic foot 
 of rotll gas. 
 
 Weight of oxy- 
 gen required. 
 
 
 
 Lbs. 
 
 B. T. U. 
 
 
 Lbs. 
 
 
 B. T. U. 
 
 Lbs. 
 
 Marsh gas 
 
 41-78 
 
 0*0447 
 
 21 690 
 
 4 
 
 0-01867 
 
 0*5061 
 
 404*9 
 
 2*024 
 
 defines . . . 
 
 2-77 
 
 0*1174 
 
 20 260 
 
 
 0*00325 
 
 0'088 
 
 65*8 
 
 0*302 
 
 Hydrogen , 
 
 39-11 
 
 00559 
 
 52 500 
 
 8 
 
 0*00218 
 
 0'059 
 
 114-4 
 
 0'472 
 
 Carbon monoxide . 
 
 8-22 
 
 0-0783 
 
 4,300 
 
 * 
 
 0*00643 
 
 0*174 
 
 277 
 
 0-099 
 
 
 8*00 
 
 0*0783 
 
 
 
 0*00626 
 
 0*170 
 
 
 
 Carbon - dioxide ) 
 and oxygen... f 
 
 0-12 
 
 01060 
 
 .. 
 
 .. 
 
 0*00012 
 
 0003 
 
 
 
 
 
 
 
 
 
 
 
 0-03691 
 
 1*000 
 
 612*8 
 
 2-897 
 
 Column (a) gives the volumes per cent of the constituents as furnished by 
 analysis. 
 
 Column (6) gives the known weights of each constituent per cubic foot at 
 32 deg. Fah. and at 14*7 Ib. per square inch in absolute pressure. 
 
 Column (c) gives the known heating value of 1 Ib. of each constituent at 32 deg t 
 Fah. and 14'71b. square inch absolute. 
 
 Column (d) gives the proportion of weight of oxygen required to combine with 
 a given weight of gas. Thus one cubic foot of marsh gas weighs 0'04471b., and 
 will require for its combustion four times that weight of oxygen : 0'0447 X 4 = 
 0'1788lb. oxygen. 
 
 From these data the remaining columns may be filled in. 
 
 Column (e). Multiply figures in column (a) by those in column (6). 
 
 Column (/) This gives the proportion of weight of constituents in 1 Ib. ol 
 coal gas. These figures are found by dividing the weight of each constituent in 
 column (e) by the total weight of one cubic foot of coal gas. 
 
 Column (g). These figures are the products of (c) x (e). 
 
 Column (A). These figures are the product of (d) x (/). 
 
SPECIFIC HEATS OF GASES AT CONSTANT VOLUME. 141 
 
 The calorific value of the gas, as given in column (#), 
 should agree with the experimental determination by 
 Junker's calorimeter when both are reduced to the same 
 standard. The calorific value for hydrogen is given minus 
 the latent heat taken up in the formation of steam ; therefore 
 no correction need be made, as described in the text relating 
 to Junker's calorimeter. 
 
 It is sometimes of interest to calculate the maximum 
 temperature possible during the combustion of the mixture 
 in the cylinder. To do this we require to know the thermal 
 value of the charge of gas, the specific heat at constant 
 volume, and the weight of the whole charge. The volume 
 of gas entering the cylinder per cycle is easily obtained 
 from the gas-meter readings. We will now proceed to find 
 the specific heats of various mixtures of coal gas and air for 
 the sample tabulated in the above table. First, to find the 
 specific heat of the coal gas. The following table gives the 
 data required : 
 
 
 (a) 
 
 Proportion by 
 weight. 
 
 (&) 
 Specific heat 
 of constituent 
 at constant 
 volume. 
 
 () 
 
 Marsh gas 
 
 0-506 
 
 470 
 
 0*2378 
 
 defines 
 
 0*088 
 
 0'332 
 
 0*0292 
 
 Hydrogen. 
 
 0*059 
 
 2'406 
 
 0-1419 
 
 Carbon-monoxide 
 
 0'174 
 
 0*173 
 
 0*0301 
 
 Nitrogen 
 
 0-170 
 
 173 
 
 0*0294 
 
 Carbon-dioxide ) 
 
 0-003 
 
 0-171 ^ 
 
 
 
 
 0-155 j 
 
 00048 
 
 
 
 
 
 
 
 
 0*4732 
 
 Column (a) gives the proportion by weight of the constituents of the coal gas. 
 These figures are copied from the previous table. 
 
 Column (6) gives the specific heats of the constituents at a constant volume. 
 
 Column (c). These figures are the result of the product of the figures In column 
 (a) by column (6), and the total is the specific heat of the gas. This equals 0-473. 
 
142 SPECIFIC HEATS OF GASES AT CONSTANT VOLUME. 
 
 Taking the specific heat of air as 0*168 at constant volume, 
 we will next work out the specific heats of the following 
 mixtures of air and gas : 
 
 Volume of air to gas. 
 
 Specific heat of mixture 
 at constant volume. 
 
 Proportion of air to gas 
 
 hy wpigrht,. 
 
 61 
 
 0-189 
 
 13-13 to 1 
 
 71 
 
 0-186 
 
 15-32 to 1 
 
 8-1 
 
 0-184 
 
 17-51 to 1 
 
 9-1 
 
 0-183 
 
 19-70 to 1 
 
 10-1 
 
 0-181 
 
 21-89 to 1 
 
 111 
 
 0-180 
 
 24-07 to 1 
 
 121 
 
 0-179 
 
 26-26 to 1 
 
 131 
 
 0-178 
 
 28-45 to 1 
 
 141 
 
 0-178 
 
 30-64 to 1 
 
 151 
 
 0-177 
 
 32-83 to 1 
 
 First find the proportion of air ' to gas by weight. The 
 ratio of the weight of one cubic foot of air to one of gas 
 
 0-0808 
 0-0369 
 
 2189. 
 
 Therefore proportion of 6 to 1 mixture by weight = 6 x 
 2189 to 1 = 13134 to 1. The other figures are similarly 
 calculated. 
 
 Proceeding now to obtain the specific heat of a 6 to 1 
 mixture, we have (proportion by weight of air) x (specific 
 heat of air) + (proportion by weight of gas) x (specific heat 
 of gas) -f- (total weight of air + gas). This, in figures, 
 
 _(1313 x 0168) + (1 x 0'473)_2-205 + Q'473 = 2 678 =0189 
 1313 + 1 1413 1413 
 
 Similarly the remaining figures in the above table are 
 arrived at. 
 
 The specific heat of any mixture may be determined 
 approximately by a formula known as Grashof's formula. 
 
SPECIFIC HEATS OF GASES AT CONSTANT VOLUME. 143 
 
 The formula is based upon a mean specific heat for coal gas, 
 assuming that its constituents never vary in their proportion 
 one to the other. 
 
 Let Cv = specific heat of mixture at a constant volume. 
 R = ratio of vols. of air to gas in the mixture ; 
 
 R x 0-168 + 0-226 
 
 then 
 
 R + 0-415 
 
 Putting in values for the above case, already worked out 
 from the exact data, we obtain, 
 
 _ f x 0168 + 0-226 
 
 I + 0-415 
 1 '234 
 
 6-415 
 = 0192, 
 
 the error in this instance amounting to about 1| per cent. 
 
 The following will serve as an example of the use of these 
 figures, until we come to consideration of the indicator 
 diagram. 
 
 Suppose 6 cubic feet of air to be mixed with 1 cubic foot 
 of the coal gas given in the previous analysis. What would 
 be the maximum temperature possible when the explosion 
 takes place at 32 deg. Fah. and 147 lb. pressure ? 
 
 Data required. (1) Specific heat of mixture ; (2) weight 
 of mixture ; (3) calorific value of the gas. 
 
 Then temperature 
 
 _ total heat present 
 
 weight of mixture x specific heat 
 = oo , 612 
 
 0-521 x 0189 
 = 32 + 6320 (nearly) 
 = 6352 deg. Fah. 
 
 In a gas engine the theoretical temperature is never 
 recorded by the indicator card, as there is a very rapid 
 transmission of heat through the walls of the cylinder. 
 
144 INDICATOR DIAGRAM CALCULATIONS. 
 
 CHAPTER XV. 
 CALCULATIONS ON THE INDICATOR DIAGRAM. 
 
 The Indicator Diagram. We have already considered the 
 means and method of actually taking an indicator diagram. 
 It now remains to be shown what results may be obtained 
 from it by simple calculations. The indicator diagram 
 furnishes information upon the following points, which we 
 will consider in order : (1) The indicated horse power ; (2) 
 the valve setting ; (3) the initial, maximum, and exhaust 
 temperatures; (4) the nature of the expansion and com- 
 pression curves. 
 
 Although the indicated horse power of a gas engine is 
 never a very certain quantity, it is always calculated in gas- 
 engine trials. The inertia of the indicator levers, the effects 
 of high temperature upon the indicator spring, and the 
 difficulty in obtaining a true average from a number of 
 diagrams when the engine is running below its full load, all 
 contribute towards inaccuracy. For these reasons the ratio 
 of brake horse power to indicated horse power cannot always 
 be relied upon as correct. High mechanical efficiencies of 
 gas engines must, therefore, be regarded with suspicion 
 unless we have complete access to all the details of the test. 
 
 The indicated horse power is found as follows : 
 
 Let I = the indicated horse power. 
 
 P = mean pressure per square inch on the piston. 
 
 L = length of stroke in feet. 
 
 E = number of explosions per minute. 
 
 A = area of cylinder in square inches. 
 
 Then T _LEAP 
 
 33000 ' 
 
 In working out a number of diagrams from one engine, it 
 will be found convenient to work out in decimal form the 
 
AVEEAGING DIAGRAMS. 145 
 
 fraction , and record this once for all as the engine 
 
 constant. 
 
 The factor we require to determine from the diagram is 
 the mean pressure per square inch. Vertical measurements 
 give pressures, and horizontal measurements distances ; 
 hence the length of an ordinate measured with the scale of 
 pressures gives the pressure per square inch at a particular 
 point of the stroke. If the pressure remained constant 
 throughout the whole stroke, the diagram would become a 
 rectangle, and the product of pressure and distance would give 
 the work done. From this we see that the length x breadth, 
 or area of the diagram, represents to some scale the work 
 done per stroke. We may therefore find the mean pressure 
 upon the piston, either by getting the average height of a 
 set of equidistant ordinates, or by dividing the area of the 
 diagram by the length. 
 
 To facilitate the work of finding the mean height, 
 instruments commonly known as averagers are used. 
 Before illustrating these we will briefly describe the most 
 expeditious way of calculating the mean height without 
 special instruments. 
 
 Divide the diagram into ten equal vertical strips. Sub- 
 divide each of these by a dotted line passing through their 
 centres. It is required to find the average height of these 
 dotted lines. Take a strip of paper, and mark off upon its 
 edge the length of the first ordinate, then add to this the 
 length of the second by applying the paper to the second 
 ordinate. In this way we obtain the sum of the lengths of 
 the ordinates. Measure this with a rule divided into inches 
 and tenths. Suppose this gives us 5 '08 in., we know that 
 the average height will be 0'508 in. If the diagram be taken 
 with a 100 spring, our mean pressure becomes 50*8 Ib. per 
 square inch. When a number of diagrams are to be dealt 
 with, it will be found convenient to construct a set of ten 
 converging lines upon tracing paper. By applying this to 
 llo 
 
146 
 
 GOODMAN'S AVERAGER. 
 
 the indicator diagrams, the ten ordinates may readily be 
 pricked off. 
 
 Where a great deal of testing is done, the work will be 
 much facilitated by the use of an averager for obtaining 
 mean pressures. Of these we shall describe very briefly the 
 two most commonly used, viz., the Goodman and Coffin 
 averagers. 
 
 FIG. 82. 
 
 Goodman's averager is shown in fig. 82, and is supplied 
 by Messrs. Jackson Brothers, Leeds. The tracing point is 
 fixed to the horizontal bar. The other leg of the instrument 
 is carefully ground to a knife edge in such a way that the 
 edge, if produced, would pass through the tracing point. 
 The distance between the knife edge and the tracing point is 
 adjustable by sliding the former along the horizontal bar, 
 by means of which the instrument is set to the exact 
 length of the diagram to be measured. Suppose the diagram, 
 fig. 83, is to be measured. Choose a point A somewhere 
 about the centre of the figure, and draw any line A B to meet 
 the boundary. The tracing point is placed upon A, and the 
 the hatchet end put so that the mean position of the instru- 
 ment is roughly square with A B, as shown at X. The position 
 of X is marked by pressing the knife edge upon the paper. 
 Now take the tracing point lightly in the fingers, so that the 
 movement of the hatchet end is not controlled by the pres- 
 sure of the hand. Travel from A to B, then follow the 
 direction of the arrows round the boundary of the figure, 
 say in a clockwise direction, returning to the point B, 
 thence back to A. The knife edge will then occupy a new 
 
GOODMAN'S AVEEAGER. 
 
 147 
 
 position at Y, which is marked as before. The indicator 
 diagram should now be turned about the point A through' 
 180 deg. The figure is again traced by the point, but 
 this time in a contra-clockwise direction, following the 
 dotted outline, fig. 83. When the point returns again to A 
 the hatchet will have travelled back towards X, say to Xj. 
 The mean side movement of the hatchet end gives at once 
 the average height of the diagram. Thus, suppose the 
 distance from Y to the dot between X and Xi is measured 
 off as 0'56 in.; then this is the average height of the diagram, 
 and multiplying 0'56 by the scale of pressures gives the 
 mean pressure upon the piston. 
 
 FIG. 83. 
 
 The ordinary hatchet planimeter was invented by Captain 
 Pry tz, of Copenhagen, but its use involved calculations far too 
 tedious for practical men. Professor Goodman has embodied 
 these calculations on a scale marked upon his hatchet plani- 
 meter, which enables an area to be measured by reading off 
 the side travel of the hatchet end as so many square inches 
 
148 COFFIN AVERAGER. 
 
 marked upon the scale. The averager described is so 
 arranged that the area of the figure traversed by the boun- 
 dary is equal to the side travel of the hatchet multiplied 
 by the length of the diagram. The complete theory of these 
 instruments is beyond the province of this work ; we may, 
 however, refer those interested to the pages of Engineering, 
 vol. Ivii., page 687, also vol. Ixii., page 225. 
 
 Before leaving the subject, it is is well to point out that 
 accuracy can only be obtained by careful manipulation. 
 The instrument should be carefully set to the total length 
 of the diagram to be measured. The hatchet end should 
 travel on a good surface, such as drawing or blotting paper, 
 and, finally, the diagram should always be reversed so that 
 the mean travel of the instrument may be read off. With 
 these precautions Goodman's averager will be found valuable 
 in quickly determining the average pressures. It further 
 possesses the important advantages that it is not liable to 
 derangement by rough usage, and it is procurable at a price 
 far below that of any other instrument used for the same 
 purpose. 
 
 The Coffin averager supplied by the Globe Engineering 
 Company, Manchester, is illustrated in fig. 84. This in- 
 strument consists of a bar, carrying a recording wheel, 
 having a tracing point at one end, and a pin at the other 
 end, moving in a vertical slot. The only movable parts are 
 the bar, with its wheel and the vertical sliding piece, marked 
 K. The diagram to be measured is placed upon the 
 instrument as shown in the illustration. The atmospheric 
 line should be parallel with the edge of the square marked 
 B. The vertical edges, marked C and K, should nearly touch 
 the extremities of the diagram. The tracing point D, when 
 moved down the inner edge of K, should then coincide with 
 the extremity of the diagram. The weight W is intended 
 to keep the instrument from lifting out of the slot when in 
 use. The recording wheel runs upon a piece of paper 
 fastened to the board upon which the other parts are 
 mounted. 
 
COFFIN AVERAGER. 
 
 149 
 
 Having placed the diagram in position, make an indenta- 
 tion at its extreme end, as at E, with the tracing point D, 
 at the same time setting the wheel to zero. Then trace out 
 
 the diagram by moving in a clockwise direction, until return- 
 ing to the point E. The area of the diagram may now be 
 
150 VALVE SETTING. 
 
 read off in square inches upon the wheel and its vernier 
 scale. We, however, require the mean height of the diagram ; 
 hence, if we move the pointer D upwards along the inner 
 edge of K until the wheel returns to zero, this vertical 
 distance E A will give the mean height. This distance 
 should be measured with the scale of pressures corresponding 
 to the spring used in taking the diagram. The distance 
 A E, as found by this instrument, would be the same as the 
 mean-side travel of the Goodman averager, and might be 
 measured in inches, and then multiplied by the scale of the 
 spring. 
 
 The well-known Amsler's planimeter is often used for 
 determining the area of indicator diagrams. It is, however, 
 much more convenient to use one of the instruments above 
 described, as they are specially designed for the purpose, 
 whereas Amsler's planimeter is designed for measuring 
 much larger areas than usually obtained on indicator 
 diagrams. 
 
 We have already incidentally touched upon the defects 
 of diagrams occasioned by bad valve setting. A weak- 
 spring diagram is necessary to show the action of the inlet 
 valves. These valves, being too small or not having 
 sufficient lift, will cause the suction line of the diagram to 
 fall below the atmospheric line. Late admission of the inlet 
 valves would, of course, be shown by a marked depression 
 of the suction line at the early part of the stroke. When 
 the valves opened the suction line would rise towards the 
 atmospheric. 
 
 With respect to the exhaust, a frequent fault is its late 
 opening. This is invariably caused by the wear of the 
 spindle. An adjusting screw is provided on the lifting 
 lever, so that the lead of the exhaust valve may be 
 increased. The effect on the diagram is to give a sharp 
 point to the toe, with a falling exhaust line. t 
 
 The next point to be considered is the temperature of the 
 gases in the cylinder during the working stroke. We have 
 already stated that without very special apparatus the 
 
i EXPLOSION TEMPERATURES. 151 
 
 temperature of the exhaust gases cannot be measured 
 experimentally. It is, however, customary to furnish in 
 reports of engine trials an account of the heat distribution. 
 For this reason we require to know the temperature of the 
 exhaust gases, so that we may determine approximately the 
 heat thrown away as the exhaust gases pass into the 
 atmosphere. 
 
 It is usually assumed that the following law holds under 
 the conditions in a gas-engine cylinder. When P = 
 absolute pressure, V = volume, and T = absolute tem- 
 perature, we have 
 
 P V 
 
 -^ = constant. 
 
 If, therefore, we determine one set of values for P, V, and 
 T, we can calculate corresponding values for any point on 
 the diagram. The values of P and V are easily measured 
 from the diagram, but at no time is the value of T actually 
 known. We do know, however, the temperature of the 
 jacket water, and it is certain that the inner walls of the 
 cylinder will be hotter than the jacket water. Then, again, 
 exhaust gases left in the clearance volume of the cylinder 
 will help to raise the temperature of the cold incoming 
 mixture. As a compromise, it may be assumed that the 
 temperature of the incoming mixture, before compression 
 commences, is 5 deg. Fah. above that of the j acket water. 
 
 Suppose the following data obtained : Engine clearance, 
 30 per cent ; length of indicator diagram, 3'5 in. ; pressure 
 during the instroke obtained from weak- spring diagram, 
 13'5 Ib. absolute per square inch ; temperature of jacket 
 water, 150 deg. Fah. Let the exhaust take place at 0'9 of 
 the stroke, and let the pressure measured off from the 
 atmospheric line at the moment of exhaust be 25 Ib. 
 
 Horizontal measurements on the indicator diagram repre- 
 sent distances travelled by the piston ; but the piston 
 diameter being constant, these same measurements may also 
 be taken to represent volumes swept out by the piston ; 
 3'5 in. therefore represents the working volume of the 
 
152 EXPLOSION TEMPERATURES. 
 
 cylinder. The clearance volume is known to be 30 per cent 
 of this. ; thus the whole volume is represented by a length of 
 
 3-5 + 3-5 x -??- = 4-55 in. 
 100 
 
 Value of V when exhaust valve opens may be measured 
 from the diagram. In our data it becomes 
 
 (3-5 x 0-9) + (3-5 x^ 
 
 which equals 415. The value of P in pounds per square 
 inch absolute = 25 + 147 = 397 ; 
 
 therefore T = 39 '? x 415 
 
 0-0982 
 
 = 1677 absolute 
 = 1217 deg. Fah. 
 
 The heat thrown away in the exhaust gases per minute 
 will therefore be (temperature Fah.) x (weight of products 
 of combustion per minute) x (specific heat at a constant 
 volume of products of combustion). 
 
 The specific heat at a constant volume, and at constant 
 pressure, may be found by Grashoff's formulae. When R = 
 ratio of air to gas by volume, 
 specific heat of products at constant pressure 
 
 = 0-2375 x R -f- 0343 . 
 R + 0-48 
 
 specific heat of products at constant volume 
 = 01684 x R + 0-286 
 R x 0'48 
 
 The temperature at any point on the indicator diagram 
 may be found as above. In calculating the maximum tem- 
 perature, as shown on the indicator diagram, a correction 
 
 * Absolute temperature = temperature in deg. Fah. plus 460. 
 
EXPANSION CURVES. 153 
 
 for an increase of volume should be made if the explosion 
 line be other than perpendicular. 
 
 The maximum temperature of the gases as shown on the 
 indicator diagram is little more than half the theoretical 
 temperature possible. This is partly owing to the rapid 
 transmission of heat through the walls of the cylinder to the 
 jacket, and partly to the fact that the whole of the heat is 
 not evolved when the highest pressure is reached. From 
 certain characteristics of the expansion curves it is probable 
 that the burning of the gases continues after the maximum 
 pressure is arrived at. 
 
 We have now discussed the deductions which are usually 
 required to be made from the indicator diagram when 
 reporting a gas-engine trial. There are many other 
 points of scientific interest. These are, however, rather 
 beyond the scope of the present work, but are com- 
 pletely discussed in a book recently published upon " Heat 
 and Heat Engines," by Mr. W. C. Popplewell. 
 
 Before leaving the subject of the indicator diagram, it 
 may be well, for the sake of completeness, to state certain 
 facts with respect to the expansion and compression curves, 
 leaving those interested to pursue the question further. 
 
 The equation to the expansion and compression curves is 
 of the form P V w = constant. When the value of n is equal to 
 
 specific heat of products of combustion at constant pressure 
 specific heat of products of combustion at constant volume 
 
 the expansion is adiabatic that is, without loss or gain of 
 heat through the cylinder walls. In a gas engine the water 
 jacket is constantly abstracting heat from the walls ; hence 
 we should expect to find the expansion curve much below 
 the adiabatic. This, however, is not the case ; indeed, in 
 practice it is found that the expansion curve is sometimes 
 above the adiabatic, though usually slightly below. In 
 view of the fact that the maximum temperature of the 
 explosion is little more than half the theoretical temperature 
 possible, and yet the expansion curve so nearly follows the 
 
154 GAS-ENGINE TRIAL. 
 
 adiabatic line, notwithstanding the rapid loss of heat to the 
 cylinder walls, it must be concluded that burning continues 
 long after the maximum pressure has been reached. 
 
 The compression curve, as might be expected, is very 
 nearly adiabatic. 
 
 CHAPTER XVI. 
 GAS-ENGINE TRIAL, OTTO CYCLE : COAL GAS USED. 
 
 WE have now explained the apparatus required for a 
 complete test of a gas engine, and we have discussed various 
 calculations in detail. It now remains for us to go through 
 the figures of a complete trial, which may be taken as a 
 typical example of modern practice. 
 
 Measurements of Enqine. 
 
 Diameter of cylinder, 9 in. ; stroke, 18 in. 
 
 Clearance volume of cylinder, 490 cubic inches. 
 
 Diameter of brake wheel, 4 ft. 
 
 Diameter of brake rope, 075 in. 
 
 Weight of unbalanced part of brake rope, including 
 
 hook, 5 Ib. 
 
 NOTE. Spring balance tested by dead weights, and found 
 to be reading 2 Ib. in excess of real weight. 
 
 Mean Results of Observations taken during a Two Hours' 
 Test. 
 Brake readings { SPB balance (uncorrected UWb. 
 
 I Load on brake (excluding hook), 2oo Ib. 
 Mean revolutions of engine by counter per minute, 160. 
 Explosions per minute, 80. 
 
 r Total gas in cubic feet, 815. 
 
 Gas-meter readings ! Temperature of meter, 65 deg. Fah. 
 I Pressure in mains, 0'9 in. water. 
 
GAS-ENGINE TRIAL. 155 
 
 {Quantity (total), 1,860 Ib. 
 Inlet temperature, 58 deg. Fah. 
 Outlet temperature, 155 deg. Fah. 
 Mean pressure on piston derived from indicator diagrams, 
 
 80'5 Ib. per square inch. 
 
 Pressure when exhaust valve opens, 41 Ib. per square inch. 
 Pressure at end of suction stroke, 13 '8 Ib. per sq. inch absolute. 
 Exhaust commences at 88 per cent of working stroke. 
 Barometric pressure, 15 Ib. per square inch. 
 Calorific value of gas per cubic foot obtained by Junker's 
 
 calorimeter at a pressure of 0'9 in. water = 615 B.T.U. 
 Temperature of calorimeter meter, 65 deg. Fah. 
 
 Calculation of Results from above Data. 
 
 T H P 9 2 x 07854 x 1*6 x 80'5 x 80 = lg . 6 
 33000 
 
 B.H.P. Corrected spring balance reading = 17*5 - 2 
 
 = 15'5lb. 
 
 Total load on brake = 255 + 5 = 260 Ib. 
 Net load on brake = 244*5 Ib. 
 
 0'75 
 Effective diameter of brake wheel = 4 + -y^- ft. 
 
 = 4-062 ft. 
 
 4-062 x 314 x 160 x 244 '5 1F . (1 
 - 33000 
 
 Mechanical efficiency = = = 0'812. 
 
 l.JtL.Jr. lob 
 
 = 81 '2 percent. 
 
 
 Gas Used per I.H.P. Hour. Total consumption of gas 
 uncorrected for pressure and temperature = 815 cubic feet. 
 Consumption reduced to 32 deg. Fah. and 147 Ib. per square 
 inch, 
 
 = 815 x 32 + 46 x 15 + (0-434 x 0'075) 
 65 + 460 14 7 
 
 = 781 cubic feet (corrected). 
 
156 GAS-ENGINE TRIAL. 
 
 Gas used per I. H. P. hour (reduced to 32 deg. and 1471b.) 
 781 - 21 cubic feet. 
 
 2 x 18-15 
 Gas used per B.H.P. hour (reduced to 32 deg. and 147 Ib.) 
 
 781 = 25-9. 
 
 2 x ioM 
 
 The ratio of air to gas should be determined from 
 analysis of the products of combustion. When not done 
 this way, account must be taken of the temperature t 
 pressure of the gas at the end of the suction stroke. '. 
 temperature is difficult to determine with precision, bu 
 usually assumed to be equal to that of the outlet ja( 
 water. Under this assumption the following figures \ 
 the ratio of air to gas by volume. 
 
 Volume of gas per cycle (corrected for temperature 
 pressure at end of suction stroke) 
 
 781 
 
 80 x 2 x 60 13'8 32 + 460 
 
 Working volume of cylinder + clearance volume = 947 
 Volume of air per cycle = 947 - 108 = 839 cubic 
 
 Ratio of air to gas = 7:7 to 1. 
 
 NOTE. When the engine is worked without a scave: 
 device the working volume only is considered. 
 
 ^by urasiioii iuruiuia; 
 
 = 0-168 x 7'2 + 286 . 
 7'2 + 048 
 
 Temperature of Exhaust Gases, by Calculation from Indicator 
 Diagram. Temperature of gases before compression com- 
 mences, assumed as temperature of jacket water + 5 deg., 
 
 = 155 + 5 = 160. 
 
 Total volume of cylinder = clearance volume + working 
 volume 
 
 = 490 + 1145 = 1635 cubic inches, 
 
 = 0*948 cubic feet. 
 
GAS-ENGINE TRIAL. 157 
 
 P V P 1 V 1 
 
 -7p- = mi when P = pressure (absolute). 
 
 V = volume (total). 
 T = temperature (absolute). 
 Filling in values 
 
 P = 13'81b. per square inch absolute. 
 V = 0*948 cubic feet. 
 T = 160 + 460 = 520. 
 
 P 1 = pressure when exhaust valve opens = 147 + 41. 
 V 1 = (1145 x 88) + 490 = 867 cubic feet. 
 
 Then 
 
 _ TP 1 Y 1 _ 55-7 x Q'867 x 520 _ 1Q9n 
 
 PV 138 x 0948 
 
 Temperature = 1920 - 460 = 1460 deg. Fah. 
 
 Heat Lost in Exhaust Gases. Weight of 1 cubic foot of 
 coal gas = 0'0369. Air weighs 2 63 more than coal gas. 
 Therefore weight of charge per cycle 
 
 = 0-948 |' 0369 x 2'63 x 72 + Q-Q369J = 081 lb. 
 L 82 8 2 J 
 
 Heat lost in exhaust gases per cycle 
 
 = O'OSl x 1460 x 0-196 = 231 B.T.U. 
 Heat lost in exhaust gases per minute 
 
 = 231 x 80 = 1848 B.T.U. 
 Heat lost in jacket water per minute 
 
 = (155 - 58) x 186 = 1503 B.T.U. 
 
 . 2 x 60 
 
 Heat equivalent of work done in the cylinder 
 
 Heat equivalent of work done on brake 
 15'6 x 33000 
 
158 GAS-ENGINE TRIAL. 
 
 Heat Received by Engine per Minute. Calorific value of 
 gas, as given by calorimeter, and corrected for hydrogen 
 only = 615 B.T.TJ. Correcting for pressure and tempera- 
 ture, we have 
 
 Calorific value at 32 deg. Fah. and 147 
 
 Heat received per minute by engine 
 
 Heat efficiency, as calculated on equivalent of work done 
 in cylinder, 
 
 = 7 ^L = 0186 = 19 per cent. 
 416o 
 
 Heat efficiency, as calculated on work done on brake, 
 
 = MA = 0151 = 15'4 per cent 
 41oo 
 
 HEAT ACCOUNT. 
 
 Heat received by engine 
 per minute 4,165 
 
 4,165 
 
 Thermal equivalent of 
 
 work done 793 
 
 Heat lost in jackets... 1,503 
 Heat lost in exhaust 
 gases 1,848 
 
 4,144 
 
 It is not unusual to find that the heat account gives a 
 slight excess of heat accounted for. The temperature of the 
 exhaust has been calculated at the point of opening of the 
 va lve that is, at 88 per cent of the working stroke. Some of 
 this heat is transmitted to the jacket water, and is therefore 
 measured twice. On the other hand, large losses due to 
 radiation are not measured. As before stated, the tempera- 
 ture of the exhaust gases is a rather uncertain quantity. 
 It therefore not infrequently works out that a percentage 
 
GAS-ENGINE DESIGN. 159 
 
 of heat is unaccounted for on one or other side of the heat 
 account. 
 
 During the progress of a trial it is extremely useful to 
 plot all quantities as they are observed upon a large sheet 
 of squared paper. Errors of observation are thus easily 
 discovered in time to be rectified. A break in the plotted 
 curves will indicate either a mistake in the readings or 
 fluctuations in the conditions of the trial. It is important 
 that either should be checked. The person responsible for 
 the trial will learn more from a casual glance at these curves 
 than by attempting to check each observer individually. 
 
 CHAPTER XVII. 
 GAS-ENGINE DESIGN. 
 
 IN writing upon the subject of gas-engine design, it will be 
 assumed that the general arrangements of a gas engine are 
 already familiar to the reader. Hitherto the calculations 
 required to determine the sizes of the various essential 
 parts have been entirely excluded from the literature upon 
 the subject. Considering the difference existing between 
 the gas engine and the steam engine, it is important that 
 calculations determining the proportions of the former 
 should be dealt with quite independently. We propose, 
 therefore, to work out, from data already established, the 
 leading dimensions of a gas engine to develop 20 horse 
 power on the brake. 
 
 The mechanical efficiency of a gas engine may reasonably 
 be supposed to reach 80 per cent. This is, of course, some- 
 times exceeded, but it is wise to underestimate, as no engine 
 will develop its highest efficiency unless perfectly adjusted, 
 and in the exigencies of practice this should not be relied 
 
1GO GAS-ENGINE DESIGN. 
 
 upon. If, therefore, we assume 80 per cent mechanical 
 efficiency, we shall require a cylinder capable of developing 
 
 ^- = 25 indicated horse power. 
 
 o 
 
 In steam-engine design, our next subject for considera- 
 tion would be the maximum boiler pressure at our disposal. 
 From this, with due regard to the number of expansions per- 
 missible, we should ultimately draw a prospective indicator 
 diagram, and so obtain the mean effective pressure per square 
 inch of piston area. Now, the pressure obtained in a gas- 
 engine cylinder depends upon three factors : (1) the mixture 
 of air and gas ; (2) the quality of the gas ; (3) the density 
 of the mixture before ignition takes place. A fourth 
 factor namely, the mean working temperature of the 
 cylinder would, if variable, greatly affect the pressure. 
 The temperature is, however, necessarily reduced by the 
 presence of the jacket water to an almost constant limit. 
 Early designers of gas engines did not attempt high 
 compression; thus, up to the year 1890, the compression 
 seldom exceeded 40 Ib. per square inch. This, however, has 
 been gradually raised, until now the compression on small 
 engines amounts to over 90 Ib. per square inch. It is 
 inadvisable to exceed, or even approach, this figure in 
 designing large power gas engines, say over 50 I.H.P., on 
 account of the liability to ignition during the compression 
 stroke. 
 
 This risk is, however, much reduced when a scavenging 
 arrangement clears the hot gases from the combustion 
 chamber, and replaces them with cooler air. In round 
 numbers, we may take it that the maximum pressure 
 obtained will be 3 '5 times the pressure before ignition. 
 Thus, compressing up to 60 Ib. per square inch, we may 
 expect, with the usual proportions of air to gas, a maximum 
 pressure of 210 Ib. Although this is the maximum pressure 
 obtainable when running at full power, it must not be 
 forgotten that when the governor cuts out the gas supply 
 
GAS-ENGINE CYLINDER DESIGN. 161 
 
 for one or perhaps more cycles, a particularly dense mixture 
 may be drawn into the cylinder. The ignition of a strong 
 mixture may thus produce a maximum pressure far in 
 excess of that calculated. For this reason a large margin is 
 necessary. 
 
 There is a marked similarity between all indicator 
 diagrams from gas engines, and from a comparison of a large 
 number of diagrams it will be found that the mean effective 
 pressure produced is roughly equal to 2 C - O'Ol C 2 when 
 C = compression in pounds per square inch above atmo- 
 spheric pressure. These figures may be varied one way or 
 other by the valve setting and cylinder proportions.* 
 
 The expansion curve of an indicator diagram may be 
 raised by decreasing the time of expansion and by reducing 
 the cylinder surface to a minimum. The first mentioned 
 condition reduces the time during which heat may be trans- 
 ferred from the burning gases to the jacket water, and the 
 second condition reduces the surface by means of which the 
 transmission of heat is facilitated. Let us first consider 
 the effects of cylinder proportions, and engine speed, upon 
 the rate of expansion, and in so doing we will assume that 
 the diameter of the cylinder and power developed remain 
 constant, whilst the revolutions per minute and the length 
 of stroke are variable. 
 
 The number of feet travelled by the piston per minute 
 is limited. For practical reasons it is unadvisable to exceed 
 700ft per minute. When L = length of stroke in feet, 
 and R = revolutions per minute, then we have the limiting 
 value of piston speed 
 
 2 L R = 700 ft. 
 
 *It will be observed that this formula gives a maximum mean pressure of 
 100 lh. per square inch when the compression reaches 100 ib. It has been shown 
 by Messrs. Mallard and Le Cbatelier that the rate of cooling follows the law 
 expressed by a Q + /3 2 where a and |8 are constants, and d = temperature. It 
 is interesting to note that the above formula, deduced from actual indicator 
 diagrams, shows that the mean effective pressure follows a similar law. The 
 formula holds only up to 100 Ib. compression, and must f-nly be regarded as giving 
 approximate values, for neither the mixture nor quality of gas form factors of 
 the expressioa. 
 12G 
 
162 GAS-ENGINE CYLINDER DESIGN. 
 
 It is evident that a variety of dimensions might be 
 chosen for L and R. As R is increased, L is diminished. 
 As R is increased the number of impulses per minute may 
 be increased, and, consequently, the time of each expansion 
 is diminished. The volume of gas used per cycle is diminished, 
 but the maximum pressure obtained need not be, if the same 
 compression be given to the mixture. Thus we see that a 
 smaller volume of gas gives the same maximum pressure per 
 square inch as a larger volume, and, further, it is expanded 
 in very much less time. If R be increased continuously, 
 and L correspondingly diminished, then at some value of R 
 the outstroke of the piston will take place in less time than 
 the charge can be effectually ignited. Let us investigate 
 the limit of speed of any engine in which the connecting rod 
 is five times the crank length. 
 
 Let T = time required for explosion pressure to arrive at 
 
 its maximum ; 
 
 C = length of crank radius in feet ; 
 R = revolutions per minute of engine j 
 then mean speed of crank pin 
 
 = 2y .Q R f t er secondt 
 
 The mean velocity of the piston during the first one-tenth 
 of its forward stroke will be very nearly 
 
 = 27r 6 P R x 0'32 ft per second. 
 
 The distance travelled 
 
 _ stroke = ^C = 
 
 Therefore time taken by piston in travelling one-tenth stroke 
 
 0-2 C 
 
 2 TT C K x 0'32 
 60 
 
 5 9 
 
 = -yf- (nearly) seconds. 
 K 
 
GAS-ENGINE CYLINDER DESIGN. 163 
 
 Now, in order that the engine may fully expand the hot 
 gases, the maximum pressure should not be reached at a 
 point later than one-tenth of the stroke. Even so late as 
 this is disadvantageous. Hence the lowest value of 
 
 -^ must equal T. 
 K 
 
 Mr. Dugald Clerk ascertained that the time required to 
 reach the maximum pressure, with a mixture of coal gas 
 and air in the proportion of 1 to 5 by volume, was 005 
 second. In his experiments the initial temperature of the 
 mixture was low, and the initial pressure was atmospheric. 
 The author has found that the time of explosion is greatly 
 diminished by raising the initial pressure, but has not yet 
 succeeded in measuring accurately by how much. Witz 
 found that the duration of explosion of a mixture of 1 to 6 '3, 
 behind a freely-moving piston, to be 0'06 second. As the 
 ignition took place at atmospheric pressure, these results 
 can hardly be applied to modern gas engines. It is probable 
 that the maximum number of revolutions approaches the 
 limit at 500 per minute. Putting this value for R in the 
 equation, 
 
 5'9 _ T 
 
 TT = 
 
 we find T = 0*0118 second. We might here appropriately 
 consider the effect of lead on the igniting valve. The total 
 time of rise of pressure in a gas-engine cylinder may be 
 divided into two distinct parts. Firstly, the time taken for 
 the flame to strike back into the mixture, and, secondly, 
 the time during which the pressure rises after this has 
 been accomplished. The former effect may be largely 
 compensated for by giving lead to the ignition valve, 
 but the latter cannot be dealt with in this way without 
 seriously increasing the liability to what is known as " back 
 explosion." The severe strains occasioned by such back 
 explosion should, of course, be avoided. In the absence of 
 experimental data with regard to the values of T, we shall 
 do well to accept the limiting speed of revolutions as 
 
164 GAS-ENGINE CYLINDER DESIGN. 
 
 approaching 500, In practice it is found advisable to run 
 large engines at a speed much below 500 (about 160) revolu- 
 tions, because of the excessive vibration due to the rapid 
 movements of reciprocating parts, and the consequent 
 stresses brought to bear upon them. It is also probable that 
 the value of T is much increased when large cylinders are 
 used. A large nun_ber of smaller engines run at from 250 to 
 300 revolutions. 
 
 The result of our investigation has, so far, led to the con- 
 clusion that the speed approaches a limit at 500 revolutions 
 per minute, and it has been further shown that this is 
 independent of the length of stroke, on the assumption that 
 the maximum pressure shall be reached during the first 
 tenth of the stroke. We have now to consider the ratio of 
 length of stroke to diameter of cylinder. 
 
 It is certain that the loss of heat increases in a greater 
 proportion than the difference in temperature between the 
 burning gases and the cylinder walls. Also the loss of heat 
 is greater, the greater the density of the charge. In other 
 words, the loss of heat is very much greater during the 
 beginning of the stroke than at any other time. If it be 
 true that the loss of heat varies directly as the density t 
 and directly as the surface, and roughly as the difference in 
 temperature, then we may express the loss of heat in a 
 time tx as 
 
 (surface x difference in temperature x density) d t. 
 
 As, however, the density varies inversely as the volume, we 
 may write the expression thus 
 
 B /Vsurface x difference in temperature \ . 
 J \ volume / 
 
 o 
 
 A and B being constants. 
 
 From this it is evident that the ratio of surface to volume 
 should be a minimum near the beginning of the stroke. To 
 
GAS-ENGINE CYLINDER DESIGN. 165 
 
 investigate the subject further, by attempting to evaluate 
 the quantities in the above expression, would be rather 
 begging the question, inasmuch as the temperature differences 
 must be estimated by references to the diagrams taken from 
 existing engines. Experiments might be devised to furnish 
 data upon this subject, but the author is not aware that such 
 independent data at present exists. 
 It is the opinion of Mr. Hamilton, the patentee of the 
 
 Premier gas engine, that the ratio sur ace should be a mini- 
 mum at about one-third of the stroke. Supposing, therefore, 
 that the combustion chamber were one-third the length of 
 the stroke, then we should require a minimum ratio of 
 
 sur ace wnen j. ne volume is actually equal to (area of piston 
 
 volume 
 
 x two-thirds stroke). Now, it is easily shown that the ratio 
 
 sur ace j g ft m inimum when the diameter of a cylinder is 
 volume 
 
 equal to its length.* Thus we see that a minimum ratio of 
 
 * Let x = diameter, c = volume (constant). Then, length 
 
 _4_c 
 
 X- IT 
 
 Surface = + _i . ; IT 
 
 OC" 7T 
 
 - = * + *.! 
 
 2 x 
 d S = - d (x 2 ) + 4 c d (_ j = 2 %x dx dx 
 
 dx ~ ~tf' 
 
 When surface is a minimum C =0. 
 d x 
 
 . . TT x* - 4 c . ' . 4 c = x :i TT. 
 
 = 4 c 
 
 x~ IT' 
 and, by putting 4 c in terms of x, we have length 
 
 But length 
 
 X L 7T 
 
 Hence length = diameter when surface is a minimum. 
 
166 GAS-ENGINE CYLINDER DESIGN. 
 
 Bur ace j g gj ven w h en the diameter is = two-thirds stroke. 
 
 volume 
 
 This, as may be supposed, is a favourite proportion. The 
 
 author is convinced that the effect of large cylinder surface 
 
 is very marked. 
 
 Cylinders of large dimensions have a much larger propor- 
 tion of volume to surface than those of small dimensions, 
 even though the proportions be chosen to favour a minimum 
 of surface in both cases. It is therefore to be expected that 
 large size gas engines will give greater economy than smaller 
 ones. This has been fully realised in practice. 
 
 We have discussed very briefly the controlling factors 
 determining the size of cylinder for a gas engine, and we 
 now proceed to apply the conclusions arrived at. We shall 
 base our calculated size upon the following data : 
 
 1. Piston speed, 500ft. per minute, 
 
 2. Compression before ignition, 80 Ib. per square inch, 
 
 3. Stroke of engine, 1^ times diameter of cylinder, 
 
 Let D = diameter of cylinder. Then stroke = 1*5 D 
 Revolutions per minute 
 
 piston speed _ 500 
 2 x stroke 3D 
 
 Explosions per minute (on Otto cycle) 
 
 _ revolutions _ 500 
 
 2 ~ 6D~ 
 
 Mean effective pressure may be taken approximately as 
 
 = 2 C - O'Ol C 2 (when C = compression pressure) 
 = 160 - O'Ol x 6400 
 = 96 Ib. per square inch. 
 
 The indicated horse power is to equal 25. Hence we have 
 
GAS-EXGINE CYLINDER DESIGN. 167 
 
 whence D 2 = 87'5 (nearly) ; 
 
 . '. D = 9 '35 in., say 9in. diameter of cylinder. 
 Stroke therefore equals 1'5 x 9'35 = 14 in., nearly. 
 
 Size of Combustion Chamber to give 80 Ib. Compression 
 (above Atmosphere). 
 
 Let V = whole volume, 
 
 = volume of combustion chamber -f volume swept 
 
 out by piston. 
 
 P = absolute pressure in pounds per square inch. 
 Then the following equation holds good for the compression 
 
 curve 
 
 PV 1 ' 3 = constant. 
 
 To simplify the figures, the numerical value of V may be 
 represented by linear inches, for, when the combustion 
 chamber is the same diameter as the cylinder, the movement 
 of the piston is proportional to the volume swept out. 
 Hence the volume of the combustion chamber may be 
 represented by 
 
 V - stroke, 
 = V - 14. 
 
 At the commencement of the compression stroke the value 
 of P will equal about 14 Ib. per square inch absolute. 
 
 At the end of the compression stroke the pressure required 
 is equal to 80 -f 15 = 95 Ib., absolute. We have therefore 
 
 Taking P = 14 Ib. per square inch, P x = 95 Ib. per square 
 inch, and V x = (V - 14), we have 
 
 14V 1 ' 3 = 95(V -14) 1 ' 3 
 then V 1 ' 3 =678(V - 14) 1 ' 3 
 
 and V = 1 ' / J/ = 678(V - 14) 1 ' 3 
 
 = (V - 14) r y "678 
 = (V - 14) 4-36, nearly, 
 from which 3'36 V = 61 ; 
 
108 CE.ANK SHAFTS. 
 
 Thus the actual volume of the combustion chamber, of 
 any shape whatever, must 
 
 = (181 - stroke of piston) x area of cylinder 
 = (181 - 14) 5fjf 
 = 283 cubic inches. 
 
 With respect to the arrangement of valves, it must be 
 pointed out that the surface of all passages leading into the 
 cylinder should be reduced as much as possible. The size of 
 valves should be such that the velocity of the gases, as 
 calculated upon the mean piston speed, is not more than 
 100ft. per second. Although it is not possible to fully 
 discuss all the details of cylinder design in the space at our 
 disposal, it is hoped that we have sufficiently indicated how 
 to determine the leading dimensions of a cylinder to develop 
 a given horse power. By reference to the illustrations and 
 descriptions of engines already given, there should be no 
 difficulty in setting out the leading dimensions of an engine 
 cylinder. It must not be forgotten, however, that the success 
 of an engine depends upon minor details, a knowledge of 
 which can only be acquired by practical experience in the 
 working of gas engines. We make no attempt at describing 
 the thousand and one minor details which will be readily 
 supplied by the practical draughtsman, our object being 
 rather to place before the reader those facts and figures 
 which are not acquired in the erecting shop. 
 
 The Crank Shaft. 
 
 A gas-engine crank shaft should be made from the very 
 best mild steel procurable. The severe stresses upon the 
 shaft render it imperative that a large margin for strength 
 be allowed. The author has known cases of gas-engine 
 crank shafts of large diameter working successfully for 
 many monthp, but which have suddenly failed, without 
 apparent cause. 
 
CRANK SHAFTS. 169 
 
 It is desirable that the crank webs be balanced by weight 
 on the crank rather than on the flywheels. When the speed 
 is high, and the flywheels of large diameter, balance weights 
 fitted or cast on the wheels cause considerable oscillation of 
 the whole engine. The bearings should be as close together 
 as possible, and have ample surface. 
 
 To ascertain the true maximum twisting moment on a 
 crank shaft when under normal conditions of working, it is 
 necessary to correct the indicator diagram for the inertia of 
 the reciprocating parts. This done, a polar diagram of 
 twisting moment may be plotted for various positions of 
 the crank. We have already stated that the maximum 
 pressure upon the piston will be approximately = 3 '5 times 
 the compression pressure. Thus, in the example under 
 consideration, we should expect a pressure of 80 x 35 = 
 280 Ib. per square inch. This pressure occurs almost on the 
 dead centre of the crank, and decreases as the tangential 
 effort increases. We shall not be far wrong in taking 70 
 per cent of the maximum pressure as the load producing 
 maximum twisting moment. Thus the maximum tangential 
 effort may be taken as compression pressure x 2 '5. When 
 the bearings are close up to the crank webs, bending stresses 
 may be neglected. 
 
 If D = diameter of shaft required, skin stress on the 
 shaft 8,000 Ib., 
 then we have 
 
 - D 3 . 8000 = 80 x 2'5 x 7 x 69 ; 
 
 D 3 = *? > L 80 _ > L?J> x 7 x 69 . 
 314 x 8000 
 
 from which D = 395, say 4 in. diameter. 
 
 We are aware that this rule gives an exceptionally large 
 diameter of shaft. We are, however, inclined to believe that 
 the wisdom of an engine builder may be measured in terms 
 of his crabk-shaft dimensions, and we are therefore inclined 
 to favour maximum dimensions. 
 
170 CKANK SHAFTS. 
 
 The crank webs should be of ample size to give rigidity. 
 The depth of the webs is, of course, largely determined by 
 the diameter of the shaft. A good rule is to make the depth 
 of web equal to 
 
 whilst the thickness of each web should be 
 
 where D = diameter of shaft. 
 
 The crank pin should not be less in diameter than 1/2 D, 
 the length being determined by the pressure upon the crank 
 pin. In a gas engine (unlike a steam engine) the pressure 
 upon the piston falls very rapidly ; hence the maximum load 
 per square inch may exceed that of steam-engine practice. 
 The maximum pressure on a gas-engine crank pin should 
 not exceed 1,000 Ib. per square inch ; but, with due regard 
 to this, the load per square inch, as calculated upon the 
 average piston pressure, may reach 400 Ib. 
 
 Applying the rules given, our crank-shaft dimensions for 
 the case under discussion will be as follow : 
 
 Depth of web 
 
 = 3-95 + ?|? = 5-26, say S^in. 
 o 
 
 Thickness of each web 
 
 = 3-95 + ?|5 = 3-46, say 3^ in. 
 
 o 
 
 The diameter of crank pin 
 
 = 3'95 x 1-2 = 474, say 4f in. 
 Then length of crank pin (calculated on maximum pressure) 
 
 = 69 x 280 = ^ 
 1000 x 4 75 
 
 Then length of crank pin (calculated on average pressure) 
 69 x 96 
 
 400 x 475 
 
 3-48. 
 
FLYWHEELS. 171 
 
 Hence the length should be (say) 4| in., to satisfy the worst 
 condition. 
 
 The main bearing pressures are largely affected by the 
 weight of the flywheels, for the maximum pressure upon 
 the bearings is the resultant of a thrust upon the crank 
 and a vertically applied load due to the flywheels. A safe 
 rule, which comprehends all conditions, is to allow 100 Ib. 
 per square inch, as calculated upon the mean piston pressure. 
 Thus, in the example before us, we should have ample 
 bearing surface by making the total lengh of bearings 
 
 69 x 96 
 
 100 x 4 
 
 = n. 
 
 Dimensions of Flywheels Required for Gas Engines. 
 
 In calculating the weight of gas-engine flywheels, we 
 shall neglect the effect of inertia of reciprocating parts. 
 Having regard to the fact that the wheels must drive for 
 three strokes out of every four without undue fluctuation of 
 speed, the effect of inertia becomes insignificant. 
 
 Let R x = maximum velocity, expressed as revolutions 
 
 per minute ; 
 R 2 = minimum velocity, expressed as revolutions 
 
 per minute ; 
 Vj = maximum velocity in feet per second at the 
 
 mean diameter of the wheels ; 
 V 2 = minimum velocity in feet per second at the 
 
 mean diameter of the wheels ; 
 W = weight of flywheel in pounds. 
 
 Then mean revolutions per minute 
 
 _ R! + R 2 
 
 2 
 
 Average work done per stroke 
 
 H.P. x 33,000 H.P. x 33000 
 
 + R 
 
172 FLYWHEELS. 
 
 Let n = number of strokes the engine runs without impulse. 
 (It may here be noted that no allowance need be made for the 
 work done in compression during the idle strokes, for the 
 energy thus absorbed is always given back again during 
 the out strokes. ) 
 
 The energy to be stored in flywheels 
 
 = H.P. x 33000 x n 
 R A + R 2 
 
 Let the fluctuations in speed be represented as 
 Is = C = 5i . 
 
 then V 2 = C V lf 
 
 and R 2 = C R x . 
 
 Now, the energy absorbed during n revolutions, without 
 impulse to the flywheels, must not reduce the speed below 
 R 2 or V 2 . Hence energy absorbed must equal 
 
 ** \y J u 
 
 Let D = mean diameter of wheels, then 
 
 Hence, by substituting this value of V^, we have work 
 done per n stroke = energy derived from fly wheels that is, 
 
 (1 - C 2 ), 
 
 H.P. x 33000 x n _ W D 2 ^ R t 2 
 R! + R 2 2 g ' 3600 
 
 and putting R 2 = C RX, we have 
 
 w= H.P. x 33000 x n x 2 g x 3600 
 R! (1 + C) D 2 *-2 IV (l - C 2 )' 
 
 and, reducing the weight to tons, we have 
 W (in tons) = 343900 H ' R x n 
 
 R x 3 x D 2 (1 - C 2 ) (1 + C)' 
 
FLYWHEELS. 173 
 
 We have here taken D as the mean diameter, instead of 
 twice the radius of gyration ; the error is, however, insigni- 
 ficant, and is more than compensated for by the absence of 
 complications in the formula. The value of c for electric 
 light work should be taken as 0'98, thus allowing a total 
 variation of speed of 2 per cent. It is, of course, desirable 
 to keep the flywheels as light as possible, not only on 
 account of cost, but with the object of increasing the 
 mechanical efficiency. For this latter reason, wheels of 
 large diameter are to be preferred, inasmuch as the required 
 weight varies inversely as the square of the diameter. 
 Although not usual in gas-engine practice, it might be 
 desirable to box in the arms of large wheels, to prevent 
 losses due to air resistance. If flywheels are made too large, 
 they become dangerous on account of centrifugal action on 
 the rim of the wheel. For safety, the outside of the wheel 
 should be limited to a speed of 100 ft. per second. 
 
 With respect to the value of n, we may say that it is 
 undesirable to work a gas engine missing fire oftener than 
 alternate cycles. This is indeed highly important with 
 large engines, because small leakages of unburnt gas pass 
 through the cylinder, and if not fired frequently, serious 
 explosions are thereby liable to occur in the exhaust pipes. 
 It is much safer, even though accompanied by loss in 
 efciency,to set the governor and gas valves to fire the engine 
 every cycle at half load. We may take the value of m 
 therefore, as being equal to seven strokes. 
 
 Taking the following values for the symbols, we can find 
 the weight of flywheels for a 25 indicated horse power 
 engine. 
 
 putting 
 
174 CONNECTING fcOD. 
 
 we have 
 
 W = 343900 x 25 xj^^ 
 215 3 x 5 2 x 0-04 x 1-08 
 
 Thus, if two flywheels be used, they should weigh a little 
 more than l tons each. We have neglected the weight 
 of the spokes and boss, but as the error favours steady 
 running, there is no necessity for further complications. 
 
 Connecting Rods. The determination of connecting-rod 
 sizes for high-speed engines has already been fully dealt with 
 in books on engine design. We may refer the reader to the 
 chapter on connecting rods in Professor Unwin's treatise 
 on machine design. Regarding gas engines as high-speed 
 engines, we shall quote a formula from the above-mentioned 
 work, which we think gives a result in agreement with 
 practice. We are quite aware that many connecting rods 
 will be found working satisfactorily which are nevertheless 
 under the size given by the formula. Considering the 
 complexity of the conditions, it is impossible, without 
 devoting much space and re-writing much that has already 
 been written, to deduce a formula embracing all conditions. 
 The value obtained by the following rule may be excessive 
 for slow-speed engines, but it may be regarded as safe for 
 all cases. 
 
 Let d = mean diameter of connecting rod ; 
 D = diameter of cylinder ; 
 I = length of connecting rod ; 
 
 p = initial pressure on the piston in Ibs. per square 
 inch ; 
 
 then d = 0038 J{VlJp}. 
 
 Having obtained a mean diameter, the rod may, if more 
 convenient, be made of oval section. The mean diameter 
 may be increased by about i in, towards the crank-pin end 
 of the rod, and diminished by the same amount at the piston 
 end. The length of rod may be from five to six times the 
 crank length. 
 
M lx xV 
 
 OF TBK ^\ 
 
 UNIVERSITY ]| 
 
 PISTON PIN SIZES. 175 
 
 For the example under consideration 
 
 d = 0-038 x / {9 35 x 35 x/280} 
 = 2 "8 in. 
 
 Hence the diameters may be, say, 2 in. and 3 in. 
 
 The Piston Pin. 
 
 The pressure on the piston pin may be much more than 
 upon the crank pin, on account of the small relative move- 
 ment of the rod at this end. A pressure of 600 Ib. per square 
 inch, as calculated upon the mean effective pressure, will 
 give suitable proportions to the pin. The length of the pin 
 may be about 1'4 x diameter. It must here be noted that 
 the design of any part should be carried out with due regard 
 to the facilities in machining. For this reason it might be 
 convenient to have the small end of the connecting rod the 
 same width as the larger end, thus avoiding troublesome 
 packings to set the rod level when machining. 
 
 The size of pin required will be as follows : 
 
 Total mean pressure 
 
 = 09 x 96 Ib. 
 Bearing area in square inches 
 
 _ 60x96 
 
 600 
 
 Diameter of cylinder is 9 in. Allowing 5| in. for the total 
 length of boss to receive the pin, we may take the length as 
 
 9| - 5J = 35. 
 Then diameter of pin 
 
 The pin should be secured in the bosses of the trunk piston 
 by set screws, to prevent rotation therein. A drip lubricator 
 should be arranged to supply oil to the pin when working. 
 
 The piston should carry from three to six cast-iron packing 
 rings, according to the pressure, of about |J x section. 
 
176 PRODUCES, GAS. 
 
 The side shaft should be driven by screw gearing, the 
 designs for which may be taken from any treatise upon 
 gearing. The governor should be driven by bevel gearing 
 from the side shaft, if requiring rotary motion. 
 
 We have already given data required for sizes of pipes, 
 tanks, and fittings external to the engine itself. Let us 
 conclude these few notes on gas-engine design with one word 
 of warning. Never rely upon lock nuts without split pins 
 to prevent them working completely off. We have known 
 serious accidents to result from overlooking this apparently 
 small detail. The vibration of all combustion engines 
 renders it absolutely necessary that steady pins and positive 
 lock nuts should be freely used in connecting the parts. 
 
 CHAPTER XVIII. 
 PRODUCER GAS, AND ITS APPLICATION TO GAS ENGINES. 
 
 IN the early days of the gas engine, one of the chief points 
 in favour of its extended use was the convenience with 
 which the ordinary coal gas of town supplies could be 
 utilised as its motive power. So long as the power developed 
 by gas engines was small, the commercial advantages 
 resulting from their use would have been seriously mini- 
 mised by the necessity of a large capital outlay. Thus, in 
 the early days of the gas-engine industry, the motive power 
 was entirely supplied from the ordinary town mains. Gas 
 engines had been established as a commercial success for 
 some four or five years before any other means of supplying 
 them with motive power was recognised as practicable or 
 economical. Nor was this state of things entirely due to 
 lack of invention. It is indeed surprising how many names 
 are associated with the development of a practical gas 
 generator, suitable not only for gas engines, but for many 
 
PRODUCER GAS. 177 
 
 industrial processes, and especially for metallurgical pur- 
 poses. In view of this, we shall merely explain the broad 
 principles upon which all gas producers work ; but we shall 
 refer specially to the work of Mr. Emerson Dowson, whose 
 name has for many years been associated with the practical 
 application of gas-producing plant for the supply of motive 
 power to gas engines. 
 
 Any combustible substance when heated to a high tempera- 
 ture, either by the process of its own combustion or by the 
 application of external heat, will be partially or wholly 
 gasified. The most familiar example of this process is the 
 ordinary household fireplace. The sudden blaze which so 
 often results from stirring the fire is merely the burning of 
 those gases which are rising continually through the red- 
 hot fuel, but which are prevented from burning when passing 
 through the hot fuel because of the inferior supply of 
 oxygen. When, however, these gases mingle with the fresh 
 air passing above the surface of the fuel, the oxygen renders 
 them combustible. If the fire be dull, the heat at this point 
 is insufficient to ignite the gases, and much of the gaseous 
 fuel passes away unburnt. If, however, the fire be stirred 
 to allow more air to enter its upper strata, then the gases 
 are kindled by the heat, and the flame, thus started, rapidly 
 spreads. 
 
 It will be seen from this that a simple gas producer might 
 be made by charging a vessel with red-hot coke, then passing 
 through it a stream of oxygen sufficient to form a combustible 
 gas, but insufficient to allow complete combustion while 
 passing through the red-hot fuel. This gas might be col- 
 lected in a suitable reservoir, and burnt at will by the 
 addition of a further supply of oxygen. The process of 
 manufacture of the gas is represented symbolically by 
 
 C0 2 + C = 2 CO. 
 
 The resulting gas, carbon-monoxide, has a calorific value 
 of about 340 B.T.U. per cubic foot. For commercial con- 
 sumption the oxygen is supplied to the hot coke or carbon 
 
 13G 
 
178 PRODUCER GAS. 
 
 by passing air through the containing vessel. The resulting 
 gas is therefore very largely diluted with nitrogen, which 
 reduces its calorific value per cubic foot to about 112 B.T.U., 
 or about one- sixth that of coal gas. 
 
 It is clear that, so long as atmospheric air alone is used as 
 a vehicle for the oxygen necessary for combination with the 
 carbon, a very inferior gas will be produced. In order to 
 avoid this, another method has been resorted to, the result 
 of which is the production of a gas known as water gas. 
 The bare outline of the operation is as follows : Atmospheric 
 air is passed through a chamber containing coke or anthracite 
 until the whole mass of fuel is at a bright red heat. The 
 gases formed during this process are not collected. The air 
 supply is then entirely cut off, and superheated steam is 
 passed through the incandescent fuel. This steam is decom- 
 posed into its elements, oxygen and hydrogen, and the 
 process may continue until the temperature of the fuel falls 
 to a point below which no further decomposition is possible. 
 By this process as much as 50 per cent of free hydrogen is 
 obtained, whilst the liberated oxygen re-combines with the 
 carbon to form about 44 per cent of carbon-monoxide, 
 together with traces of free oxygen, carbon-dioxide, and 
 marsh gas. The calorific value of this gas is about 240 
 B.T.U. per cubic foot, and, moreover, the percentage of 
 combustibles approaches 100. Hence this gas is eminently 
 suited for gas-engine purposes, and would be largely used 
 excepting for the intermittent action of the apparatus. In 
 generators of this gas the production can only continue for 
 periods of about 15 minutes. After this time steam must be 
 turned off, and air again passed through the fuel, in order to 
 raise it to the incandescent state. Some success has been 
 attained by duplicating this apparatus, but the additional 
 capital outlay required, and the introduction of simpler 
 methods of equal efficiency as regards gas production, has 
 prevented the general adoption of intermittent producers. 
 
 The method so successfully carried out by Mr. Emerson 
 Dowson embraces a combination of the above-described 
 
PRODUCER GAS. 
 
 179 
 
 processes. From a specially-constructed nozzle a jet of 
 steam is blown into the incandescent fuel. The steam 
 induces a current of air to enter the fuel at the same time. 
 The oxygen thus continuously supplied to the fire maintains 
 it at a proper temperature, whilst the resulting gas is greatly 
 enriched by the decomposition of the steam ; and, moreover, 
 by the judicious regulation of the weight of air and steam 
 passing through the fuel, the process is continuously carried 
 on. 
 
 FIG. 85. Dowson gas plant. 
 
 Figs. 85 and 86 show an elevation and plan of Dowson gas 
 plant, of which the following is a brief description. 
 
 Referring to fig. 85, the boiler for generating the steam is 
 shown at A. This boiler is kept at a pressure of from 30 Ib. to 
 60 Ib. per square inch, according to the size of the generator. 
 Superheated steam passes over from the boiler through the 
 pipe O to the gas generator shown at C. Beneath the firebars 
 of the generator the steam passes through a nozzle of special 
 construction, and so draws air with it into the closed ashpit 
 
180 
 
 PRODUCES GAS. 
 
 E. The gases resulting from the contact of the air and 
 steam with the incandescent fuel in C pass up through the 
 delivery pipes to the stand pipes F. The gases are here 
 cooled, and are afterwards cleansed by passing through 
 water in the box H. From H the gases pass to the scrubber 
 J, and afterwards through the coke scrubber K placed inside 
 the gasholder L, where they remain until conducted to the 
 gas engine. The pressure in the gasholder is allowed to 
 remain at about 1| in. of water, and the pressure produced 
 
 FIG. 86. Plan of Dowson gas plant. 
 
 in the closed ashpit by the action of the steam jet iff 
 sufficient to maintain proper circulation of the gases from 
 the producer to the gasholder. The supply of steam to the 
 generator is regulated by a cock at O, the handle of which 
 is attached to the upper moving portion of the gasholder L, 
 thus automatically preventing the generation of gas when 
 the holder becomes fully charged. 
 
 The fuel most suitable for consumption in the generator 
 is anthracite, because of the absence of sulphur, smoke, and 
 
PRODUCER GAS. 181 
 
 other impurities. Its cost may be taken at from 12s. to 15s. 
 per ton delivered, and the fuel required per 1,000 cubic feet 
 of gas equals about 13'51b. Hence the cost of fuel in the 
 generator per thousand cubic feet may be taken as about 
 one penny. Including all charges, depreciation, wages, &c., 
 upon a small plant capable of delivering 1,000 cubic feet per 
 hour, the cost is worked out at 4|d. per 1,000 cubic feet.* 
 A larger plant, producing 3,000 cubic feet per hour, works 
 out to about 2-J-d. per 1,000 cubic feet. It must here be 
 remembered, however, that about five volumes of Dowson 
 gas are required to evolve the same heat as one volume of 
 coal gas. Hence the cost per 1,000 cubic feet must be 
 multiplied by five before a comparison can be made between 
 Dowson and coal gas. Thus we see that with large pro- 
 ducers a saving of 67 per cent is effected, and with small 
 producers a saving of 41 per cent, when coal gas is taken at 
 3s. per 1,000. 
 
 The lowest fuel consumption obtained with Dowson plant, 
 working a twin-cylinder Crossley engine, indicating 118 
 horse power, was 76 Ib. per indicated horse power hour 
 during a working test of eight hours. When the losses in 
 the generator due to clinkering and standing all night were 
 added, the total consumption was 08731b. per indicated 
 horse power hour. The cost per indicated horse power hou^ 
 exclusive of wages and incidental expenses, is therefore 
 O'OGd. This exceptionally low consumption of fuel can 
 hardly be expected under ordinary working conditions. 
 The average consumption may be taken as 1^ Ib. of anthracite 
 per indicated horse power hour. 
 
 Referring to analyses previously given of Dowson gas, it 
 will be seen that it consists largely of carbon-monoxide, 
 which gas is very poisonous. It is therefore important that 
 all fittings should be tight, and that there should be good 
 ventilation in the engine-room. 
 
 * lust. Civil Engineers, Proceedings, vol. Ixxiii. 
 
182 
 
 PRODUCER GAS. 
 
 It has already been mentioned that the calorific value of 
 producer gas is much lower than that of coal gas ; also that 
 the air volume required for its combustion is less than for 
 coal gas. Although some difficulty is found in practice in 
 keeping the composition of the gas quite uniform, owing to 
 choking up of the firebars, &c., yet an average composition 
 of Dowson gas will be found to require, theoretically, I'Ol 
 volumes of air to 1 of gas. In gas engines, 1'5 volumes of 
 air to 1 of gas are usually adopted. 
 
 The following table, worked out similarly to that given 
 previously for coal gas, will be found useful in working out 
 the results of trials. For the purposes of these calculations, 
 an average analysis is quoted : 
 
 
 
 "3 
 
 j>> 
 
 z> <-< 
 
 k'S'S 
 
 -^ *s 
 
 
 Ivolnme pe 
 cent. 
 
 li, 
 
 D bo 
 
 Prr portion 
 weight. 
 
 J3 3 g^-u 2J 
 
 
 Weight of o 
 gen requi 
 per pounc 
 gas. 
 
 -Id 
 
 Hvdro^en 
 
 20 
 
 o-ooin 
 
 0-008 
 
 58"27 
 
 0-064 
 
 0-01924 
 
 Carbon-monoxide .. 
 
 23 
 
 01SOO 
 
 0-136 
 
 77-40 
 
 0-077 
 
 0-02352 
 
 Carbon-dioxide 
 Nitrogen 
 
 7 
 50 
 
 0-07420 
 0-03915 
 
 0-561 
 0*295 
 
 % 
 
 .... 
 
 0-OP593 
 0-05103 
 
 
 
 
 
 
 
 
 
 100 
 
 6-13246 
 
 1-00 
 
 135-67 
 
 0-141 
 
 0-18972 
 
 From the above table we see that 0'1411b. of oxygen ia 
 required to combine with 1 Ib. of Dowson gas. Hence 
 
 0141 
 0-23 
 
 = 0-613 Ib. 
 
 of air will be required to supply this weight of oxygen, 
 lib. of air occupies (at 32 deg. Fah. and 147 Ib.) "08059 
 cubic feet. Volume of air required therefore 
 
 0-613 
 008059 
 
 = 7 '61 cubic feet (nearly). 
 
EFFECTS OF THE PRODUCTS OF COMBUSTION. 183 
 
 1 lb. of Dowson gas occupies 
 
 - = 7 '56 cubic feet (nearly). 
 
 013246 
 The ratio of air to gas theoretically required is, therefore, 
 
 ^i = l-01(ne rly). 
 
 7 'o(5 
 
 CHAPTER XIX. 
 
 THE EFFECTS OF THE PRODUCTS OF COMBUSTION UPON 
 EXPLOSIVE MIXTURES OF COAL GAS AND Am 
 
 EXPERIMENTS previously made upon gaseous mixtures have 
 been directed towards the investigation of the actual 
 pressures produced by the combustion of an inflammable 
 gas, in the presence of oxygen or pure air only. Thirty 
 years ago such experiments were conducted in chemical 
 laboratories on a very small scale quite incomparable with 
 the volumes of the cylinders which are used in practical 
 work. The practical difficulties which beset the develop- 
 ment of the gas engine retarded, rather than stimulated, 
 any very complete research upon the behaviour of explosive 
 mixtures. Practical men were satisfied with an approxima- 
 tion to the maximum pressure which might be expected 
 with any given mixture ; this information was amply pro- 
 vided by Hirn, Bunsen, and later by Berthelot. As the 
 mechanical arrangements became more efficient, the need was 
 felt for further data regarding the comparative economy of 
 various mixtures. 
 
 The most complete practical contribution upon this 
 subject has been afforded by the experiments of Mr. Dugald 
 Clerk, which enabled him to estimate the most economical 
 mixture to be used in a non-compression engine, but no 
 account was taken of the effects of the products of combus- 
 tion which are present in the cylinder of a gas engine, 
 
184 EFFECTS OF THE PRODUCTS OF COMBUSTION. 
 
 notwithstanding that the early engines were constructed 
 with a clearance volume of 60 per cent. To obtain some 
 definite data upon this important subject the author has re- 
 cently carried out a series of experiments in the Engineering 
 Laboratory of the Yorkshire College, Leeds. 
 
 It has been generally inferred that the products of com- 
 bustion, when mixed with a fresh charge of coal gas and 
 air, will decrease the maximum pressure and thereby 
 reduce the efficiency of the charge ; if this inference be 
 correct, it would justify the efforts of gas-engine manu- 
 facturers in introducing a scavenger stroke, in order to 
 minimise the evils of the presence of residual gases. The 
 experiments carried out by the author show that the 
 presence of the products of combustion in certain mixtures 
 actually raise, rather than diminish, the maximum pressure 
 obtained. It is, however, incontestible that the expulsion 
 of the products of combustion has effected an increase in the 
 economy of gas consumption ; but, in the author's opinion, 
 the reduced consumption per horse power is due to the 
 increase in the effective cylinder capacity of any given 
 engine when the products of combustion are replaced by 
 explosive mixture, as well as to the cooling effects of the 
 scavenging stroke. Following the example of previous 
 experimenters upon gaseous mixtures, the author made use 
 of a closed vessel of constant volume, and ignited the charges 
 at atmospheric pressure. The use of such apparatus is 
 considered by some to detract from the practical value of the 
 results obtained, but if the presence of certain constituents 
 affects the rise of pressure in a vessel of constant volume, it 
 may be asked, Why should it not also affect the rise of 
 pressure in the expanding chamber of a gas engine 1 The 
 author is now continuing this series of experiments with 
 compressed mixtures, the result of which he hopes shortly 
 to publish. 
 
 The apparatus used for the explosion of the gases con- 
 sisted of a thick cast-iron cylinder, flanged at both ends, of 
 
EFFECTS OF THE PRODUCTS OF COMBUSTION". 
 
 185 
 
 1 cubic foot capacity. The cylinder (A, fig. 87) was bolted 
 vertically to the column B. 
 
 Igniting Arrangements. The charge was ignited by 
 passing an electric spark in the mixture by means of a 
 secondary battery and induction coil, the wires from which 
 passed through the insulated brass plug C on the top cover. 
 
186 
 
 EFFECTS OF THE PRODUCTS OF COMBUSTION. 
 
 The insulation at first gave a great deal of trouble through 
 moisture collecting on the under surface of the insulating 
 material at the point where the wires penetrated (see 
 fig. 88, a). The difficulty was ultimately overcome by 
 adopting the arrangement shown in fig. 88, b. The brass 
 casing, by projecting beyond the surface of the insulation, 
 
 / ^ 
 
 I'rftfical Engineer 
 
 FIG. 88. The Igniter. 
 
 prevented the water, during the filling of the cylinder, from 
 rising to within the cup thus formed. By thoroughly 
 drying, then oiling the surface of the insulation, the spark- 
 ing was made certain. If any mixture failed to explode, 
 the igniting gear was carefully tested, and the inflammatory 
 nature of the gas ascertained by applying flame as it was 
 gently driven from the experimental cylinder. 
 
 Temperatures. The temperatures were measured, before 
 ignition, by means of a thermometer enclosed in a wrought- 
 iron case, containing mercury, which penetrated into the 
 gas chamber. 
 
EFFECTS OF THE PRODUCTS OF COMBUSTION. 
 
 187 
 
 Rt cor ding Gear. The pressures were recorded by means 
 of a Crosby indicator, the pencil of which was arranged to 
 scribe upon a continuously-revolving drum, Sin. diameter, 
 driven by clockwork. The exact speed of this drum was 
 checked by the vibrating spring, fig. 89, adjusted to make 
 four complete oscillations per second, the dimensions of 
 the spring being such that the inertia enabled it to over- 
 come the slight friction of the pen during one experiment. 
 By allowing this pen to remain stationary during one 
 revolution of the drum, a zero line was traced, which was 
 crossed at every eighth of a second by the wave line 
 produced by the vibrating spring. This period of time is 
 represented by a linear distance of 0'3 in., and thus further 
 sub-divisions may be made if desired. It was inconvenient 
 to arrange the pen of the time recorder immediately over 
 
 FIG. 89. The Recorder. 
 
 the indicator pencil, but due correction has been made by 
 transferring the required portion of the time wave to its 
 proper position on the diagram. 
 
 In all the experiments the volumes were measured by filling 
 the cylinder with water, and afterwards allowing the gas to 
 enter as a measured quantity of water flowed out. After 
 
188 EFFECTS OF THE PKODUCTS OF COMBUSTION. 
 
 firing the charge, a known volume of water was injected into 
 the cylinder (care being taken that no air was inhaled by the 
 partial vacuum formed by the condensation of the previous 
 charge), thus expelling all but the required volume of 
 residual gas ; this, together with a fresh supply of air, and 
 the same volume of coal gas as before, formed the next 
 charge. It need hardly be pointed out that the presence 
 of products of combustion reduced the proportion of gas to 
 pure air, for if this ratio had remained constant during the 
 whole series, it would have necessitated a reduction in the 
 quantity of coal gas, and consequently a reduction of the 
 maximum explosion pressure. Undoubtedly the best 
 arrangement would have been one in which the cylinder 
 
 FIG. 90. Indicator diagram, one volume coal gas, eight volumes air. 
 
 itself could have been subject to alteration in volume, for 
 each experiment, sufficient to contain the quantity of 
 products which might be present. The conclusions which 
 may be deduced from these experiments involve the careful 
 consideration of this point. The gases for each experiment 
 were taken into the cylinder in the following order : 
 
 1. The products of the previous combustion, if any. 
 
 2. Half the volume of pure air. 
 3 The coal gas. 
 
 4. The remaining air required to complete the charge. 
 
 No appreciable time was allowed for the diffusion of the 
 mixture, it having been fired immediately after taking the 
 temperature. It is impossible to estimate how the diffusion 
 
EFFECTS OF THE PRODUCTS OF COMBUSTION. 
 
 180 
 
 of the gases in a working gas-engine cylinder is assisted by 
 the rapid motion of the piston, though it is probable that 
 the coal gas and air are more intimately mixed with one 
 another than with the residual gases. In all experiments 
 great care was taken to obtain atmospheric pressure in the 
 cylinder, special precaution being taken to allow the excess 
 of pressure of the coal gas to escape after uncoupling the 
 supply. 
 
 The coal gas used throughout the experiments was taken 
 from the service pipes of the Leeds gas supply. From 
 analyses of three samples made by the author after the 
 experiments, the composition was found to be as follows : 
 
 ANALYSIS OF LEEDS COAL GAS. 
 
 
 
 4J 
 
 ^ 
 
 
 $-1 
 
 _ 
 
 M K 
 
 
 
 
 
 8 
 
 & 
 
 | 
 
 a 
 
 
 
 +J-( 
 
 2 
 
 
 ^3 
 
 .2 
 
 3 
 
 1 
 
 
 
 
 '5 % 
 
 g 
 
 
 8 
 
 a 
 
 g 
 
 
 1 
 
 3 
 
 >, O'C 
 
 H 
 
 Constitueuts. 
 
 
 "8 
 
 = 5 
 
 
 
 
 y3 
 
 
 ji? <1> c 
 
 a *" 
 
 
 
 Volume 
 
 I 
 
 II 
 
 1 
 
 a 
 g 
 
 Calori 
 pounc 
 
 A 
 
 
 
 ft) 
 
 j_> ,-J 
 
 ^ ft 
 
 
 
 LV'S. 
 
 Lbs. 
 
 
 B.TU. 
 
 
 
 Lbs. 
 
 
 35-2 
 
 0*0447 
 
 "015 T3 
 
 0*514 
 
 21 690 
 
 11 148 
 
 4 
 
 2*056 
 
 Olefines 
 
 4-2 
 
 0-1174 
 
 00493 
 
 0*161 
 
 20 260 
 
 3 261 
 
 >* 
 
 
 Hydrogen 
 
 52'9 
 
 0-00559 
 
 00295 
 
 0*096 
 
 52 500 
 
 5 040 
 
 8 
 
 0*768 
 
 Carbon monoxide. . . 
 
 65 
 
 0-0783 
 
 0050S 
 
 0166 
 
 4300 
 
 713 
 
 I 
 
 0-094 
 
 Nitrogen 
 
 o-i 
 
 0783 
 
 00078 
 
 0-025 
 
 
 
 
 
 Carbon dioxide and 
 
 
 
 
 
 
 
 
 
 oxygen 
 
 11 
 
 0-1060 
 
 00116 
 
 O'OSS 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 100-0 
 
 
 
 03063 
 
 1-000 
 
 
 
 20,162 
 
 
 3-47 
 
190 EFFECTS OF THE PRODUCTS OF COMBUSTION. 
 
 The mixtures experimented upon were as follow : 
 
 Volume of coal 
 gas. 
 
 Volumes of 
 air. 
 
 Volume of coal 
 gas. 
 
 Volumes of 
 air. 
 
 *1 
 
 *16 
 
 1 
 
 12 
 
 1 
 
 15 
 
 1 
 
 10 
 
 1 
 
 14 
 
 1 
 
 8 
 
 1 
 
 18 
 
 1 
 
 fi 
 
 * Failed repeatedly to explode. 
 
 Keferring to Table 1. (Appendix), it will be seen that after 
 igniting a mixture of 1 volume of coal gas and 15 volumes 
 of air, the residue was added to the next charge in the 
 proportion of 5 per cent of the cylinder volume. Column 5 
 gives the actual pressure above the atmosphere as recorded 
 by the indicator. In calculating the calorific values of the 
 volumes of coal gas present in the experiments, the original 
 temperatures and barometric pressures were allowed for ; 
 in other cases, the slight variations in initial temperature 
 and barometric pressures have been neglected. 
 
 In Column VI. is recorded the rise of pressure above, or 
 fall below, that obtained with a pure mixture, when the 
 various percentages of residue were added. The greatest 
 rise of pressure, namely, 19 Ib. per square inch, took place 
 when the products of combustion amounted to about 30 per 
 cent of the whole cylinder volume, which left a proportion 
 of coal gas to pure air of about 1 to 9^ by volume. The air 
 was again decreased by the addition of residual gases until, 
 in the proportion of 1 volume of gas to 6 of air, the rise in 
 pressure was only 8 Ib. per square inch above that recorded 
 for the pure mixture. Fifty-five per cent of products was 
 found to be the greatest quantity that might be introduced 
 without preventing ignition. The results given in 
 Table II. show that when the mixture was originally in 
 the proportion of 1 volume of gas to 14= of air the explosions 
 were somewhat less regular than in the previous cases, and 
 
EFFECTS OF THE PRODUCTS OF COMBUSTION. 
 
 191 
 
 the maximum rise of pressure was greatest when the 
 residual gases were present in the proportion of from 10 to 
 30 per cent, but only 12 Ib. per square inch rise above the 
 normal was recorded in this series. 
 
 The accompanying diagram, fig. 91, has been prepared 
 from the figures in Column VI. of the tables. The ordinates 
 above the line A B represent the excess of pressure above 
 that recorded when mixtures of pure air and gas were 
 exploded. Below the line the fall in pressure is plotted. 
 
 FIG. 91. Diagram showing rise and fall of maximum pressure due to the 
 addition of products of combustion. 
 
 The abscissae represent the percentages of residue present. 
 The curve marked a represents the rise of pressure with a 
 mixture of 15 to 1. Similarly, the curve marked 6, with 
 a mixture of 14 to 1. When the volume of coal gas is as 
 great as one-tenth of the cylinder volume, the addition of 
 residual gases immediately reduces the pressure below that 
 of a pure mixture. The general conclusions which may be 
 be drawn from this diagram are as follow : That with weak 
 mixtures the maximum pressure is least when the excess of 
 pure air is greatest. That the maximum pressure, obtained 
 from a given quantity of coal gas, rises as the excess of air 
 
192 
 
 EFFECTS OF THE PRODUCTS OF COMBUSTION. 
 
 is diminished by the addition of products of combustion ; 
 but this no longer holds when the volume of pure air to 
 coal gas approaches the proportion of 10 to 1. 
 
 From an examination of fig. 92 it is seen that the explosion 
 becomes more rapid as the excess of air is replaced by 
 neutral gases. If the time is short compared with that of 
 the outstroke of the motor, then the expansion line of an 
 indicator diagram, taken under ordinary conditions, would 
 be below the adiabatic curve. If, on the contrary, the 
 burning continues during the whole of the outstroke, the 
 expansion curve would probably be above the adiabatic, 
 and would consequently give a greater mean effective 
 pressure. Although in the diagrams before us we have only 
 the curve due to cooling, we may still estimate the relative 
 economy of a charge by measuring the area enclosed by the 
 curve of pressure from its commencement to a point 
 determined by an ordinate of time, which shall be chosen 
 equal to that of the outstroke of a motor. Assuming the 
 
 PRACTICAL ENGINtEfl. 
 
 Fio. 92. Diagram showing increase in the rate of combustion due to the 
 addition of neutral gases. 
 
 rather slow speed of 25 second as being the duration of the 
 outstroke, we obtain the following mean pressures derived 
 from the areas of the pressure-time diagrams up to 025 
 second. It appears from the following calculations that 
 the mean pressure is not influenced by the quantity of the 
 products of combustion present, but it is always a maximum 
 
EFFECTS OF THE PRODUCTS OF COMBUSTION. 
 
 193 
 
 when pure air is present in the proportion of about 10 to 1 
 of gas. 
 
 TABLE OF GREATEST MEAN PRESSURES, CALCULATED FROM 
 PRESSURE-TIME DIAGRAMS, UP TO 0'25 SECONDS FROM 
 COMMENCEMENT OF EXPLOSION. 
 
 Proportion 
 of volume of 
 
 coal gas to 
 volume of 
 cylinder. 
 
 Greatest mean \ . 
 pressure recorded, f 1 
 
 I % 
 Pounds per square I 
 
 inch. ) 
 
 The percentage 
 of products 
 
 (estimated as , a ~- 
 clearance in an I -g 
 
 engine) was / 
 
 111 
 
 The proportion 
 
 of pure air 
 
 to coal gas, 
 
 of 
 
 * 1 
 
 9 
 20 
 
 None 
 30 
 
 15 to 1 
 11 tol 
 
 * { 
 
 11 
 "23 
 
 None 
 20 
 
 14tol 
 11 tol 
 
 * { 
 
 11 
 
 21 
 
 None 
 18 
 
 13 tol 
 11 tol 
 
 " { 
 
 10 
 21 
 
 None 
 18 
 
 12 tol 
 10 tol 
 
 * { 
 
 32 
 33 
 
 None 
 
 17-8 
 
 11 tol 
 
 8'3 to 1 
 
 * { 
 
 42 
 40 
 
 None 
 11 
 
 8 tol 
 
 7 to 1 
 
 * { 
 
 30 
 26 
 
 None 
 5-2 
 
 6 tol 
 
 5-6 tol 
 
 From the composition of the coal gas used, it has been 
 calculated that 57 volumes of air are required for complete 
 combustion ; and the experiments show that, as long as this 
 ratio is not widely departed from, the mixture is explosive, 
 notwithstanding the presence of 55 per cent of inert gases. 
 
 The results of all the experiments point conclusively to 
 the fact that the ratio of air to gas alone determines the 
 possibility of an explosion, and that the explosion pressure 
 depends chiefly upon a suitable ratio being chosen. 
 
 The total heat evolved in each of the experiments was as 
 follows : 
 Uo 
 
194 
 
 EFFECTS OF THE PRODUCTS OF COMBUSTION. 
 
 Heat values of the 
 volumes of coal 
 gas used. 
 
 Proportion of volume 
 of gas to volume of 
 cylinder. 
 
 Volume of air to 
 volume of gas. 
 
 B.T.U. 
 
 37 
 
 Per cent. 
 6-2 
 
 15 to 1 
 
 39 
 
 6-6 
 
 14 to 1 
 
 42 
 
 7-1 
 
 13 to 1 
 
 45 
 
 7-7 
 
 12 to 1 
 
 53 
 
 91 
 
 10 to 1 
 
 66 
 
 111 
 
 8 to 1 
 
 86 
 
 14-2 
 
 6 to 1 
 
 It was found, as might be anticipated, that the explosion 
 pressures varied directly as the number of thermal units 
 generated from the combustion of the gas. In fig. 93, the 
 explosion pressures have been plotted as ordinates, and the 
 number of British thermal units as abscisase. 
 
 The diagrams in all cases show an alteration in the rate 
 of combustion, in two places, but in no instance is there any 
 sign of an actual reduction in pressure merely an alteration 
 in the rate of increase. Mr. Dugald Clerk, in his work 
 upon the "Explosion of Gaseous Mixtures," produces an 
 indicator diagram taken during the explosion of a mixture 
 of one volume of gas to five volumes of air, in which the 
 pressure curve distinctly falls after having reached a pressure 
 of 60 Ib. per square inch ; it then rises to its maximum of 
 nearly 100 Ib. In none of the diagrams taken by the author, 
 or in any single instance, out of some hundreds of experi- 
 ments made by the Students of the College, has any such 
 fall been recorded. It is therefore impossible to believe in a 
 theory which attempts to explain it. Mr. Clerk's theory 
 supposes that, after complete inflammation has taken place 
 the pressure is further raised as the constituents of the coal 
 gas comb'ne with oxygen, only at a much slower rate than 
 during inflammation. So far this may be satisfactory, but 
 
EFFECTS OF THE PRODUCTS OF COMBUSTION. 
 
 195 
 
 to believe that this explains the momentary fall in pressure 
 is, in the author's opinion, carrying the theory beyond its 
 legitimate application. 
 
 It is found from an examination of the diagrams taken 
 during the explosions of pure mixtures of air and coal gas, 
 that the first alteration in the slope of the rising pressure 
 curve occurs at 0'4 of the maximum height of the diagram. 
 Whatever may be the real cause of this characteristic of all 
 
 4 
 
 30- 
 
 40 
 
 So 60 70 60 
 
 Fia. 93. Diagram showing relation between maximum pressure 
 obtained and thermal units present in the charge. 
 
 coal-gas explosion diagrams, it is interesting to note that 
 the heat generated by the combustion of the hydrogen and 
 the defines combined is found to be just 0'4 of the total heat 
 of the whole of the constituents of the coal gas experimented 
 upon. Experiments have been made by Messrs. V. Meyer 
 and F. Fryer, to determine the temperature at which each 
 of the constituents of coal gas combkie with oxygen. The 
 results of their experiments are as follow : 
 
196 EFFECTS OF THE PRODUCTS OF COMBUSTION. 
 
 IGNITION TEMPERATURES OF EXPLOSIVE GASEOUS 
 MIXTURES BURNT IN A CLOSED BULB. 
 
 Hydrogen combined with oxygen at 1,124 deg. Fah. 
 
 Olefines combined with oxygen at 1,124 deg. Fah. 
 
 Marsh gas combined with oxygen at 1,201 deg. Fah. 
 
 Carbon monoxide combined with oxygen at 1,347 deg. Fah. 
 
 It will here be noted that the hydrogen and olefines burn 
 at the lowest temperature, namely, 1,124 deg. Fah., that the 
 marsh gas burns next in order, and lastly the carbon 
 monoxide at the highest temperature. From the calorific 
 values of the constituents of the gas, which are given in a 
 previous table, 1 Ib. of coal gas is found to contain 
 
 8,301 thermal units due to hydrogen and olefines. 
 1 1,148 thermal units due to marsh gas. 
 
 7l3 thermal units due to carbon monoxide. 
 
 20,162 total. 
 
 The total heat may, therefore, be divided into three parts, 
 the proportion of each to the whole being 
 
 0'417 of total heat due to hydrogen and olefines. 
 0'548 of total heat due to marsh gas. 
 0*035 of total heat due to carbon monoxide. 
 
 The annexed diagram (fig. 94) shows the application of 
 these figures to the results. This diagram was taken 
 during the explosion of 1 volume of gas mixed with 12 
 volumes of air. If horizontal lines are drawn through the 
 points A, B, C, the figure is divided into three bands. If 
 the specific heat be sensibly constant, then the pressures 
 produced will be proportional to the heat developed at a 
 given time. If, then, 0'41 of the total heat were developed 
 first, we should expect to find the height of the first band to 
 be 0'41 of the total height of the diagram. From several 
 diagrams taken with pure mixtures the mean height of this 
 band was found to be 0'403 of the total height. In the 
 shaded diagram illustrating this point, the heights of the 
 
EFFECTS OF THE PRODUCTS OF COMBUSTION. 
 
 197 
 
 bands are not quite in agreement with the previous figures, 
 but very nearly so. The diagrams, taken when the residual 
 gases are present in large volumes, show the point A is very 
 much raised, probably on account of the hydrogen and 
 defines combining with the greater part of the oxygen 
 present, the remainder being insufficient to completely burn 
 the marsh gas and carbon monoxide. In other words, the 
 total height of the diagram is not what it would have been 
 had the heat due to the marsh gas and carbon monoxide 
 been wholly developed. 
 
 PRACTICAL ENGINEER.. 
 
 FIG. 94. 
 
 In every experiment the pressures recorded were low 
 compared with those obtained by Clerk. Fig. 94 shows a 
 maximum pressure of 38 Ib. per square inch above the 
 atmosphere, from which the maximum temperature works 
 out to 1,350 deg. Fah. The maximum temperature possible 
 was 3,250 deg. Fah., showing a loss of heat of 58*5 per cent, 
 or 41'5 per cent accounted for, and the maximum pressure 
 correspondingly higher. The difference is no doubt due to 
 the fact that water was present on the walls of the cylinder, 
 much of the heat thereby rendered latent in converting it 
 into vapour. The pressure at the point A on fig. 6 is about 
 18 Ib. per square inch above the atmosphere, and the heat 
 accounted for at this point is found to be 49 per cent of that 
 due to the hydrogen and olefines. That the heat accounted 
 for at points on the pressure line near the atmospheric is 
 more than at the maximum is what we should anticipate, 
 for the cooling curves are all much steeper at the top thai* 
 
198 
 
 EFFECTS OF THE PRODUCTS OF COMBUSTION. 
 
 lower, showing that the rate of transmission of heat is 
 greatest at the highest temperature. 
 
 A summary of the whole serie of these experiments is 
 shown by fig. 95. Plotted to a scale of 20 Ib. = 1 in. are the 
 pressures recorded during each experiment. The dotted 
 line passes through all the points of pressure ; the full line 
 shows the mean rise and fall of pressure throughout. 
 
 Diagram showing Volumes of Coal Gas, Air, and Xeutral Gases, with 
 Maximum Pressures in pounds per square inch for each mixture. 
 
 Coal Gas Volumes, per cent. 
 FIG. 95. 
 
 CONCLUSIONS. 
 
 From these combined results the following conclusions 
 may be drawn : 
 
 1. That the highest pressures are obtained when the 
 volume of air present is only slightly in excess of the 
 amount required for complete combustion. 
 
EFFECTS OF THE PRODUCTS OF COMBUSTION. 199 
 
 2. That higher pressures are recorded when residual 
 gases take the place of an excess of air. 
 
 3. That when the volume of the products of combustion 
 does not exceed 58 per cent of the mixture, then it is 
 explosive, provided the volume of air is not less than 5 '5 
 times the volume of coal gas. 
 
 4. That the time of an explosion is much reduced when 
 excess of air is replaced by products of combustion. 
 
APPENDICES TO PART I. 
 
 TABLES OF RESULTS OF EXPERIMENTS ON THE EFFECTS OF 
 THE PRODUCTS OF COMBUSTION UPON EXPLOSIVE 
 MIXTURES OF COAL GAS AND AIR. (REFERRED TO IN 
 CHAPTER XLX.) 
 
 TABLE I. 
 
 Col. I. 
 
 Col. II. 
 
 Col. III. 
 
 Col. IV. 
 
 Col. V. 
 
 Col. VI. 
 
 
 
 
 
 
 Rise <i 
 
 Ratio of 
 volume of 
 coal gas to 
 volume of 
 cylinder. 
 
 Ratio of 
 volume of 
 coal gas to 
 volume of 
 air. 
 
 Ratio of 
 volume of 
 products to 
 volume of 
 cylinder. 
 
 Ratio of 
 volume of 
 products to 
 volume of 
 air + coal gas. 
 
 Maximum 
 pressures, 
 pounds per 
 square inch. 
 
 pressure 
 above (+) 
 or fall 
 below (-) 
 maximum 
 for pure 
 
 
 
 
 
 
 mixture. 
 
 Actu'J 
 
 Per 
 
 ct-nt 
 
 Actual 
 
 Per 
 
 cent 
 
 Actu'l 
 
 Per 
 cent. 
 
 Actu'l 
 
 Per 
 
 cent. 
 
 
 
 1/16 
 
 62 
 
 1/15 
 
 6-6 
 
 
 
 
 
 
 
 
 
 16 
 
 .. 
 
 ii 
 
 i) 
 
 1/14-2 
 
 7-0 
 
 1/20 
 
 5 
 
 1/19 
 
 5-2 
 
 22 
 
 + 6 
 
 ii 
 
 
 
 1/13-2 
 
 7-4 
 
 1/10 
 
 10 
 
 1/9 
 
 11-1 
 
 34 
 
 + 18 
 
 II 
 
 ii 
 
 1/12-6 
 
 7'9 
 
 1/6-6 
 
 15 
 
 1/5-6 
 
 17-8 
 
 34 
 
 + 18 
 
 1) 
 
 ii 
 
 1/ll-S 
 
 8-4 
 
 1/5 
 
 20 
 
 1/4 
 
 25 
 
 34 
 
 + 18 
 
 II 
 
 ,, 
 
 l/ll 
 
 90 
 
 1/4 
 
 25 
 
 1/3-3 
 
 30 
 
 34 
 
 + 18 
 
 !1 
 
 ii 
 
 1/10'2 
 
 9-8 
 
 1/3-3 
 
 30 
 
 1/2-3 
 
 45 
 
 35 
 
 + 19 
 
 fl 
 
 ,, 
 
 1/8-6 
 
 11-6 
 
 1/2-5 
 
 40 
 
 1/1-5 
 
 66 
 
 28 
 
 + 12 
 
 
 
 >i 
 
 1/7-8 
 
 12-8 
 
 1/2-2 
 
 45 
 
 1/1-2 
 
 89 
 
 32 
 
 + 16 
 
 
 
 
 
 1/7 
 
 14-3 
 
 1/2 
 
 50 
 
 1/1 
 
 100 
 
 32 
 
 + 16 
 
 1) 
 
 ,, 
 
 1/6-1 
 
 16-2 
 
 1/1-8 
 
 55 
 
 .. 
 
 
 24 
 
 + 8 
 
 II 
 
 ii 
 
 1/5-3 
 
 18-6 
 
 1/1-6 
 
 60 
 
 
 
 Failed. 
 
 
 TABLE II. 
 
 1/15 ; 6-6 
 
 1/14 
 
 7'1 
 
 
 
 
 
 
 
 24 
 
 .. 
 
 II 
 
 
 
 1/13-3 
 
 75 
 
 1/20 
 
 5 i 1/19 
 
 5-2 
 
 24 
 
 
 
 > 
 
 ! 
 
 1/12-5 
 
 8 
 
 1/10 
 
 10 j 1/9 
 
 in 
 
 33 
 
 + 9 
 
 II 
 
 ,, 
 
 1/11-8 
 
 S-4 
 
 1/6-6 
 
 15 ; 1/5-6 
 
 17-8 
 
 34 
 
 + 10 
 
 II 
 
 
 
 1/11 
 
 9-1 
 
 1/5-0 
 
 20 
 
 1/4 
 
 25 
 
 36 
 
 + 12 
 
 II 
 
 
 
 1/9-7 
 
 10-3 
 
 1/4 
 
 25 
 
 1/3-3 
 
 30 
 
 28 
 
 + 4 
 
 M 
 
 Jt 
 
 1/9-5 
 
 10-4 
 
 1/3-3 
 
 30 
 
 1/2-3 
 
 45 
 
 36 
 
 + 12 
 
 )t 
 
 ,, 
 
 1/8 
 
 12-4 
 
 1/2-5 
 
 40 
 
 1/1-5 
 
 66 
 
 31 
 
 + 7 
 
 11 
 
 II 
 
 1/6-6 
 
 15-1 
 
 1/2 
 
 50 
 
 1-1 
 
 100 
 
 33 
 
 + 9 
 
 II 
 
 II 
 
 1/6-2 
 
 16-1 
 
 1/1-9 
 
 52 
 
 .. 
 
 .. 
 
 35 
 
 + 11 
 
 II 
 
 II 
 
 1/5-7 
 
 17-3 
 
 1/1-8 
 
 55 
 
 .. 
 
 .. 
 
 23 
 
 - 1 
 
 M 
 
 II 
 
 1/5-2 
 
 19-2 
 
 1/1-7 
 
 58 
 
 .. 
 
 .. 
 
 25 
 
 + 1 
 
 " 
 
 
 
 1/5 
 
 19-9 
 
 1/1-6 
 
 60 
 
 
 
 F^lerl. 
 
 
APPENDICES TO PART I. 
 
 201 
 
 TABLE III. 
 
 Col. I. 
 
 Col. II. 
 
 Col. III. 
 
 Col. IV. 
 
 Col. V. 
 
 Vol. VI. 
 
 
 
 
 
 
 Rise of 
 
 Ratio of 
 volume of 
 coal gas to 
 volume of 
 cylinder. 
 
 Ratio of 
 volume of 
 coal gas 
 to volume 
 of air. 
 
 Ratio of 
 * olume of 
 products to 
 volume of 
 cyliuder. 
 
 Ratio of 
 volume of 
 products to 
 volume of 
 air + coal gas. 
 
 Maximum 
 pressures, 
 pounds per 
 square inch. 
 
 pressuro 
 above (-f ) 
 or fall 
 below (-) 
 maximum 
 for pure 
 mixture. 
 
 Actu'l 
 
 1/14 
 
 Pel- 
 cent 
 7'1 
 
 Actual 
 1/13 
 
 Per 
 
 c-nt 
 
 7-7 
 
 Actu' 
 
 
 Per 
 cent 
 
 
 \ctu' 
 
 
 Per 
 
 cent 
 
 
 31 
 
 
 ,, 
 
 
 
 1/12-3 
 
 8-1 
 
 1/20 
 
 5 
 
 1/19 
 
 5-2 
 
 27 
 
 4 
 
 ,, 
 
 ,, 
 
 1/11-6 
 
 86 
 
 1/10 
 
 10 
 
 1/9 
 
 11-1 
 
 40 
 
 + 9 
 
 ,, 
 
 ,, 
 
 1/11 
 
 9-1 
 
 1/6-6 
 
 15 
 
 1/5-6 
 
 17-8 
 
 37 
 
 + 6 
 
 
 
 ,, 
 
 1/10-2 
 
 9-8 
 
 1/5 
 
 20 
 
 1/4 
 
 25 
 
 39 
 
 + 8 
 
 ,, 
 
 ,, 
 
 1/9-5 
 
 10-5 
 
 1/4 
 
 25 
 
 1/3-3 
 
 30 
 
 30 
 
 1 
 
 
 
 ,, 
 
 1/8-8 
 
 11-3 
 
 1/33 
 
 30 
 
 1/2-3 
 
 45 
 
 34 
 
 + 3 
 
 ,, 
 
 ,, 
 
 1/8-1 
 
 12-3 
 
 1/2-8 
 
 35 
 
 1/1-8 
 
 54 
 
 37 
 
 . +6 
 
 ,, 
 
 ,, 
 
 1/7-4 
 
 13-o 
 
 1/2-5 
 
 40 
 
 1/1-5 
 
 66 
 
 87 
 
 + 6 
 
 ,, 
 
 
 
 1/6-7 
 
 14-9 
 
 1/22 
 
 45 
 
 1/1-2 
 
 89 
 
 36 
 
 + 5 
 
 ,, 
 
 ,, 
 
 1/6 
 
 16-9 
 
 1/2 
 
 50 
 
 1/1 
 
 100 
 
 37 
 
 + 6 
 
 " 
 
 " 
 
 l/5'3 
 
 18-9 
 
 1/1-8 
 
 55 
 
 I/ 
 
 
 Failed. 
 
 
 TABLE IV. 
 
 1/13 
 
 7-7 
 
 1/12 
 
 8-3 
 
 
 
 
 
 
 
 | 36 
 
 .. 
 
 ,, 
 
 ,, 
 
 1/11-3 
 
 8-8 
 
 1/20 
 
 5 
 
 1/19 
 
 5-2 | 36 
 
 
 
 ,, 
 
 ,, 
 
 1/107 
 
 9-3 
 
 1/10 
 
 10 
 
 1/9 
 
 11-1 j 40 
 
 + 4 
 
 ,, 
 
 ,, 
 
 1/10 
 
 9-9 
 
 1/6-6 
 
 15 
 
 1/5-6 
 
 17-8 
 
 42 
 
 + 6 
 
 ,, 
 
 ,, 
 
 1/9-4 
 
 10-6 
 
 1/5 
 
 20 
 
 1/4 
 
 25 
 
 41 
 
 + 5 
 
 
 
 ii 
 
 1/8-7 
 
 11-4 
 
 1/4 
 
 25 
 
 1/33 
 
 30 
 
 43 
 
 + 7 
 
 ,, 
 
 ,, 
 
 1/8-1 
 
 12-3 
 
 1/3-3 
 
 30 
 
 1/2-3 
 
 45 
 
 35 
 
 -1 
 
 >> 
 
 ,, 
 
 1/68 
 
 147 
 
 1/2-5 
 
 40 
 
 1/1-5 
 
 66-7 
 
 39 
 
 t 3 
 
 
 
 ,, 
 
 1/5-5 
 
 18-3 
 
 1/2 
 
 50 
 
 1/1 
 
 100 
 
 32 
 
 4 
 
 it 
 
 ,, 
 
 1/4-8 
 
 20 8 
 
 1/1-8 
 
 55 
 
 
 .. 
 
 34 
 
 2 
 
 ,, 
 
 ,, 
 
 1/4-4 
 
 227 
 
 1/17 
 
 58 
 
 
 
 Failed. 
 
 .. 
 
 " 
 
 " 
 
 1/4-1 
 
 24-3 
 
 1/1-6 
 
 60 
 
 
 
 Failed. 
 
 
 
202 
 
 APPENDICES TO PART I. 
 
 TABLE V. 
 
 Col. I. 
 
 Col. II. 
 
 Col. III. 
 
 Col. IV. 
 
 Col. V. 
 
 Col. vr. 
 
 
 
 
 
 
 Rise of 
 
 Ratio of 
 volume of 
 coal gas to 
 volume of 
 cylinder. 
 
 Ratio of 
 volume of 
 coal gas 
 to volume 
 of air. 
 
 Ratio of 
 volume of 
 products to 
 volume of 
 cylinder. 
 
 Ratio of 
 volume of 
 products to 
 volume of 
 air+ coal gas. 
 
 Maximum 
 pressures, 
 pounds per 
 square inch. 
 
 pressure 
 above (+) 
 or fall 
 below (-) 
 maximum 
 for pure 
 mixture. 
 
 Actu'l 
 1/11 
 
 Per 
 
 cent 
 9-1 
 
 Actual 
 1/10 
 
 Per 
 
 cent 
 10 
 
 Actu'J 
 
 
 Per 
 
 cent 
 
 
 Actu'J 
 
 
 Per 
 
 cent 
 
 
 48 
 
 
 
 
 ,, 
 
 1/9-4 
 
 10-6 
 
 1/20 
 
 5 
 
 1/19 
 
 5-2 
 
 49 
 
 + 1 
 
 
 
 ,, 
 
 1/8-9 
 
 11-2 
 
 1/10 
 
 10 
 
 1/9 
 
 11-1 
 
 48 
 
 
 
 
 
 ,, 
 
 1/83 
 
 12 
 
 1/6-6 
 
 15 
 
 1/5-6 
 
 17-S 
 
 50 
 
 + 2 
 
 ,, 
 
 ,, 
 
 1/7-8 
 
 12-8 
 
 1/5 
 
 20 
 
 1/4 j 25 
 
 45 
 
 3 
 
 
 
 
 
 1/7-2 
 
 13-8 
 
 1/4 
 
 25 
 
 1/3-3 
 
 30 
 
 44 
 
 4 
 
 it 
 
 ,, 
 
 1/6-7 
 
 14-9 
 
 1/3-3 
 
 30 
 
 1/2-3 
 
 45 
 
 47 
 
 1 
 
 11 
 
 
 
 1/5-6 
 
 17-9 
 
 1/2-5 
 
 40 
 
 1/1-5 
 
 66-7 
 
 45 
 
 3 
 
 ,, 
 
 
 
 1/5 
 
 19-8 
 
 1/2-2 
 
 45 
 
 1/1-2 
 
 82 
 
 35 
 
 13 
 
 
 
 ,, 
 
 1/4-9 
 
 20-2 
 
 1/2-1 
 
 46 
 
 1/1-7 
 
 85 
 
 
 
 
 
 
 
 
 1/4-7 
 
 21'3 
 
 1/2-08 
 
 48 
 
 1/1-1 
 
 92 
 
 
 
 
 
 
 " 
 
 1/4-5 
 
 22*2 
 
 1/2 
 
 50 
 
 1/1 
 
 100 
 
 
 
 
 TABLE VI. 
 
 1/9 
 
 11-1 
 
 1/8 
 
 12-5 
 
 
 
 
 
 
 
 
 
 62 
 
 I 
 
 ,, 
 
 ,, 
 
 1/7-5 
 
 13-4 
 
 1/20 
 
 5 
 
 1/19 
 
 5-2 
 
 63 
 
 + 1 
 
 
 
 
 
 1/7-1 
 
 14-2 
 
 1/10 
 
 10 
 
 1/9 
 
 11-1 
 
 57 
 
 - 5 
 
 ,, 
 
 
 
 1/6-6 
 
 152 
 
 1/6-6 
 
 15 
 
 1/5-6 
 
 17-8 
 
 57 
 
 - 5 
 
 
 
 ,, 
 
 1/6-2 
 
 16-1 
 
 1/5 
 
 20 
 
 1/4 
 
 25 
 
 43 
 
 19 
 
 ,, 
 
 
 
 1/5-7 
 
 17-4 
 
 1/4 
 
 25 
 
 1/3-3 
 
 30 
 
 36 
 
 - 26 
 
 
 
 
 
 1/5-3 
 
 is-s 
 
 1/3-3 
 
 30 
 
 1/2-3 
 
 45 
 
 35 
 
 27 
 
 ,, 
 
 ,, 
 
 1/4-8 
 
 20-6 
 
 1/2-8 
 
 35 ' 1/1-8 
 
 54 
 
 35 
 
 -27 
 
 
 
 
 
 1/4-6 
 
 21-8 
 
 1/2-6 
 
 38 1/1-6 
 
 61 
 
 Failed 
 
 
 TABLE VII. 
 
 1/7 
 
 14-3 
 
 1/6 
 
 16-6 
 
 
 
 
 
 
 
 (I 
 
 62 
 
 
 ,, 
 
 ,, 
 
 1/5-8 
 
 177 
 
 1/20 
 
 5 
 
 1/19 
 
 52 
 
 59 
 
 -3 
 
 ,, 
 
 
 
 1/53 
 
 18-6 
 
 1/10 
 
 10 
 
 1/9 
 
 11-1 
 
 60 
 
 _ 2 
 
 ti 
 
 >' 
 
 1/4-9 
 
 20-2 
 
 1/6-6 
 
 15 
 
 1/5-6 
 
 17-8 
 
 Failed 
 
 
APPENDICES TO PART I. 
 
 203 
 
 . a 
 
 O jn 
 
 -111 
 
 fl M a . J 
 
 ^ 1 S 
 
 n 2 t, o 3 -; 
 
 C<! CO CO O t- A-l 
 
 
 : S 
 
 00 O . ^H O I- 
 
 GO 00 CO * 1^ f-~ Jb* 
 
 ooo b o 
 
 O 00 CO >* 
 
 flfi 
 
 b- CJ 
 
204 
 
 APPENDICES TO PART I. 
 
 ffi W 
 
 S*5 
 
 
 
 
 
 M ? 
 
 p ( Q q) 
 
 
 o 
 
 TJ 
 
 fl 
 
 
 
 ^ ^ 
 
 |||| 
 
 O ip CO CO 
 
 H 
 
 - ^ 
 
 S S s fe 
 
 O O 4% ea 
 
 
 CO H 
 
 c2"S< 
 
 * b b Af 
 N <M CM T-H 
 
 O 
 
 -5 
 
 * * o ^ 
 
 II II II II 
 
 
 
 
 
 
 "5 '3 
 
 
 
 S ^ 
 
 3- 
 
 
 
 
 *S 
 ^ 
 
 b Q *i ^ ^5 
 
 
 O 
 
 
 
 
 
 ^ooo^^o 
 
 
 ^ CO 
 
 Ig 
 
 fill 
 
 0000 
 CO S S? 00 
 
 
 
 1- 
 
 
 
 H ^ 
 
 CU a 3 
 
 
 H 
 
 e >> 
 
 
 
 ** M 
 
 i o g 
 
 
 a 
 
 ^1 
 
 
 
 
 
 
 
 53 *S 
 
 O O O h 
 
 
 *H 
 
 P.S 1 * 
 
 1*4 
 
 CO 
 
 fe 
 
 if 
 
 CM CN S o5 
 
 b b b b 
 
 
 u O 
 
 
 
 O ^ -^i CO 
 
 _ 
 
 "o 'o 
 
 
 
 02 3 
 
 ^ IS 
 
 O O O O 
 
 ^j CO 
 
 <^ -*-* 
 
 
 
 o 5 
 
 Iff 
 
 
 
 
 cf t*^ 
 
 
 
 o 1 ' 
 
 K|| 
 
 
 IEORET: 
 
 IMENSI 
 
 j Theoreti 
 efficienc 
 
 S S o - 
 
 4l ^p *0l ^i 
 
 o o o o 
 
 a 
 
 
 
 ||| 
 
 CO CO CO 1C 
 
 fH 1-5 
 
 S c 
 
 
 r- i H 
 
 
 
 J 
 
 O < 
 
 S H 
 
 01 
 
 
 EH 
 fe H 
 
 || 
 
 p fc s 
 
 rH CO r-1 (M 
 
 1 
 
 IW 
 
 
 ^ 
 
 III 
 
 OS C? UO* l^ 
 
 2 
 
 
 B> 
 
 fe 
 
 ^>> ^ "^ 
 
 1 Hi S 
 
 
 
 S 
 
 3g 
 
 o^S.2 
 
 
 g % 
 
 
 
 1 
 
 a 
 
 
 t I 
 
 
 
 
 
 * 03 
 
 
 Q fe 
 
 Kl p 
 
 % o 
 
 -s; a 
 
 g 3 
 
 111 
 
 |ii 
 
 ^ b b b 
 || J II ^ II 
 
 H cc 
 
 HH M 
 
 
 d, -^ <D 
 
 | | | | 
 a' " "* 
 
 --- 
 
 i 
 
 
 OS 
 
 bib b o b b "^ b 
 
 H g 
 
 
 4 3 3 
 
 ^ r-l 00 10 
 
 3 
 
 ETICAL I* 
 IONS, COM3 
 
 lf]j 
 
 CO O O 
 r-4 fH r- CSI 
 
 boob 
 
 ARISON OF 
 ENG 
 
 1 
 
 ^ >, i-l W 
 
 s 1 * * 
 
 III 
 
 I'll 
 
 1 
 
 ING THEOR 
 COMPRESS 
 PRESSIONS. 
 
 III 8 
 
 
 &i 
 
 1 
 
 s 
 
 Engine cyl: 
 
 1 S = 5 
 
 t~ S o? S 
 
 
 ill 
 
 IP 
 
 
 1 
 
 |1| 
 
 111 
 
 Ifl 
 
 111 
 
 ? * 1 P 
 
 b b b b 
 
 o 
 
 a 
 
 CO 
 
 
 1 
 
 
 w ^ S 
 
 fi*S 
 
 
 
 
 
 CT 1 
 
 
 S Q cc 
 
 0*3 O^ 
 
 
 _^ 
 
 
 5 
 
 
 ^J 
 
 *^3 o s 
 
 
 C" 
 
 
 ^|> 
 
 
 H 
 % 
 
 I 8 
 
 
 * 
 
 
 1 
 
 
PART II. 
 
 PETROLEUM ENGINES. 
 
 CHAPTER XX. 
 
 THE discovery of petroleum in large quantities in Russia 
 and America has materially stimulated inventors to devise 
 means of utilising this enormous latent energy. Before 
 dealing with the oil engines which have attained success, 
 it will be well to discuss the physical properties of the 
 substance known generally as petroleum. The chemistry 
 and the commercial production of petroleum are subjects 
 beyond the scope of this work. Those physical properties, 
 however, which more directly affect the design of oil engines 
 should be carefully studied. 
 
 Petroleum has been found at various times in nearly all 
 parts of the world, and has received a variety of names. 
 The most productive American area lies within the 
 boundaries of New York and Pennsylvania. In Canada 
 there has been a large production, and the famous Baku 
 wells of Russia are well known. Petroleum consists 
 chemically of hydrogen and carbon. It is probably of 
 vegetable origin, though this point is vigorously contested 
 by geologists supporting other theories. Crude petroleum 
 is usually of a dark green hue by reflected light, showing 
 red by transmitted light. 
 
 An idea of the commercial value of petroleum and the 
 variety of its uses will be gathered from the following 
 table : 
 
206 
 
 PETROLEUM. 
 
 ONE HUNDRED GALLONS OF CKPDE PKTRO'KUM WILL YIELD UPON FRACTIONAL 
 DISTILLATION : 
 
 
 Gallons. 
 
 Specific 
 gravity. 
 
 Flashing point. 
 
 Benzene light oil \ 
 
 1 
 
 0-725 
 
 - 10 deg. Ceil. 
 
 Gasolene V oil engines 
 
 3 
 
 0775 
 
 + ,, ,, 
 
 Kerosene / 
 
 27 
 
 0-822 
 
 25 ,, 
 
 Saliarovi \ 
 
 12 
 
 0-870 
 
 100 
 
 Veregenni I 
 
 10 
 
 0-890 
 
 150 ,, 
 
 > lubricating 
 
 
 
 
 Lubricating oil | 
 
 17 
 
 0-905 
 
 1"5 
 
 Cylinder oil / 
 
 5 
 
 0-915 
 
 200 
 
 Vaseline 
 
 1 
 
 0-925 
 
 
 Residuum liquid fuel 
 
 14 
 
 
 
 
 10 
 
 
 
 
 
 
 
 The oils used in oil engines are benzene, gasolene, and 
 kerosene, the two former of which are generally known 
 as light oils. They are extremely volatile at ordinary 
 atmospheric temperatures, and are consequently dangerous 
 when carelessly handled. In order that these light oils 
 may be used by owners of oil engines without the incon- 
 venient restrictions imposed by the Petroleum Acts of 
 1871-9, regulations were issued on November 3rd, 1896, by the 
 Secretary of State. For the purposes of these regulations 
 a light oil is defined as having a flashing point* of less than 
 73 deg. Fah. When the flashing point exceeds 73 deg. Fah. 
 the oil is said to be heavy, and to this class the above 
 regulations do not apply. Moreover, engines driven by the 
 heavy oils are regarded as safe, and for this reason they 
 will no doubt in time supersede the light oil motors which 
 are at the present time used for driving vehicles. Such oils 
 may be stored in air-tight tanks, all openings from which 
 
 * The temperature at which oil commences to give off inflammable vapour, 
 when under ordinary atmospheric pressure, is termed the flashing point. 
 
FLASHING POINT. 
 
 207 
 
 must be covered with gauze of 400 mesh. No tank may 
 exceed 20 gallons capacity; they must be kept in well- 
 ventilated places. When used in connection with motors 
 for driving vehicles the quantity carried by each vehicle 
 must not exceed 40 gallons. 
 
 A^*ough test of the flashing point may be made in the 
 following way : Take a small quantity of the oil to be tested 
 
 FIG. 96. Abel tester. 
 
 in a metal vessel. Heat the vessel gradually by means of a 
 lighted taper. Apply the light occasionally to the surface 
 of the oil. When the oil ignites immediately put out the 
 flame and take the temperature of the oil with a Fahrenheit 
 thermometer. This will give the flashing point approxi- 
 mately. 
 
208 ABEL TESTER. 
 
 In England and Canada an instrument known as the Abel 
 Tester is legally recognised as giving the flashing point. 
 This instrument is shown in fig. 96. The copper vessel W 
 contains water, which is heated by the lamp B to 130 deg. 
 Fah. The oil to be tested is placed in the chamber P, which 
 is surrounded by an air jacket A. A gas flame/, outside the 
 chamber P, is arranged to swing round until the flame shoots 
 through a small hole in the cover of the oil chamber, when 
 the hole is uncovered by the moving of the slide S. To test 
 a sample of oil, it should first be cooled to about 60 deg. Fah. ; 
 a quantity is then placed in the chamber P, until the pointer 
 p is just covered. Light the gas at the nozzle /. For each 
 degree rise of the thermometer t, the slide s is opened, and 
 the flame gently tilted into the oil chamber. A pendulum is 
 usually provided to time this operation to four oscillations, 
 the slide being closed on the fourth swing. When a blue 
 flash is obtained on the application of the test flame, the 
 flashing point of the oil is read from the thermometer t. 
 
 The specific gravity of shale oils, as sold in Britain, varies 
 0'78 to 85, and the flashing point varies with the density 
 from 76 deg. Fah. to 225 deg. Fah. The following table has 
 been obtained by Professor Eobinson, who has carried out 
 an interesting series of experiments, the results of which 
 may be more fully referred to in The Engineer, September 
 llth, 1891. 
 
 The chief difficulty in the construction of oil engines is 
 the design of the vaporiser, and in this connection the 
 figures given in the latter columns of the above table are 
 particularly interesting. We see that American Royal 
 Daylight oil boils at a temperature of 144 deg. Cen., but at 
 215 deg. Cen. only one quarter of the volume of oil treated 
 is vaporised. At a temperature of 230 deg. Cen. 65 per 
 cent by volume still remains. Thus it appears the tempera- 
 ture must be raised far beyond this point in order that the 
 whole volume may be distilled. But it is just here that 
 the designer meets with fresh difficulties, owing to the fact 
 that if the temperature be raised too much decomposition 
 
OJ 
 
 a 
 
 2 CO 
 
 CO TH CO 
 
 CO . -0 (N CO COCO(N(N 
 
 H 
 
 w 
 
 
 
 
 ,43* 
 
 
 
 
 5 i; a 
 
 
 
 atunioA 
 
 (S| 
 
 CO CO 
 
 
 ,-. 
 
 G 
 
 
 
 1 la 
 
 co 
 
 8 S 3 
 
 S .cooo S S S o? 
 
 1 a* 
 
 bo N 
 
 
 
 
 E$^ 
 
 c 
 
 
 
 
 Illll 
 
 3 
 
 .2^ 43 
 
 ^ 
 
 O 1C 
 
 
 Oi O <M 
 C<I O <N 
 
 . 1 
 
 w 1 
 
 >r * w 
 
 (S 
 
 
 fa 
 
 to 
 
 
 
 
 I'll 
 
 bO rX 
 
 '-5 
 
 S : S S S 
 
 tt & 
 
 
 
 
 
 s "P 
 
 , M 
 
 00 t- 
 
 
 3 * 
 
 
 
 .9-g 
 
 o ^ 
 
 
 (^ 3 rH 
 
 it 
 
 . 
 
 
 
 s ft 
 
 i2 2 
 
 S o CO 
 
 S : 3 : : : | 
 
 Itl 
 
 ^ 
 
 . 
 
 CO -f 
 
 to'"' 
 
 
 
 
 000 
 
 TIT?^ 'Sop 09 
 ^B ^^lABoS agpedg 
 
 o 
 
 ? ? 
 
 000 
 
 S S 2 S? S S S 
 
 CO CO 00 CO CO 00 00 
 
 o o o o o o o 
 
 uon^S lad 
 
 -aO>3 UI p3J3Al[9p 
 
 1 
 
 1 * 
 
 sr * 
 
 .... .... 
 
 
 
 
 . .. 
 
 S3 
 & ^ ^ ^ S3 fl 
 
 C 
 
 c3 
 
 i co 
 I -g 1 
 
 III ^ 1 1 1 
 
 
 
 
 to g OT 
 
 
 3 
 
 | 
 
 tc |o *& 
 
 S S 02 .S M J3 
 
 
 h3 
 
 
 232 -goo 
 
 
 
 
 * . . 
 
 
 
 
 : : : : : : 
 
 
 
 : : 
 
 1 ; - : '! 
 
 d 
 
 02 ^3 
 
 ^ .SP 
 
 <X> i 
 
 1 ! ^" 3 
 
 
 
 
 
 "3 
 
 bL ^* 
 
 "2 
 
 03 .2 g* S <u S 
 
 CJ 
 
 S ft 
 
 t? & * 
 
 X s "S -5 a 
 
 1 
 
 li 
 
 1 1 2 
 
 ^3 o ^Saj^'g^ 
 
 g S S .3 -S a -e 
 
 ^ag^algo 
 
 
 American 
 
 1 1 1 
 g - g 
 
 s s s 
 
 ||| | 1 1 1 
 III 1 1 1 I 
 
 K M eq q oo 02 i-5 1 
 
 10G 
 
210 LENOIR ENGINE. 
 
 takes place, and a carbon residue remains, which, if present 
 in large quantities, is quite fatal to the successful action of 
 a vaporiser. The best method of dealing with this class 
 of oils is to raise the temperature to a safe limit, then allow 
 hot air to pass over the oil. In the description of the 
 various methods of vaporising we shall see how this is 
 practically carried into effect. Suffice it to say that in the 
 
 FIG. 97. Longitudinal section of Spiel engine. 
 
 design of any oil engine the brand of oil to be used must 
 be the first consideration of the designer. 
 
 The earliest oil engines worked with the light oils, such 
 as gasolene, which vaporise at ordinary atmospheric tempera- 
 tures. The method of working was to cause air to pass 
 through scattered particles of the oil, and thus become 
 saturated with oil vapour. It is said that the Lenoir oil 
 engine developed 1 horse power hour by the consumption 
 of 0'9lb. of oil of 65 density. 
 
SPIEL'S ENGINE. 
 
 211 
 
 In 1873 Brayton constructed an engine driven by a safe 
 oil of 0'85 density, the consumption of which is stated to be 
 2 Ib. per horse power hour. In this engine air was mixed 
 with petroleum vapour by forcing the air at considerable 
 pressure through the liquid. The mixture then passed to 
 the cylinder through a network of wire gauze, the latter 
 being kept hot by the exhaust products. 
 
 The Spiel engine, a longitudinal section of which is shown 
 at fig. 97, is worked with light oil of specific gravity about 
 
 FIG. 98. Supply valves of Spiel engine. 
 
 07. The liquid is contained in the upper chamber, from which 
 it runs down the pipe F to the supply valves shown in detail 
 in fig. 98. These valves act in the following way : A spiral 
 spring B, forces a double seated valve P towards the end V. 
 Spirit, therefore, passes through the open valve and remains 
 beneath the plunger E. A cam shown at M, fig. 99, gives a 
 
212 
 
 SPIEL'S ENGINE. 
 
 vertical motion to the crosshead N, fig. 98. Upon the down 
 stroke of N, the roller S forces the upper end of the lever U 
 outwards, thereby forcing the spindle V inwards, closing 
 
 FIG. 99. Sectional plan of Spiel engine. 
 
 the spirit valve P, and compressing the spiral spring R. The 
 down stroke is prolonged until the spirit is discharged by tl^e 
 plunger E against a cone-shaped disc in the chamber X. The 
 
 FIG. 100. End view of Spiel engine. 
 
 function of this cone is to spray the spirit and to cause it to 
 mix with air. During the downward movement of the 
 plunger E the admission valve to the cylinder is opened, and 
 
SIMPLEX CARBURATOR. 
 
 213 
 
 the piston, on its outward stroke draws the spirit vapour 
 and air down the pipe H. Ignition is caused by a moving 
 slide, shown in figs. 97, 99, 100. The action of the slide is 
 very similar to that of the early Otto slide ignition engines. 
 Simplex Carburettor. Messrs. Delamare-Devoutteville and 
 Malandin have adopted an arrangement for supplying gas 
 
 FIG. 101 Simplex Carburator. 
 
 to their Simplex gas motors where a supply is otherwise 
 unavailable. Fig. 101 shows a sectional elevation of this 
 carburator. The spirit (density about 07) is contained in 
 the tank Z. The pipe E conducts hot water from the engine 
 
214 SIMPLEX CARBURATOR. 
 
 jacket to a cylindrical casing T, whence it passes away 
 through the pipe H. The central chamber contains a spiral 
 wire brush, down which the spirit drops. Through the cock 
 R a quantity of hot water is allowed to pass down the 
 spiral ; this assists in heating the spirit to completely 
 vaporise it. 
 
 The water falls into the chamber C, whence it escapes 
 through the pipe N. A perforated cork floats on the water, 
 thus preventing the suction of water spray into the cylinder 
 through the suction pipe A. The valve B is attached to the 
 suction pipe A, and is used to prevent the return flow of 
 high-pressure gases from the cylinder to the carburator. 
 The supply of spirit is controlled by the action of the 
 governor on the valve S. 
 
 The use of petroleum for firing boilers containing water 
 has been successfully carried out, and economy has also been 
 sought by evaporating spirit instead of water in the boiler. 
 This arrangement is attended with considerable danger 
 unless very carefully handled ; at the same time it is true 
 that certain thermal advantages result from the evaporation 
 of spirit instead of water, on account of the smaller quantity 
 of heat supplied to the spirit being rendered latent. These 
 arrangements cannot be regarded as internal-combustion 
 engines, and will not therefore be described. 
 
PRIESTMAN OIL ENGINE. 215 
 
 CHAPTER XXL 
 
 IT is not our intention to traverse the early history of the 
 oil engine. We shall at once, therefore, enter upon a 
 description of the engines which have been made within the 
 last few years. It will be understood that the main difference 
 between the gas and oil engine is the addition to the former 
 of some means of causing the oil to form an explosive 
 mixture with air. This being so, our attention need only 
 be directed to this matter ; indeed, many makers are now 
 building gas engines which may with very little trouble be 
 converted into oil engines. This necessitates the addition 
 of a vaporising chamber, which need be no encumbrance 
 when the engine is worked by a supply of gas. 
 
 It has been the object of many inventors to construct an 
 engine capable of working with heavy petroleum oils, 
 instead of petroleum spirit. The inflammability of spirit, 
 even when a light is applied at a distance of 2 in. or 3 in. 
 from the spirit, renders its use highly dangerous ; whereas 
 petroleum oil will extinguish a lighted taper when, at 
 ordinary atmospheric temperatures, the taper is plunged 
 into the oil. 
 
 Although many attempts were made to utilise the heavy 
 oils, none led to practical results until Messrs. Priestman, 
 of Hull, having acquired Eteve's patents, persevered with 
 the constructive details, and were thus able to place a satis- 
 factory engine upon the market in the year 1888. 
 
 In the Priestman engine the idea of spraying oil mingled 
 with air into a heated chamber, called the vaporiser, was 
 first brought forward. 
 
 Although it is now well known that oil may be vaporised 
 without being finely divided into a spray and mixed with 
 air, it is a noteworthy fact that the development of this 
 class of motor proceeded on the lines indicated by Messrs. 
 Priestman, and very great credit is due to this firm for their 
 
216 
 
 PRIESTMAN OIL ENGINE 
 
PRTESTMAN'S GOVERNOR 
 
 217 
 
 persevering efforts, which have resulted in the production 
 of a motor of great economy and usefulness. The following 
 description, together with the illustrations, will suffice to 
 explain the working of the Priestman engine. 
 
 The oil is contained in a cast-iron tank Y, fig. 102, formed 
 in the bed of the engine. The oil, with a supply of air 
 under a pressure of about 8 Ib. per square inch, is carried 
 to the spray maker S ; here they mix and break up into a 
 fine vapour. This is then passed through a heating chamber, 
 called the vaporiser (which latter is warmed by a lamp I 
 
 FIG. 103. Section and End View of Spray Maker of Priestman Engine. 
 
 before starting, and is kept hot, when running, with the 
 exhaust products). From this chamber the vapour is drawn 
 into the engine cylinder, compressed by the piston, and 
 ignited by an electric spark, or lamp and hot tube, as pre- 
 ferred. With the Edison L'alande system of creating the 
 spark, an engine can be run for 500 to 1,000 hours without any 
 attention being given to the two small cells. The governing 
 is effected by regulating the quantity of oil, instead of by 
 entirely stopping the supply. The piston is lubricated by 
 the deposit of oil on the surface of the cylinder. 
 
 Fig. 103 shows a section of the spraying nozzle, and the 
 wing valve for controlling the air supply. Oil is forced into 
 
218 PRIESTMAN OIL ENGINE. 
 
 the nozzle E by means of compressed air in the oil tank. 
 On its way to the nozzle the oil passes through a tapered 
 hole in the plug H, to the lower end of which the wing valve 
 G is attached. The plug H is controlled by the governor, 
 and it will be obvious that both the oil and air are 
 simultaneously throttled, thus keeping the mixture uniform 
 The nozzle E, a larger view of which is shown in fig. 104, is of 
 special construction, and is the result of a long series of ex- 
 periments. Oil enters the nozzle through the central opening, 
 and the air is forced in through the annular space. It is an 
 important feature in the construction of the nozzle that the 
 air current is directed against the flow of oil. In the earlier 
 
 FIG. 104. Spraying Xozz'e, Priestman Oil Engine. 
 
 nozzles used by Messrs. Priestman the air and oil were made 
 to flow in the same direction, where they mingled together. 
 The form was, however, gradually altered, until it was found 
 to give the best spray when made as in fig. 104. 
 
 For the ignition of the charge a simple bichromate cell 
 may be used, with one induction coil. 
 
 The charge recommended for the battery is 
 
 8 parts by weight of bichromate of potash. 
 6 parts by weight of sulphuric acid. 
 80 parts by weight of water. 
 
 This charge will, if made up in ounces, last for about 30 
 hours, at a cost of about 6d. 
 
PRIESTMAN OIL ENGINE. 
 
 219 
 
 The following figures have been obtained by Professor 
 Unwin in a systematic test of a Priestman oil engine : 
 
 
 
 
 
 PM' 
 
 a 
 
 fl 
 
 
 E 
 
 
 
 
 >> 
 
 w . 
 
 "o 
 
 "o ?>" 
 
 fl-8 
 
 -B*s * 
 
 
 
 
 "s 1 
 
 !* 
 
 " ID 
 
 5 * 
 
 -53 a 
 
 2 "S^ 
 
 Loads. 
 
 I.H.P. 
 
 B.H.P. 
 
 II 
 
 
 Is 3 
 
 jg,g 
 
 11 
 
 |ij 
 
 
 
 
 ai 
 
 a'' 5 (2 
 o 
 
 1 
 
 ca 
 
 1^1 
 
 h 
 
 l; 
 
 Full load.. 
 
 7-40 
 
 6-76 
 
 0-91 
 
 0*946 | Russolene 
 
 41-38 
 
 207-73 
 
 264-5 
 
 Half load.. 
 
 4-70 
 
 3-62 
 
 0-76 
 
 1-381 Russolene 
 
 25-48 
 
 214-29 
 
 254-0 
 
 
 
 
 
 I.H.P. 
 
 
 
 
 
 No load. . . 
 
 0-889 
 
 .. 
 
 .. 
 
 5-734 
 
 Russolene 
 
 5-51 
 
 187-3 
 
 260-5 
 
 
 
 
 
 
 Royal 
 
 
 
 
 Full load.. 
 
 9-36 
 
 772 
 
 0-S2 
 
 0842 
 
 Daylight 
 
 53-2 
 
 204-33 
 
 268-0 
 
 The dimensions of the engine from which these figures 
 were obtained are as follow : 
 
 Diameter of cylinder 8'51in. 
 
 Stroke 120in. 
 
 Clearance in percentage of working 
 
 volume 53 per cent. 
 
 Volume of air-compressing pump per 
 
 stroke 0'0329 cu. ft. 
 
 Diameter 512 in. 
 
 Diameter of flywheel 4'61ft. 
 
 Weight of engine without flywheel 26 cwt. 
 
 Weight of flywheel 10 cwt. 
 
 The oils used in the above trials yielded the following 
 constituents : 
 
 Russolene. Royal Daylight 
 
 Per cent. Per cent. 
 
 Carbon 85'88 84'62 
 
 Hydrogen 14'07 1486 
 
 Oxygen, &c., by difference 005 
 
 10000 
 
 0-52 
 
 100-00 
 
220 
 
 PRIESTMAN OIL ENGINE. 
 
 Density at 60 deg. Fah 822 0793 
 
 Flashing point (Abel test)... 86 deg. Fah. 77 deg. Fah. 
 The calculated total calorific 
 
 value 21,180 21,490 
 
 The calculated calorific value 
 
 when steam formed is not 
 
 condensed 19,957 20,198 
 
 The following is a resume of the results of a trial 
 conducted by Professor Unwin and Mr. D. Pidgeon, at the 
 Royal Agricultural Show, Plymouth, in 1890 : 
 
 PRIESTMAN OIL-ENGINE TRIALS. 
 
 
 
 cu 
 
 > a 
 
 
 
 
 
 g 
 
 a v? 
 
 
 '"i g '^ 
 
 ' ^ 
 
 Loads. 
 
 I.H.P. 
 
 B.H.P. 
 
 
 sjjl 
 
 Brand of oil 
 used. 
 
 "l-s 
 
 I 1 
 
 
 
 
 SH 
 
 b 
 
 
 ri 
 
 
 Full load .... 
 
 5-24 
 
 4-49 
 
 0-85 
 
 1-066 
 
 Broxburn oil 
 
 33-96 
 
 179-5 
 
 Half load .... 
 
 3-21 
 
 2-36 
 
 0-73 
 
 1-460 
 
 Broxburn oil 
 
 20-65 
 
 180-8 
 
 ANALYSIS OF BROXBURN OIL. 
 
 Carbon 86 '01 percent. 
 
 Hydrogen 13'90 
 
 Undetermined 0'09 
 
 Calculated calorific value of Broxburn oil, when steam 
 formed is not condensed, 19,700 British thermal units. 
 
 The Priestman oil engine has been successfully used for 
 pumping, electric lighting, fog signalling, and for driving 
 launches and barges. For the latter purpose Messrs. 
 Priestman supply a patent reversible screw propeller, the 
 blades of which can be inclined to drive ahead or astern 
 without stopping the motor. 
 
 The Samuelson Oil Engine (Griffin's Patent}. This engine 
 made by Messrs. Samuelson and Co. Limited, Banbury, and 
 
THE SAMUELSON OIL ENGINE. 
 
 221 
 
 the Priestman engine, are the only two manufactured which 
 work with the spray maker. The oil sprayer is shown in fig. 
 105. Its action is somewhat different from the Priestman 
 nozzle, inasmuch as the oil is sucked up into the annular 
 space surrounding the air nozzle by the action of the 
 horizontal air jet ; whereas in the Priestman engine the 
 
 FIG. 105. Griffin Patent Oil Sprayer. 
 
 oil occupies the central nozzle, and the air meets it at the 
 orifice. The air pressure is maintained by an eccentric- 
 driven pump, at about 12 Ib. above atmosphere. The 
 vaporiser is a long corrugated chamber, heated by the 
 
222 
 
 THE SAMUELSON OIL ENGINE. 
 
 exhaust products. The governor is of the ordinary centri- 
 fugal pattern, and acts by cutting off the air supply, at the 
 same time preventing the opening of the exhaust and 
 admission valves. This method of governing has the 
 advantage that the cylinder is not cooled by the admission 
 of cold air when the engine is missing fire. 
 
 FIG. 106. Samuelson's Igniting Arrangement. 
 
 The igniting arrangements are shown in fig. 106. An air 
 jet plays upon a piece of bent wire, which latter dips into a 
 chamber filled with oil to a constant level. The air jet 
 draws a film of oil up the wire, carries the oil forward in 
 the form of spray, and the mixture burns around the 
 ignition tube. In starting this engine a hand pump is used 
 to compress air for the ignition spray and for the vaporiser 
 sprayer. Both are lighted, and the engine allowed to stand 
 for about ten minutes, until the vaporiser is hot enough to 
 
THE HORNSBY-AKROYD OIL ENGINE. 
 
 223 
 
 work. The length of the ignition tube has apparently no 
 little influence upon the form of card obtained, due no doubt 
 to the fact that the charge does not get compressed into 
 the hottest part of the tube when the latter is too short. 
 
 The Horn&by-Akroyd Oil Engine. This engine is made by 
 Messrs. Hornsby and Sons Limited, Grantham, and works 
 upon an entirely different principle to those previously 
 described. No sprayer is employed, neither is there an 
 ignition tube to the engine. Fig. 107 shows a transverse 
 section through the vaporiser. The action of the engine is 
 extremely simple. A small quantity of oil is injected into 
 the vaporiser, with a little air. The oil quickly spreads 
 
 Fio. 107. Section through Valves and Cylinder of Hornsby-Akroyd Oil Engine. 
 
 itself over a large surface of the hot walls of the vaporiser, 
 and is consequently quickly evaporated. The vaporising 
 chamber contains only the products of combustion of the 
 previous charge at the moment the oil is injected. As the 
 engine piston makes its outward stroke, fresh air is drawn 
 into the cylinder through valves and passages which have 
 no connection with the vaporising chamber ; hence at the 
 end of the outstroke of the piston the cylinder is filled with 
 pure air, whilst the vaporiser is charged with oil vapour, 
 mixed with the products of combustion of a previous charge. 
 The charge in the vaporiser is not combustible until more 
 fresh air is forced into it by the return stroke of the piston. 
 During the instroke the piston drives the air from the 
 
224 
 
 THE HORNSBY-AKROYD OIL ENGINE. 
 
 cylinder into the vaporiser, thus rendering the charge in the 
 latter highly explosive. When the compression is completed 
 the temperature of the vaporising chamber is sufficiently 
 high to cause ignition of the charge. It will be understood 
 that correct proportion between the cylinder volume and 
 that of the vaporising chamber is essential for the successful 
 working of the engine. To determine this Messrs. Hornsby 
 carried out numerous experiments. The area of the neck of 
 the vaporising chamber is another important factor. This 
 must be small enough to prevent the diffusion of the gases 
 during the outstroke of the piston. Another rather 
 important function of the small neck of the vaporising 
 chamber is to control the flow of heat from the vaporiser 
 
 FIG. 108. Indicator Diagram from Hornsby- Akroyd Oil Engine. 
 
 to the cylinder water jacket. By making the vaporiser 
 without jacket, and suitably proportioning the area of 
 metal at the neck, a proper temperature is maintained to 
 ignite the charge every cycle. A lamp is provided to heat 
 the vaporiser before starting the engine; afterwards the 
 heat of combustion is sufficient to maintain it at a proper 
 temperature. 
 
 The Hornsby engine was awarded the first prize in com- 
 petition with nine other makers at the Royal Agricultural 
 Show at Cambridge. According to the report of the trials, 
 written by Professor Capper, this engine ran without hitch 
 of any kind from start to finish. One attendant only was 
 employed all through the trials, and he started the engine 
 easily and with certainty, after working the hand blast to 
 
CROSSLEY OIL ENGINE. 225 
 
 the lamp for eight minutes. The longest time taken to start 
 was nine minutes, and the shortest seven. The revolutions 
 were very constant, and the power developed did not vary 
 one-quarter of a brake horse power from day to day. The 
 oil consumption, reckoned on the average of the three days' 
 run, was 0*919 Ib. per brake horse power hour, including 
 the oil used for the starting lamp. The oil used was 
 Russolene, sold at Cambridge at that time for 3|d. per gallon. 
 The cost of oil per brake horse power hour is therefore about 
 |d. At half -load the oil consumption was found to be T49 Ib. 
 per brake horse power hour. 
 
 The Hornsby engine is governed by the interception of 
 the oil supply through a by-pass valve, thus decreasing the 
 explosion pressures as the speed rises. It is important that 
 the firing should take place regularly every cycle, as other- 
 wise the temperature of the vaporiser is liable to become 
 too low. 
 
 Fig. 108 shows a mean indicator diagram, taken from a 
 Hornsby- Akroyd engine on full-power trial. The maximum 
 explosion pressure was about 130 Ib. per square inch, the 
 mean pressure about 29 Ib. per square inch. 
 
 Crossley Engine. Fig. 109 shows a sectional elevation 
 and plan of the Crossley vaporiser. The valve to the left 
 of the plan view is held upon its seat by a spiral spring, 
 but opens communication to the cylinder during the out- 
 stroke of the piston. The governor acts upon this valve by 
 a hit-and-miss arrangement, so that as the speed of the 
 engine rises the valve is not opened. During the outstroke 
 of the piston air follows through a separate automatic valve, 
 held upon its seat by a spring ; consequently there is a 
 considerable suction through the passages connected to the 
 valve shown when the latter is opened. Through these 
 passages hot air and oil are simultaneously drawn, the oil 
 entering at the pipe shown in the plan. The air enters the 
 annular space round the chimney shown in elevation, and 
 passes downwards to meet the oil. 
 
 A lamp is continually burning beneath the vaporiser, and, 
 16G 
 
226 
 
 CROSSLEY OIL ENGINE. 
 
 besides heating the latter, its flame plays upon the ignition 
 tube, which is connected to the combustion chamber of the 
 engine. The products of combustion of the heating lamp 
 pass up through the square holes shown in the vaporiser. 
 In this way the heating surface is largely increased. The 
 
 Fio. 109. Sectional Elevation and Plan of Crossley's Vaporiser 
 
 average time taken to start the Crossley engine, at the 
 Royal Agricultural Show trials at Cambridge, was, accord- 
 ing to Professor Capper's report, 16 minutes. The engine 
 ran without giving trouble, the governing was good, and 
 the power developed was uniform throughout the trials. 
 
 The Wells Engine. The oil feed to this engine is delivered 
 by means of a rotating plug, driven by the link R, fig. 110 
 
WELLS' OIL ENGINE. 
 
 227 
 
 This plug allows a measured quantity of oil to drop on the 
 inclined surface shown at V. Here it becomes vaporised by 
 the high temperature maintained by an oil lamp and auto- 
 matic air blast. The lever L is operated by a cam on the 
 side shaft, and rocks upon a pin fixed in the boss E. The 
 lower end of the lever opens the exhaust valve when pressed 
 inwards against the tension of the spiral spring placed at 
 its upper end. On the return stroke of the lever the oil 
 plug is rotated, and oil emitted to the vaporiser at V ; at the 
 same time the air valve A is opened by the adjustable stud 
 
 FIG. 110. Arrangement of Valves and Levers in Wells' Oil Engine. 
 
 on the lever L. By this ingenious arrangement all the 
 valves are controlled by the one lever. 
 
 The governing is effected by the horizontal link C. This 
 link is balanced so that its position of equilibrium is such 
 that the upper end of the lever L clears the point K when 
 the former is moving from right to left. When the upper 
 end of the lever L is moved outwards by the cam, the 
 horizontal lever C is drawn downwards ; when released it 
 recovers its position of equilibrium in a definite time, 
 depending upon the spring adjustment at M. If the speed 
 of the engine be increased so that L moves inwards before K 
 has risen, then the further movement of the lever is arrested 
 by the catch K. In this way the exhaust valve is kept open, 
 
228 
 
 THE TRUSTY OIL ENGINE. 
 
 allowing the hot products of combustion to return into the 
 cylinder; at the same time the oil and air valves are 
 prevented from opening. The ignition of the charge is 
 effected by means of a tube, heated by an auxiliary lamp. 
 
 Trusty Oil Engine. Sectional views of this engine, which 
 is made by Messrs. Weyman and Hitchcock, are shown at 
 figs. Ill and 113, and an end view at fig. 112. The oil 
 supply is pumped, together with a little air, into the tube C, 
 fig. Ill, and in falling to the bottom of the annular space 
 
 FIG. 111. Section through Cylinder and Vaporiser of the Trusty Oil Engine. 
 
 becomes vaporised. The mixture then rises through the 
 vapour valve, fig. 113, into the inner chamber, and passes 
 away from there into the combustion chamber. Here it is 
 mixed with more air, which enters the cylinder through 
 valve F, fig. Ill, on the outstroke of the piston. Ignition 
 may be effected by contact with the hot chamber, though 
 usually the ordinary hot tube igniter is used. 
 
 The following is an extract of a trial made by Mr. W. 
 Worby Beaumont, M.Inst.C.E., on the Trusty engine, in 
 1893. The dimensions of the engine were as follow : 
 Diameter of cylinder, 7*375 in. ; length of stroke, 14 in. ; 
 
THE TRUSTY OIL ENGINE. 
 
 229 
 
 FIG. 112. Arrangement of Valves, Lever?, &c.. in Weyman and Hitchcock's 
 Trusty Oil Engine. 
 
 FIG. 113. -End Section of FIG. 113A. Sectional Plan of 
 Trusty Vaporiser. Trusty Vaporiser. 
 
230 
 
 THE TRUST* OIL ENGINE. 
 
 revolutions per minute, 248 ; flywheel diameter, 5 ft. Accord- 
 ing to Mr. Beaumont's report, the engine works equally well 
 with all kinds of lamp oil. 
 
 " The duration of the several trials was : Full power, four 
 hours ; half power, two hours ; running light, one hour. 
 The intended full-power load was 6 horse power, and the 
 brake load was adjusted as nearly as possible to this. The 
 running of the engine was very regular, and when at half 
 power the speed was within 0'4 of 1 per cent of that of 
 full power. The oil used in the trials was Royal Daylight, 
 costing in quantities about 4d. per gallon. Its speciric 
 gravity was found to be 0'802, one pint thus weighing 
 1-0025 Ib." 
 
 The general results of the trials are as follow : 
 
 
 Full 
 power. 
 
 Half 
 
 power. 
 
 Running 
 light. 
 
 Brake horse power 
 
 5-98 
 
 3-32 
 
 
 Indicated horse power 
 
 7-04 
 
 5-48 
 
 272 
 
 
 847 
 
 61 / 
 
 
 Oil used per B H P hour . . Ibs 
 
 0'82 
 
 1-12 
 
 
 Oil used per I H P hour Ibs. 
 
 0-69 
 
 0-68 
 
 0'82 
 
 
 
 
 
 COST OF FUEL FOR TRUSTY OIL ENGINE WITH OIL AT 
 
 FOURPENCE PER GALLON. 
 
 
 CostperI.H.P. 
 hour. 
 
 Cost per B.H.P. 
 hour. 
 
 
 0'348d 
 
 0'410d 
 
 Engine with half load 
 
 0'342d 
 
 0'565d. 
 
 
 0-406J. 
 
 
 
 
 
 The above figures do not include the cost of oil for heating 
 the lamp. This is, however, small, and was found to be 
 from 6 oz. to 7 oz. per hour. Thus, on the average, the total 
 cost may be taken as not exceeding 0'4d. per I.H.P. hour. 
 
CAMPBELL'S OIL ENGINE. 
 
 231 
 
 CHAPTER XXII. 
 
 Campbell's Oil Engine. The vaporiser of this engine is 
 roughly shown in fig. 114. It is of simple construction, and 
 acts in the following way : The automatic valve is held up 
 to its seating by the spiral spring. Beneath the valve is a 
 port communicating with the cylinder, and through this 
 port the charge of vaporised oil and air is drawn. Oil is 
 led through the small pipe into the annular passage sur- 
 rounding the valve. When the piston moves outwards it 
 
 FIG. 114. Campbell's Vapiriser 
 
 acts as a pump and draws air past the valve. At the same 
 time the annular passage is thus opened for the admission 
 of oil to the vaporiser. The oil drops upon the walls of the 
 vaporiser, which is heated by a lamp beneath it. This lamp 
 also heats the ignition tube. The governing is effected by 
 intercepting the oil supply and by holding open the exhaust 
 valve, thus allowing the exhaust products to return to 
 the cylinder instead of drawing air through the automatic 
 valve. The time taken to start the engine during the Royal 
 Agricultural Society's trials was somewhat lengthened by 
 
232 
 
 THE BRITANNIA OIL ENGINE. 
 
 attempting to start before the vaporiser was hot enough. 
 With a larger lamp the time might have been reduced to 
 about 10 minutes to heat the vaporiser. 
 
 The Britannia Company's Engine is the invention of 
 Mr. Roots. Fig. 115 shows a section through the vaporiser 
 of the Britannia oil engine made under Roots' patents. 
 I is the ignition tube, which is surrounded by the casing H. 
 
 Fia. 115. Section through the Vaporiser of the Britannia Oil Engine. 
 The arrows show the direction of flow of the air. 
 
 Air enters the casing, and in passing round the spiral 
 passages it becomes heated by the same lamp used to heat 
 the ignition tube. During the outstroke of the piston the 
 air is sucked from H, through the elbow at X, to Y, and so 
 passes into the cylinder through the valve A V. As the air 
 sweeps through the passages marked X, it carries away with 
 
THE BRITANNIA OIL ENGINE. 
 
 233 
 
 it a certain quantity of oil from the grooves of a spindle 
 which enters the chamber X at the point marked F. The 
 oil is vaporised by contact with the hot walls of the 
 chamber Y, and passes off with the air to the cylinder. On 
 the return stroke of the piston the vapour is compressed 
 into the ignition tube and ignited. 
 
 The way in which the oil is carried into the chamber, 
 shown at X, fig. 115, is further illustrated in fig. 116. The 
 spindle is here shown at A A, and the side view of the 
 
 FIG. 116. Arrangement of Oil Supply and Governor Gear of the Britannia 
 Oil Engine. 
 
 chamber X shows the spindle A A passing through it. This 
 spindle has grooves cut upon it, which are shown in the oil 
 cavity C, fig. 116. The quantity of oil entering the chamber 
 X is regulated by the governor. As the speed rises the 
 piece B is lifted, so that all the grooves on the spindle do 
 not enter the chamber X ; thus the oil supply is reduced. 
 
 According to Professor Capper's report, from which we 
 have produced the illustration, this engine gave no trouble 
 in working. The average revolutions on each of the three 
 days of the Cambridge trials did not vary, though there 
 was some racing of the engine in spite of the ingenious 
 method of governing. 
 
234 
 
 THE CAPITAINE OIL ENGINE. 
 
 The Capitaine Oil Engine is made by Messrs. Tolch and 
 Co., of London, and its distinguishing features are shown 
 in figs. 117 and 118. Fig. 117 shows the upper end of the 
 vertical combustion chamber. On the outstroke of the 
 piston air enters the automatic valve shown at B. The 
 greater volume of air passes into the cylinder round the 
 outside of the case D, whilst some air necessarily passes 
 through the conical hole C. The part C is kept at a high 
 temperature by the explosion, its heat being retained by 
 the non-conducting substance surrounding it. 
 
 Fio. 117. Section through Vaporiser and Combustion Chamber of Tolch and Co.'i 
 Capitaine Engine. 
 
 Oil enters by the pipe A, and, passing through the hole in 
 the air valve B, drops into C, and is vaporised. The central 
 spindle within the valve B closes the oil passage to C, when 
 the valve is in its uppermost position. The governing is 
 effected by holding open the exhaust valve, so that B 
 
THE CAPITAINE OIL ENGINE. 
 
 235 
 
 remains closed during the outstroke, and no air or oil is 
 passed into the cylinder. 
 
 The delivery of oil to the pipe A is effected by the plunger 
 pump B, shown in fig. 118 This pump works in a glycerine 
 bath. Oil floats upon the surface of the glycerine and passes 
 into the pipe A, fig. 117, through the slide valve A, fig. 118. 
 
 FIG. 118. Glycerine Pump of Tolch and Co.'s Capitaine Engine. 
 A Slide Valve. B Pump Plunger. 
 
 Professor Capper says : " The oil ordinarily used in this 
 engine is Tea Rose, and some difficulty was experienced in 
 keeping the vaporiser hot enough to work with Russolene. 
 Experience with the use of Russolene oil may be expected to 
 overcome these difficulties, and the engine certainly deserves 
 praise for its particularly quiet running, and for the ingenuity 
 as well as simplicity of its working parts." 
 
236 
 
 TANG YE' S OIL ENGINE. 
 
 Tangye's vaporiser, made under Pinkney's patents, is 
 shown in fig. 119. The vaporising chamber is in direct 
 communication with the cylinder. Air is drawn in through 
 the mushroom valve, and meets oil fed by gravity from a 
 tank to the hole on the valve seating. Thus, when the 
 valve opens by the suction of the piston, oil drops through 
 into the vaporiser. To govern the engine the exhaust valve 
 
 Fio. 119. Tangye's Vaporiser. 
 
 is kept open and the automatic valve remains closed during 
 the suction stroke, thus preventing the admission of fresh 
 air or oil to the cylinder. A lamp placed beneath the ignition 
 tube heats both the vaporiser and the tube. Before starting, 
 the lamp is placed immediately beneath the vaporiser. 
 
 The Fielding Oil Engine is constructed by Messrs. Fielding 
 and Platt Limited, of Gloucester. The engine works in a 
 similar way to a gas engine, with the exception of the 
 additional parts required to vaporise the oil. A longitudinal 
 
THE FIELDING OIL ENGINE. 
 
 237- 
 
 section of the vaporiser is shown in fig. 120. This vaporiser 
 consists of two tubes, one above the other, the lower one 
 being heated by the lamp to a bright red, whilst the upper 
 one is kept at a lower temperature. Two copper tubes, at a 
 high temperature, connect the air inlet shown to the upper 
 vaporiser tube. V is an automatic valve, drawn open by 
 the outstroke of the engine piston. The action of the engine 
 
 FIG. 120. The Fielding Vaporiser. 
 
 is as follows : During the outstroke of the engine piston, air 
 is drawn into the cylinder through a valve in the combustion 
 chamber (not shown). The valve V also opens, and a current 
 of hot air sweeps through the ignition tube into the cylinder. 
 Oil vapour, which has been pumped into the vaporiser tube 
 mixes with this air, but is not ignited by the ignition tube 
 because of the inferior quantity of air drawn through the 
 valve V. When the vapour charge enters the cylinder it 
 
238 
 
 CLAYTON AND SHUTTLEWORTH IGNITER. 
 
 mingles with the pure air entering by another valve, and 
 thus forms an explosive mixture, which ignites when part of 
 it is compressed back into the ignition tube. Fig. 121 shows 
 an end view of the engine. 
 
 The speed is controlled by a centrifugal governor, which 
 props up the exhaust valve lever and the oil pump rod. 
 Hence no oil is injected into the vaporiser, nor is pure air 
 
 /7//? VffL V TO 
 VfiPOR/SER. 
 
 Fio. 121. Elevation showing Arrangement of Valve Levers and Trip Gear of 
 Fielding and Platt's Oil Engine. 
 
 drawn into the cylinder through either of the valves. This 
 engine ran very steadily during the Cambridge trials, start- 
 ing readily with one attendant after the vaporiser tubes 
 were heated for about 20 minutes. This engine was, how- 
 ever, withdrawn from the competitive trials on account of 
 slight defects in the heating lamp, which have since been 
 remedied. 
 
 A section of Messrs. Clayton and Shuttle worth's method 
 of igniting is shown in tig. 122. A steel needle projects into 
 
THE DAIMLER MOTOR. 239 
 
 the combustion chamber. This needle is surrounded by a 
 number of strips of asbestos millboard. The heat retained 
 by this arrangement keeps the needle at a sufficiently high 
 temperature to ignite the charge when it is compressed. 
 A tube shown in the illustration serves for ignition when 
 starting the engine. The igniter is said to work satisfac- 
 torily even at low loads. 
 
 As illustrating a class of oil engines working with light 
 oil now upon the market, we reproduce a section of tho 
 
 FIG. 122. -Section of Clayton and S buttle worth's Oil Engine Igniter. 
 
 Daimler motor, tig. 123. This engine has been very success- 
 fully applied to motor cars, and works FS follows : The 
 small tank P is fed with benzolene. A float in the tank 
 operates a check valvf, which keeps the oil at a proper level 
 in the tank. Upon the outstroke of the piston air enters 
 the passage g, and, passing over the oil pipe v, induces a 
 small quantity of benzolene to pass into q in the form of a 
 fine spray. The air and benzolene pass into the cylinder 
 through the automatic valve m. The piston returns, closes 
 ?7i, and compresses the charge into the igniting tube o. The 
 
240 
 
 THE DAIMLER MOTOK. 
 
 exhaust takes place through the valve n, which is lifted 
 by the rod X. The governing is effected by holding the 
 
 FIG. 123. Daimler Motor (1895). 
 
 exhaust valve open, so that the burnt gases return to the 
 cylinder, thereby destroying the partial vacuum, which 
 would otherwise open the valve m. 
 
OIL ENGINE TESTING. 241 
 
 CHAPTER XXIII. 
 OIL ENGINE TESTING. 
 
 IN a previous chapter we have described in great detail the 
 method of testing gas engines. Much that has been said in 
 reference thereto will readily be applied to the testing of 
 oil engines. There are, however, some points arising in 
 connection with oil-engine testing which require special 
 consideration. 
 
 In the first place, it is convenient to arrange the oil supply 
 in tanks calibrated to give the weight of oil. The calibration 
 should be separately determined for each class of oil used, on 
 account of the different density of the brands of oil now 
 upon the market. 
 
 In testing gas engines the quantity of air per stroke is 
 not difficult to obtain by difference, when the volume of 
 gas in the cylinder is known. The measurement of the 
 air supplied for combustion of the charge in an oil engine 
 may present some difficulty, but unless this be determined 
 the heat account cannot be ascertained. When air is 
 supplied by a pump, an estimate of the volume may be made 
 from the pump dimensions, though this is not satisfactory. 
 It may, in some instances, be necessary to measure the air 
 by arranging a suitable air trunk, and in it placing an 
 anemometer. In the author's opinion this instrument is 
 not reliable ; nevertheless, if it be carefully tested, a correc- 
 tion curve may be plotted, by the application of which 
 fairly accurate results may be obtained. 
 
 In dealing with gas-engine trials, a simple method of 
 analysing the gases used was described. We shall now 
 describe an equally simple method of making a determination 
 of the carbon and hydrogen in a sample of petroleum. 
 Although most engineers, in conducting trials, usually place 
 samples for analysis in the hands of trained chemists, it is, 
 
 17G 
 
242 T^TltOLEUM ANALYSIS. 
 
 nevertheless, a great advantage, when much testing is being 
 done, to be able to perform these simple analyses oneself. 
 
 The various petroleum oils consist chiefly, as we have 
 previously pointed out, of carbon and hydrogen. The sum 
 of the weights of these two substances is found to be very 
 nearly 100 per cent. The undetermined weight may amount 
 to 5 per cent, and this may be neglected in engine trial 
 calculations. The method about to be described is due to 
 Liebig, who made use of the fact that if the substance to be 
 analysed be mixed with an excess of copper oxide, and the 
 whole be heated to redness, the carbon will reduce the oxide 
 and pass off as carbon dioxide; at the same time the hydrogen 
 will pass off as water vapour. The carbon dioxide is 
 absorbed by caustic potash; the water is absorbed by 
 calcium chloride. The additional weight of these substances 
 after absorption has taken place gives the carbon and 
 hydrogen. 
 
 The analysis may be done in the following way : A com- 
 bustion tube of hard glass about f in. bore and 2 ft. long, 
 open at one end and drawn out to a point at the other, is 
 supported upon a light iron stand. A row of Binsen 
 burners is placed beneath the tube, and a sheet of asbestos 
 arranged round the tube and the flames, in order to con- 
 centrate the latter upon the tube. Take about 5 gram of 
 petroleum, and place it in a small soft glass tube, the weight 
 of which (empty) is known. Fill the tube, seal at both 
 ends, and carefully weigh again. Subtract from the total 
 weight, when full, the weight of the glass tube, and so obtain 
 the net weight of petroleum. When heavy oil is being 
 analysed the tube need not be sealed, but in the case of light 
 volatile oils the evaporation constantly going on at atmo- 
 spheric temperature renders it impossible to determine 
 the weight accurately unless the tube is sealed. Place 
 the sealed tube near the closed end of the combustion tube. 
 Then fill the remainder of the combustion tube with copper 
 oxide. 
 
 Attach to the open end of the combustion tube a bulb 
 
PETROLEUM ANALYSIS. 243 
 
 containing porous calcium chloride, so that all the gas 
 generated passes through this bulb first. From this bulb 
 the gas must pass into another vessel, containing a concen- 
 trated solution of caustic potash. When the weights of 
 these vessels are ascertained, and all the joints of the tubes 
 carefully tested, the analysis may be made. First light the 
 burners near the open end of the combustion tube until the 
 copper oxide becomes heated. Gradually heat the tube 
 towards the sealed end. When the petroleum sample becomes 
 hot the soft glass vessel in which it is contained will burst 
 by the expansion of the liquid, and the process of reduction 
 will commence. To prevent a violent explosion, the liquid 
 should completely fill the sealed vessel. The products of 
 combustion will bubble through the bulbs until the petroleum 
 is consumed. When the gases cease to pass through, break 
 off the sealed end of the combustion tube, and inspire the 
 gaseous contents of the latter into the absorption vessels by 
 means of a rubber tube. 
 
 Weigh the absorption bulbs after the process is completed, 
 and so obtain the weights of the carbon dioxide and water. 
 The carbon dioxide is composed of 12 parts of carbon to 32 
 parts of oxygen ; hence it or rr of the weight of carbon 
 dioxide gives the weight of carbon in the petroleum sample. 
 The weight of hydrogen will be found by multiplying the 
 weight of water formed by i. 
 
 This method is satisfactory, provided you have a large 
 excess of freshly ignited copper oxide in the combustion 
 tube. There is, however, a liability to the formation of 
 carbon in the combustion tube, in which case the analysis is 
 incomplete. A modification of the above apparatus is as 
 follows : Arrange the combustion tube with a free passage 
 through both ends. To the end of the potash absorption 
 bulb attach an aspirator tube, so that air or oxygen may be 
 passed over to consume the petroleum. Whether air or 
 oxygen be. used, a calcium chloride drier should be attached 
 to the combustion tube where the air enters, in order to 
 absorb all moisture that might be otherwise carried ovei 
 
244 CALORIFIC VALUE OF OIL. 
 
 into the absorption bulb, and credited to the hydrogen. 
 This method has the advantage that the reduced copper may 
 be re-oxidised by passing over a large excess of oxygen while 
 the analysis is being carried out. 
 
 With respect to the heat given to the engine, this may be 
 calculated from the calorific value of the oil. To determine 
 the calorific value of oil two methods may be adopted. The 
 determination may be made by calculation from the analysis, 
 or by the direct combustion of a known quantity of oil in 
 some form of calorimeter. Neither method is in itself quite 
 satisfactory. Calculated values are often found to be higher 
 than those obtained by calorimetric tests. On the other 
 hand, the calorimeters at present available are not perfectly 
 satisfactory. 
 
 To calculate the heat value of oil from the analysis 
 Let C = the percentage of carbon by weight ; 
 
 H = the percentage of hydrogen by weight. 
 Then, 14500 jC + 4'28H( = total heat value per pound of oil. 
 
 This value includes the heat given up by cooling all the 
 products of combustion to atmospheric temperature (60 deg. 
 Fah.). For each pound of oil burnt about If Ib. of water 
 (according to the weight of hydrogen) is formed. If, there- 
 fore, this product passes off in the form of steam, as is 
 inevitable in the case of internal combustion engines, the 
 question arises, should the latent heat of the steam be deducted 
 from the calorific value in working out the heat efficiency of 
 the motor *? The answer to the question is more a matter of 
 opinion than of right. Both methods have been adopted in 
 the treatment of trials. This point should, however, be 
 referred to when reporting an engine trial, and the method 
 adopted should be clearly stated. The author prefers to 
 deduct from the calorific value of the oil the units of heat 
 passing away with the steam. 
 
 The following example of this calculation will suffice : 
 Upon analysis the oil yields Carbon, 84*92 per cent; 
 hydrogen, 15 '03 per cent ; undetermined, 0'05 per cent. 
 
CLASSIFICATION OF VAPORISERS. 245 
 
 14,500 is taken as the calorific value of carbon, and 14,500 
 x 4 '28 is the calorific value of hydrogen ; hence the formula 
 
 14500 {C + 4'28 H} = total heat. 
 In the example given, 
 
 14500 {0-8492 + 4 28 x 01503} = total heat, 
 
 = 21641 B.T.U. 
 
 But lib. of oil will yield 9 x 015031b. of water. Each 
 pound of water will carry away 966 units ; therefore, the 
 units to be deducted from the total value = 9 x 1503 x 
 9t>6 = 1306. 
 
 Therefore effective calculated value 
 = 21641 - 1306, 
 = 20335 units. 
 
 With respect to the indicating of oil engines, there is 
 great difficulty in obtaining satisfactory diagrams with some 
 engines, more especially when the engine is being forced by 
 the use of excessive oil supply. Hence the indicated horse 
 power is seldom reliable, and in all cases comparisons should 
 be made on the brake power. 
 
 After reading the description of the vaporisers used by the 
 leading makers of oil engines, it will be noticed that all 
 engines working with heavy oils vaporise in one of the 
 following ways : (1) Oil is injected, together with the whole 
 of the air supply, into a large vaporising chamber, separate 
 from the cylinder, or (2) oil is injected into a small 
 vaporising chamber, together with some air, but the greater 
 volume of air enters the cylinder through a separate air 
 valve ; (3) same as (2), but all the air goes through the 
 vaporiser ; (4) oil is injected into the combustion chamber, 
 and is vaporised therein. Air is drawn into the cylinder 
 through a separate air valve, and mingles with the oil 
 vapour when compressed into the combustion chamber. 
 The first method undoubtedly possesses the advantage that 
 the mixture of the vapour with the air is more complete. 
 
 Detracting somewhat from this, there is the fact that a 
 large volume of highly-explosive vapour is contained in the 
 
246 CLASSIFICATION OF VAPORISERS. 
 
 vaporiser, and back firing might occur with serious results. 
 It will be observed that in this type of engine the whole of 
 the air is heated before entering the cylinder; hence the 
 weight of the charge is considerably reduced, and the mean 
 pressure thereby reduced. It is sometimes suggested that 
 difficulties in the governing are encountered by the use of 
 this vaporiser. For if the explosions are discontinued by 
 the cutting-out system of governing, then the vaporiser 
 cools. Messrs. Priestman obviate this by their regulation 
 of the oil and air supply, thus keeping the engine running 
 even at light loads without missing fire. This method 
 seems to be prejudicial to the economic consumption of oil 
 at light loads. According to Professor Unwin's papei* on 
 the Priestman oil engine, the consumption, when running 
 with TIO load, was over 5 Ib. of oil (Eussolene) per hour 
 whereas when the engine was doing nearly 7 brake horse 
 power the oil consumption was only 3 Ib. per hour that is, 
 0*988 Ib. per brake horse power. Thus it appears that the 
 method of reducing the oil supply, even though the air is 
 also reduced, is prejudicial to economy. 
 
 From an analysis of the trials (see table of oil engine 
 dimensions and consumptions) there appears little to choose 
 between the various classes of vaporisers. It is important, 
 however, to note the difference existing between vaporisers 
 working with all the air passing through and those with 
 only a portion. The cooling effect of the air in the 
 former case should be carefully considered in designing the 
 vaporiser. Results show that either methods give excellent 
 results, as far as consumption. The mean pressure in 
 Crossley's and Fielding's engines exceed those of other 
 types. The mean pressure in the Hornsby-Akroyd engine 
 (of class 4) is lower than any other. A large size of cylinder 
 is therefore required. This is, however, compensated for by 
 simplicity of parts and general neatness. 
 
 * Inst. C.E. Proceedings. Vol. cix. 1892. 
 
I 
 
 
 
 Professor Unwin 
 
 Professor Capper 
 
 Professor Capper 
 
 1 
 1 
 
 1 
 "p 
 
 
 
 cod 
 
 i* 
 
 ^ 
 
 Professor Capper 
 
 Professor Capper 
 
 Professor Capper 
 
 ho 
 11 
 
 si 
 
 1 
 
 g^nuiui jgd 
 snoimiOAQg 
 
 s o 
 
 5! 
 
 o 
 
 1 
 
 1 
 
 oo 
 
 CM 
 
 1 
 
 CO 
 
 
 CM 
 
 m suoiqinggp ggg) 
 .lasuodtJA jo ssujo 
 
 
 
 o 
 
 
 CO 
 
 o 
 
 fe 
 
 CO 
 
 o 
 
 A 
 
 co 
 
 cJ 
 
 
 ll 
 be 
 
 Variable oil and air 
 supply 
 
 Variable oil supply 
 
 'o 
 
 f! 
 
 d 
 
 Exhaust kept open. 
 Oil inte: cepted 
 
 g 
 
 1 
 
 kept open. 
 ;ercepted 
 
 a 
 1 
 
 hrottled in 
 f cylinder. 
 
 "3 
 
 O 
 
 
 Exhaust 
 Oil in 
 
 O 
 
 Charge t 
 enterin 
 
 1 
 
 Oil con- 
 sumption 
 at half load. 
 
 unoq 
 
 <TH-a 
 
 jl| 
 
 ? 
 
 CO 
 
 co 
 
 ? 
 
 g 
 
 
 
 g 
 
 CO 
 
 g 
 
 anoq 
 J HI -isj 
 
 
 ' 
 
 : 
 
 
 o 
 o 
 
 
 : 
 
 : 
 
 
 " ~cS 
 
 unoq 
 M'HTI 
 
 75 1 
 
 1-1 
 
 OJ 
 
 
 o 
 
 ? 
 
 1 
 
 9 
 
 o 
 
 ? 
 
 1 
 
 anoq 
 
 o 
 
 CO 
 
 o 
 
 CO 
 
 o 
 
 o 
 o 
 
 
 
 o 
 
 P 
 
 o 
 
 2 
 
 
 1 
 
 i 
 
 1 
 
 jl 
 
 Russolene 
 
 s 
 
 OS 
 
 Russolene 
 
 11 
 
 Russolene 
 
 
 1 
 
 Benzolinc 
 
 Russolene 
 
 Maximum. 
 
 ^uqo?W 
 
 ls 
 
 CO 
 
 % 
 
 co 
 
 00 
 
 S 
 
 t t 
 
 
 
 i 
 
 ^ 
 
 CO 
 
 9 
 
 v> 
 
 1 
 
 S 
 
 s 
 
 
 1 
 
 M-ITI 
 
 eo 
 os 
 
 CO 
 
 o 
 
 
 
 co 
 
 s 
 
 o 
 10 
 
 CO 
 
 
 ' 
 
 Full load 
 indicator 
 diagrams. 
 
 U ( SS9-ldtUOQ 
 
 ^- s 
 
 s 
 
 CO 
 
 
 00 
 
 .8 
 
 
 
 
 % 
 
 9J uSw d 
 
 ^s 
 
 1 
 
 t2 
 
 o 
 
 5 
 
 o 
 
 t- 
 
 
 g 
 
 ganssaad 
 
 ^ '^ iO 
 
 * 0} ^ 
 
 1 
 
 1 
 
 
 s 
 
 00 
 
 ^ 
 
 : 
 
 CO 
 
 Cylinder 
 
 frizes. 
 
 90^9^0 
 
 -II 
 
 CO 
 CO 
 
 o 
 
 CM" 
 
 1 
 
 
 
 1 
 
 
 
 FMW 
 
 .9 S 
 
 s 
 
 2 
 
 US 
 
 3 
 
 ^ 
 
 2 
 
 CO 
 
 o 
 
 a^nnrKl 
 
 .2 oo" 
 
 
 
 - 
 
 % 
 
 s? 
 
 J* 
 
 I? 
 
 g 
 
 CO 
 
 ' It 
 
 Priestman 
 
 Hornsby-Akroyd. 
 
 1 
 
 Wells' "Premier" 
 
 Weyman & Hitch- 
 cock's "Trusty" 
 
 Camplell 
 
 Britaunia 
 
 Clai ke-Charman . 
 
 Fielding and Platt 
 
INDEX TO ILLUSTRATIONS. 
 
 A 
 
 Abel Oil Tester 207 
 
 Atkinson's Cycle Engine 20, 23 
 
 Atkinson's Differential Engine 21 
 
 Averager for Indicator Diagrams, Coffin 149 
 
 Averager for Indicator Diagrams, Goodman 146 
 
 B 
 
 Britannia Governor 233 
 
 Britannia Vaporiser 232 
 
 c 
 
 Calorimeter, Junker's 119, 121 
 
 Campbell Vaporiser 231 
 
 Capitaine Vaporiser 234 
 
 Carburator, Simplex 213 
 
 Clayton and Shuttleworth Igniter 239 
 
 Clerk-Lanchester Starter 76 
 
 Clerk Pressure Starter 75 
 
 Clerk Two-cycle Engine 83 
 
 Coffin Averager 149 
 
 Crosby Indicator 99, 100 
 
 Crossley Cylinder Section 35, 36 
 
 Crossley Double-cylinder Gas Engine 31 
 
 Crossley Tandem Gas Engine 29 
 
 Crossley Tube Igniter 32 
 
 Crossley Vaporiser 226 
 
 Crossley Weak-Spring Diagram 34 
 
 Cycles, Diagram of Gas Engine 26, 46 
 
 D 
 
 Daimler Oil Engine 240 
 
 Day Gas Engine 78 
 
 Diagrams of Combustion Experiments 198 
 
 Diagrams of Combustion, Stage 192 
 
 Diagrams of Combustion 188191 
 
 Diagrams of Comparatives Cycles 26, 46 
 
 Diagrams of Heat and Pressure of Explosion 197 
 
 Diagrams, Indicator, from Dougill Engine 81 
 
 Diagrams, Indicator, from Fielding Gas Engine 60 
 
 Diagrams, Indicator, from Hornsby Oil Engine 224 
 
 Diagrams, Indicator, from Scavenging Engine 33, 34 
 
 Diagrams, Indicator, from Self-starters 57 
 
 Diagrams, Indicator, from Wayne Indicator 107 
 
 Diagrams, Indicator, from Wells Gas Engine 66 
 
 Diagrams showing Effects of Products of Combustion 188, 191, 192, 197, 198 
 
 Dougill Gas-engine Governor 80 
 
 Dowson Gas Plant 179, 18C 
 
INDEX TO ILLUSTRATIONS. 249 
 
 E PAGE 
 
 Electric Igniter 186 
 
 Explosion Apparatus, Experimental 183 
 
 F 
 
 Fielding Gas Engine 59 
 
 Fieldiug Oil Engine ; 237 
 
 Forward Gas Engine 68, 69 
 
 G 
 
 Gas Analysis, Hempel's Apparatus 127131 
 
 Gas Engine, Crossley 2936 
 
 Gas Engine, Day 78 
 
 Gas Engine, Dougill 80 
 
 Gas Engine, Fielding 59 
 
 Gas Engine, Fo. ward 68, 69 
 
 Gas Engine, Griffin 47 
 
 Gas Engine, Kilmarnock 49, 51 
 
 Gas Engine, Koerting-Lieckfeldt 90 
 
 Gas Engine, Otto E*rly 9 
 
 Gas Engine, Stockport 3844 
 
 Gas Engine, Tangye 55 
 
 Gas Engine, Wells Premier 66 
 
 Green Self-starter 73 
 
 Griffin Gas Engine 47 
 
 Griffin Oil Sprayer 221 
 
 Grover Indicator Gear 110 113 
 
 H 
 
 Hempel's Gas Analysis Apparatus 127 
 
 Hornsby Oil Engine 223 
 
 Indicator Diagrams from Crossley Engine 33, 34 
 
 Indicator Diagrams from Dougill Engine 81 
 
 Indicator Diagrams from Fielding Gas Engine 60 
 
 Indicator Diagrams from Hornsby Oil Engine 224 
 
 Indicator Diagrams from Self-starters 57 
 
 Indicator Diagrams from Wayne Indicator 107 
 
 Indicator Diagrams from Wells Gas Engine , 66 
 
 Indicator Testing 97 
 
 Indicator, Crosby 99, 100 
 
 Indicator, Simplex 108 
 
 lunicator, Tabor 101103 
 
 Indicator, Wayne 105, 106 
 
 J 
 
 Junker's Calorimeter 119121 
 
 K 
 
 Kilmarnock Gas Engine 49, 51 
 
 Koerting-Lieckfeldt Gas Engine 9.5 
 
 L 
 Lanchester Self-starter 72 
 
 M 
 
 lloscrop Recorder Diagram 45 
 
250 INDEX TO ILLUSTRATIONS 
 
 o 
 
 Oil Engines (see Vaporisers). 
 
 Otto Early Gas Engine , 
 
 P 
 
 Pipettes for Gas Analysis 128131 
 
 Pries tman Oil Engine 216218 
 
 s 
 
 Sanrielson Spray Maker 222 
 
 Self-starter, Clerk 75 
 
 Self-starter, Clerk-Lan Chester 76 
 
 Self-starter, Green 73 
 
 Self-starter, Lanchester 72 
 
 Simplex Carburator 213 
 
 Simplex Indicator 108 
 
 Spiel Oil Engine 210-212 
 
 Stockport Gas Engine 38 -44 
 
 T 
 
 Tabor Indicator 101103 
 
 Tangye's Gas Engine 55 
 
 Tangye's Oil Engine 236 
 
 Tangye's Self-starter 56 
 
 Trusty Oil Engine 22S, '22'J 
 
 V 
 
 Vaporiser, Britannia 232 
 
 Vaporiser, Capitaine 234 
 
 Vanoriser, Campbell 231 
 
 Vaporiser, Crossley 226 
 
 Vaporiser, Fielding 237 
 
 Vaporiser, Griffin Sprayer 221 
 
 Vaporiser, Hornsby 223 
 
 Vaporiser, Priestman 217, 218 
 
 Vaporiser, Sarnuelson Sprayer 222 
 
 Vaporiser, Spiel Sprayer 211 
 
 Vaporiser, Tangye 236 
 
 Vaporiser, Trusty 228, 22 
 
 Vaporiser, Wells 2-7 
 
 w 
 
 Wayne Indicator 105, 106> 
 
 Weak Spring Indicator Diagram 34 
 
 Wells Premier Gas Engine 6f> 
 
 Wells Oil Engine 227 
 
INDEX TO SUBJECT MATTER. 
 
 A P-'GE 
 
 Abel Oil Tester 207 
 
 Absorption of Gases 126 
 
 Acme Gas Engine 81 
 
 Air, Composition of 13!) 
 
 Air Proportions in Explosive Mixtures 190 
 
 Air Pulsation in Gas-engine Rooms 10 
 
 Air, Weight of 139 
 
 Amsler's PlaTiimeter 150 
 
 Analysis of Coal Gas 12(5, 135, 189 
 
 Analysis of Dowson Gas 182 
 
 Analysis of Petroleum 242 
 
 Appendices, Gas-engine Trial Results 203 
 
 Appendices, Products of Combustion Experiments 200 
 
 Atkinson's Differential Engine 19 
 
 Atkinson's Cycle Engine 22 
 
 Averager fur Indicator Diagrams, Coffin's 149 
 
 Averager for Indicator Diagrams, Goodman's 146 
 
 B 
 
 Barber's Patent Gas Engine 2 
 
 harnard's Patent Gas Engine 5 
 
 Bariiett s Patent Gas Engine 3 
 
 Barsanti and Mat tenci's Engine 3 
 
 Beau de Rochas Cycle 5 
 
 Bearings, Pressures on 70 
 
 Belt Drives ! IS 
 
 Benz Gas Engine 88 
 
 Brayt oil's Oil Engine 211 
 
 Britannia Oil Engine 232 
 
 Brown's Engine 2 
 
 C 
 
 Calorific Value Calculations 120, 245 
 
 Calorimeter, Junker's . . 117 
 
 Campbell Oil Engine 231 
 
 Capitaine Gas Engine 92 
 
 Capitaine Oil Engine 234 
 
 Carbon-dioxide, Weight of 140 
 
 Carbon-monoxide. Ignition Temperature of 1 96 
 
 Carbon, Weight of 141 
 
 Carbon, Calorific Value of 140 
 
 Charon Gas Enoine 87 
 
 Circulating Tanks, Arrangement of 17 
 
 Cler k 's Self-starter 75 
 
 ClTk-Lanch ester Self-s'arter 76 
 
 Clayton and Sbuttleworth's Igniter 2S' 
 
 Coal-gas Constituents , 1-2(5, 140 
 
 Coal Ga?, Analyses and other Da f a 135, 138, MO, IsD 
 
252 INDEX. 
 
 PAGE 
 
 Coffin Averager 149 
 
 Combustion Chamber, Size of 167 
 
 Comparative Efficiencies of Engines with different Dimensions, but having 
 
 the same Compression 204 
 
 Compounding Gas Engines 6 
 
 Compression, Amount of 85 
 
 Connecting Rods 1 174 
 
 Consumption of Gas by Engines 22, 24, 50, 54, 71, 61, 155 
 
 Consumption of Oil by Engines 247 
 
 Crauk Shafts 168 
 
 Crosby Indicator 99 
 
 Crossley Gas Engine 27 
 
 Crossley Oil Engine 225 
 
 Cycles of Gas E gines 25, 26, 46 
 
 * ycle Engin , Atkinson's 22 
 
 Cylinder Proportions 163165 
 
 D 
 
 Daimler Oil Engine 240 
 
 Daimler Gas Engine 91 
 
 Day Gas Engine 77 
 
 Differential Gas Engine 19 
 
 Dougill Governor 79 
 
 Dowdon Gas Analyses 138, 182 
 
 Dowson Gas, Air Required for Combustion of 182 
 
 Dowson Gas Plant 179 
 
 Dynamometer, Absorption . . 93 
 
 Economy, Effects of Governor on 246 
 
 Effects of Products of Combustion 183198 
 
 Efficiencies of Gas Engines 158, 204 
 
 Electric Igniter 186 
 
 Engine-room for Gas Engine 7 
 
 Engines, Gas (see Gas Engines). 
 
 Engine Trials, Analysis of Gas 12 '138 
 
 Engine Trials, Calculations for Reports 1391*8 
 
 Engine Trials, Calorimetry 117 
 
 En ,ina Trials, Gas Measur 6 ment 117 
 
 Engine Trials, Indicating vti, 10t, 144 
 
 Engine Trials, Reducing Gear 109, 114 
 
 Engine Trials, Water-jacket Measurements 115 
 
 Exhaust Pipe for Gas Engine 14 
 
 Exhaust, Temperature of 124 
 
 Explosive Mixtures, Proportions 190 
 
 Flashing Point 206 
 
 Flywheels, Sizes of 37, 171 
 
 Fielding's Gas Engine 58 
 
 Fielding's Oil Engine 236 
 
 Forward Gas Engine 68 
 
 Friction Brake 93 
 
 Frost in Water Jackets 39 
 
 G 
 
 Gas Analysis 126, 138, 189 
 
 Gas Bags on Mains 13 
 
 Gas, Temperature of Exhaust 124, 151 156 
 
 G-iS, Exhaust, Loss of Heat 157 
 
 Gas Measurements on Trials 117 
 
 Gas Producer 176, 183 
 
 Gas Engine, Atkinson's Cycle 22 
 
 G is Engine, Atkinson's Differential 19 
 
 Gas Engine, Barber's 2 
 
 Gas Engine, Barnard's 5 
 
INDEX. 253 
 
 PAGE 
 
 Gas Engine, Barnett's 3 
 
 Gas Engine, Barsanti 3 
 
 Gas Engine, Beau de Rochas 5 
 
 Gas Enyine, Benz 88 
 
 Gas Engine, Brown's 2 
 
 Gas Engine, Capitaine 92 
 
 Gas Engine, Charon 87 
 
 Gaa Engine, Crossley 27 
 
 Gas Engine, Daimler 91 
 
 Gas Engine, Day 77 
 
 Gas Engine, Design of 159 
 
 Gas Engine, Dougill 79 
 
 Gas Engine, Efficiency of 160 
 
 Gas Engine, Fielding 58 
 
 Gas Engine, Forward 68 
 
 Gas Engine, Griffin 46, 53, 82 
 
 Gas Engine, Kilmarnock : 47 
 
 Gas Engine, Everting- Lieckfeldt 89 
 
 Gas Engine, L-albin 8S 
 
 Gas Engine, Lebon 2 
 
 Gas Engine, Lenoir 87 
 
 Gas Engine, Norris 84, 85 
 
 Gas Engine, Niel S3 
 
 Gas Engine, Otto 8, 27 
 
 Gas Engine, Palatine 79 
 
 Gas Engine, Premier 61 
 
 Gas Engine, Robey 84 
 
 Gas Engine, Simplex 86 
 
 Gas Engine, Stockport 36, 37 
 
 Gas Engine Trials with Goal Gas 203 
 
 Gas Engine Trials with Dowson Gas 203 
 
 Gas Engine, Tangye 53 
 
 Gas Engine, Weights of 37 
 
 Goodman Averager 146 
 
 Governor, Dnugill 79, 80 
 
 Governor, Stockport 43,44 
 
 Governing Oil Engines 246 
 
 Grashoff's Formula 152 
 
 Griffin Gas Engine 46, 53, 82 
 
 Green's Self-starter 74 
 
 Grover's Indicator Gear 110113 
 
 LJ 
 
 Heat Account of Gas Engine 158 
 
 Hempel's Gas Apparatus 126 
 
 History of the Gas Engine 17 
 
 Hornsby-Akroyd Oil Engine 223 
 
 Horse Power, Brake 12, 96, 92 
 
 Horse Power, Nominal 12 
 
 Horse Powers, Weights, and Overall Dimensions of Gas Engines 37 
 
 Hydrogen, Calorific Value 140 
 
 Hydrogen, Ignition Temperature 196 
 
 Hydrogen, Weight of 140 
 
 Igniter, Electric 186 
 
 Ignition Temperatures of Coal-gas Constituents 196 
 
 Ignition Tubes, Crossley 3032 
 
 Ignition Tubes, Stockport 39-41 
 
 Indicator Diagrams, Averagers for 146 149 
 
 Indicator Diagrams, Calculations on 144 
 
 Indicator Diagrams, Dougill Governor 81 
 
 Indicator Diagrams, Light Loads 60 
 
 Indicator Diagrams, Mean Pressures of Oil Engines 247 
 
 Indicator Diagrams, Premier Gas Engine 66 
 
 Indicator Diagrams, Starting 57 
 
 Indicator Diagrams, Wayne In dtcator 107 
 
 Indicator Diagrams, Weak-Spring 34 
 
254 INDEX. 
 
 PAOE 
 
 Indicators, Crosby 99 
 
 Indicators, Reducing Gears 109 
 
 Indicators, Simplex 108 
 
 Indicators, Tabor 101 
 
 Indicators, Testing Springs 97 
 
 Indicators, Wayne 104 
 
 J 
 
 Jacket Water, Measuring Quantity of 1(3, 115 
 
 Jacket Water Temperatures 116 
 
 Junksr's Calorimeter 117 
 
 ^f 
 
 Kilmarnock Engine 47 
 
 Koerting-Lieckfeldt Engine 89 
 
 Lalbin Engine 88 
 
 Lanchester Starter 72 
 
 Langen Engine 7 
 
 Lead on Gas Valve 163 
 
 Lebon Engine 2 
 
 Lenoir Gas Engine 4, 87 
 
 Lenoir Oil Engine 210 
 
 Marsh Gas, Calorific Value 140 
 
 Marsh Gas, Determination of 135138 
 
 Marsh Gas, ignition Temperature 196 
 
 Marsh Gas, Weight of 140 
 
 Mean Pressures in Oil Engines 247 
 
 Mechanical Efficiencies of Gas Engines 30, 71, 156 
 
 3Iethods of Governing Oil Engines 247 
 
 Million 5 
 
 Moscrop Recorder Diagram 45 
 
 N 
 
 Kiel's Engine 88 
 
 Nitrogen. Weight of 140 
 
 Norris Engine 84, 5 
 
 o 
 
 Oil, Analysis of 242 
 
 Oil, Boiling Points 209, 247 
 
 Oil, Calorific Value of 24t 
 
 Oil, Colour of 209 
 
 Oil Engine, Brayton 211 
 
 Oil Engine, Britannia 232 
 
 Oil Engine, Campbell 231 
 
 Oil Engine, Capiraine 234 
 
 Oil Engine, Clayton and Shuttle-worth's 239 
 
 Oil Engine, Consumption of Oil 247 
 
 Oil Engine, Crossley 225 
 
 Oil Engine, Daimler 240 
 
 Oil Engines, Dimensions and Consumptions of 247 
 
 Oil Engine, Economy Affected by Governor 246 
 
 Oil Eng-iie, Fielding 247 
 
 Oil Engine, Governing 247 
 
 Oil Engine, Hornsby 223 
 
 Oil Engine, Lenoir 210 
 
 Oil Engine, Mean Pressures of Indicated Diagram 247 
 
 Oil Engine, Priestman 215 
 
 Oil Engine, Samuelson 220 
 
INDEX. 255 
 
 P(.E 
 
 Oil E -gine, Simplex 213 
 
 Oil Engine, Spiel 21] 
 
 Oil Engine, Tangye 236 
 
 Oil Engine, Testing 241 
 
 Oil Engine, Trusty 228 
 
 Oil Engine, Wells 226 
 
 Oil, Flashing Point 206 
 
 Oil, Legal Restrictions in Using 206 
 
 Oil, Products of, by Distillation 206 
 
 Oil, Specific Gravity of 209, 210, 247 
 
 Olefines, Calorific Value 146 
 
 Olefines, Ignition Temperature of 196 
 
 Olefines, Weight of 1 10 
 
 Otto Gas Engine 9 
 
 Otto and Langen Engine 3-7 
 
 P 
 
 Palatine Gas Engine 79 
 
 Petroleum, Constituents and Data 206, 208 
 
 Petroleum, Dis jovery of 205 
 
 Petroleum, Distillation of 206 
 
 Phosphorus, Casting of 1 32 
 
 Pinkney's Self-starter 56 
 
 Pipettes for Gas Analysis 128 
 
 Piston Pin Sizes 175 
 
 Piston Speeds 70, 161 
 
 Planimeter, Amsler's 150 
 
 Premier Gas Engine 61 
 
 Priestman Oil Engine 215 
 
 Priestman Oil Engine Trials 21y. 220 
 
 Prud ucer Gas 176 
 
 Products of Combustion 183 
 
 Products, Effects on Time of Explosion 192, 199 
 
 Proportions of Explosive Mixture 190 
 
 p 
 
 Reducing Gear for Indicators 109 113 
 
 Revolutions, Counting ' ' ' ' '123 
 
 Robey Gas Engine ) ." j 4 
 
 Rope Drive .'. 50 
 
 s 
 
 Samuelspn's Oil Engine 220 
 
 Scavenging, Advantages of " " i4 
 
 Scavenging, Crossley Engine 33 
 
 Scavenging, Premier 04, 65, 67 
 
 Self-starter, Clerk 75 
 
 Self-starter, Clerk-Lanchester ' ] 73 
 
 Self-starter, Crossley ' 32 
 
 Self-starter, Fielding 61 
 
 Self-starter, Green 73 74 
 
 Self-starter, Stockport ' ' 39] 42 
 
 Self-starter, Tangye 54J 56 
 
 Silencers 14 
 
 Simplex Gas Engine gg 
 
 Simplex Oil Engine .'.".'." 213 
 
 Simplex Indicator ' ' ' IQS 
 
 Steam v. Gas " 2 
 
 Stockport Gas Engine 36, S7 
 
 Stratification 28 
 
 Street's Patent '.'.'.'.'. 2 
 
 Specific Gravity of Oils ' . ' ] 208 
 
 Specific Heat by Grashoffs Formula [\ 153 
 
 Spiel Oil Engine '....'. 211 
 
256 INDEX. 
 
 T PAGE 
 
 Tabor Indicator 101 
 
 Tandem Engines 51, 52 
 
 Tangye's Gas Engine 53 
 
 Tangye's Oil Engine 236 
 
 Temperature of Jacket Water 110 
 
 Temperature of Ignition of Carbon-monoxide 19(5 
 
 Temperature of Ignition of Hydrogen 196 
 
 Temperature of Ignition of Marsh Gas 196 
 
 Temperature of Ignition of defines 196 
 
 Temperature of Explosions 151, 196 
 
 Temperature of Exhaust 124 
 
 Testing Gas Engines 2, 114, 203 
 
 Testing Indicators 97 
 
 Testing Oil Engines 241 
 
 Time of Explosions 163 
 
 Time Recorder 1ST 
 
 Timing Valve Adjustment 40 
 
 Trials of Gas Engines 92, 114, 203 
 
 Trials, Calorimeter for 117 
 
 Trials, Counting Revolutions 123 
 
 Trials, Gas Measurements 117 
 
 Trials, Jacket-water Measurement 115 
 
 Trials, Sampling Gas 123 
 
 Trials, Temperature of Exhaust 124 
 
 Trusty Oil Engine 228 
 
 Two-cycle Engines 46, S3 
 
 u 
 
 Unit of Heat 120 
 
 v 
 
 Vaporisers, Classification of 245 
 
 Vaporiser, Britannia 232 
 
 Vaporiser, Campbell 231 
 
 Vaporiser, Capitaine 234 
 
 Vaporiser, Crossley 225 
 
 Vaporiser, Fielding 237 
 
 Vaporiser, Hornsby 223 
 
 Vaporiser, Priestman 215 
 
 Vaporiser, Samuelson 320 
 
 Vaporiser, Simplex 213 
 
 Vaporiser, Spiel 261 
 
 Vapoi iser, Tangye ." 2?6 
 
 Vaporiser, Trusty 223 
 
 Vaporiser, Wells 226 
 
 Velocity Through Valves 70, 168 
 
 Vibration of Gas Engines 10 
 
 w 
 
 Water Jackets 15 
 
 Water Jackets, Quantity of Water for It; 
 
 Water Jackets, General Arrangement 17 
 
 Water Formed by Combustion 245 
 
 Wayne Indicator '. 104 
 
 Weak-Spring Diagram 34 
 
 Wells Gas Engine 61 
 
 We Is Oil Engine 226 
 
 Printed by JOHN HEYWOOD, Excelsior Works, Manchester. 
 
THE. 
 
 "CAMPBELL" 
 
 GAS ENGINE 
 
 (OTTO CYCLE). 
 
 We have recently re-designed all our 
 
 and can now offer an Engine which CANNOT BE 
 EQUALLED by any other maker. 
 
 SELF STARTERS FITTED. 
 
 SOLE M .A. IKI IE IRS = 
 
 THE CAMPBELL GAS ENGINE GO. 
 
 HALIFAX, ENGLAND. 
 
We are the largest makers of Oil 
 Engines in the World. All sizes and 
 types from 1 to 50 brake horse-power. 
 
 SOLE MAKERS: 
 
 THE CAMPBELL GAS ENGINE CO, 
 
 Halifax, ENGLAND. 
 
Published simultaneously in London and Manchester, 
 
 EVERY FRIDAY, PRICE TWOPENCE. 
 THE BEST WEEKLY JOURNAL 
 
 OF 
 
 MECHANICAL AND MOTIVE POWER ENGINEERING 
 
 WRITTEN BY PRACTICAL ENGINEERS FOR PRACTICAL ENGINEERS. 
 
 The Journal has a large staff of Contributors, of special experience 
 in every branch of Engineering to which it is devoted, so that its readers 
 are kept fully informed with respect to the latest improvements and 
 most modern practice in Steam and Gas Engines, Hydraulic Engineering, 
 Electrical Engineering, Boiler Making, Machine Tools, &c. 
 
 SUBSCRIPTION IDS. PER ANNUM 
 
 (Including Free Copy of Pocket-Book), 
 POST FREE TO ALL PARTS OF THE WORLD. 
 
 Foreign Remittances by International Money Order. 
 
 THE TECHNICAL PUBLISHING CO, LIMITED, 
 
 31, WHITWORTH STREET, MANCHESTER, 
 
 SEND POST CARD FOR SAMPLE COPY. 
 
THE TECHNICAL PUBLISHING CO.'S PUBLICATIONS. 
 
 Post Free to any Part of the World. 
 
 MECHANICAL ENGINEERING MATERIALS. By EDWARD 
 
 C. R. MARKS, Assoc.M.Inst.C.E., M.I.M.E. Crown 8vo, Is. 6d., post free. 
 
 PRACTICAL NOTES ON THE CONSTRUCTION OF CRANES 
 
 AND LIFTING MACHINERY. By EDWARD C. R. MARKS, 
 Assoc.M.Inst.C.E., M.I.M.E. Crown 8vo, cloth, lettered, 2s. 6d., post free. 
 
 INJECTORS: THEIR THEORY, CONSTRUCTION AND 
 WORKING. By W. W. F. POLLEN, Wh.Sc. Cr. 8vo, cloth, 3s. 6d., post free. 
 
 THE CENTRIFUGAL PUMP, TURBINES AND WATER 
 MOTORS. By CHAS. H. INNES, M.A. Crown 8vo, 3s. 6d., post free. 
 
 THE NAVAL ENGINEER AND THE COMMAND OF THE 
 SEA. By FRANCIS G. BURTON. Cr. 8vo, cloth, price 2s. 6d. net, post free. 
 
 MODERN GAS AND OIL ENGINES. Profusely illustrated. By 
 FREDERICK GROVER, Assoc.M.Inst.C.E. Second edition, Crown 8vo., 
 cloth, price 4s. 6d. net, post free. 
 
 THE INDICATOR AND ITS DIAGRAMS : WITH CHAPTERS 
 ON ENGINE AND BOILER TESTING. By CHARLES DAY, 
 Wh.Sc. Second Edition, crown 8vo, cloth, price 4s. 6d. net, post free. 
 
 PROBLEMS IN MACHINE DESIGN. For the use of students, 
 
 Draughtsmen, &c. By CHAS. H. INNES, M.A. Crown 8vo, cloth, 
 price 3s. 6d. net, post free. 
 
 THE APPLICATION OF GRAPHIC METHODS TO THE 
 DESIGN OF STRUCTURES. Profusely illustrated. By W. W. 
 F. PULLEN, Wh.Sc., &c. Crown 8vo, cloth, price 6s. net, post free. 
 
 ENGINEERING ESTIMATES AND COST ACCOUNTS. By F 
 
 G. BURTON. Crown 8vo, cloth, price 3s. net, post free. 
 
 OPENING BRIDGES. By G. WILSON, M.Sc. Cr. 8vo, 2s. net, post free. 
 
 GRAPHIC METHODS OF ENGINE DESIGN. By A. H. BARKER, 
 B.A., B.Sc., Wh.Sc., author of "Graphical Calculus," &c. Crown 8vo, 
 cloth, price 3s. 6d. net, post free. 
 
 TTEAT AND HEAT ENGINES. A Treatise on Thermo-dynamics 
 1 as Practically Applied to Heat Motors. By W. C. POPPLEWELL, M.Sc. 
 Crown 8vo, cloth, price 6s. net, post free. 
 
 MARINE ENGINEERS: THEIR QUALIFICATIONS AND 
 DUTIES. By E. G. CONSTANTINE, Assoc.M.Inst.C.E., &c. Crown 8vo, 
 cloth, price 5s. net, post free. 
 
 THE A B C of the DIFFERENTIAL CALCULUS. By W. D. 
 
 WANSBROUGH. Crown 8vo, cloth, price 3s. net, post free. 
 
 THE MANUFACTURE OF IRON AND STEEL TUBES. By 
 
 EDWARD C. R. MARKS. Crown 8vo, cloth, price 5s., post free. 
 TURNING LATHES. A Manual for Technical Schools and Apprentices. 
 
 Edited by J. LUKIN, B.A. Demy 8vo, cloth, price 3s., post free. 
 SCREWS AND SCREW-MAKING. Including an account of the 
 
 origin of the Whitworth's, Swiss, British Association, and American 
 Threads, &c. ; with 95 Illustrations. Demy 8vo, cloth, price 3s. post free. 
 
 THE TECHNICAL, PUBLISHING CO. LIMITED, 
 
 31, Whitworth Street, Manchester, England ; 
 JOHN HEYWOOD, London and Manchester ; and all Booksellers. 
 
UNIVERSITY OF CALIFORNIA LIBRARY, 
 BERKELEY 
 
 THIS BOOK IS DUE ON THE LAST DATE 
 STAMPED BELOW 
 
 Books not returned on time are subject to a fine of 
 50c per volume after the third day overdue, increasing 
 to $1.00 per volume after the sixth day. Books not in 
 demand may be renewed if application is made before 
 expiration of loa^i period. 
 
 FEB 
 
 OCT 3 1927 
 
 DEC 1 5 1931 
 
/